The Priority of Democracy: Political Consequences of Pragmatism [Course Book ed.] 9781400840335

Citation preview

T h e P r i o r i t y o f D e mo c rac y

This page intentionally left blank

The Priority of Democracy Political Consequences of Pragmatism

Jack Knight and James Johnson

r u s s e l l sag e fo u n dat i o n • n e w yo r k princeton university press • princeton and oxford

Copyright © 2011 by Princeton University Press and the Russell Sage Foundation Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540 In the United Kingdom: Princeton University Press, 6 Oxford Street, Woodstock, Oxfordshire OX20 1TW press.princeton.edu Russell Sage Foundation, 112 East 64th Street, New York, New York 10065 russellsage.org All Rights Reserved Library of Congress Cataloging-­in-­Publication Data Knight, Jack, 1952–­ The priority of democracy : political consequences of pragmatism / Jack Knight and James Johnson. p. cm. Includes bibliographical references and index. ISBN 978-­0-­691-­15123-­6 (cloth : alk. paper) 1. Democracy—­Philosophy. I. Johnson, James, 1955–­II. Title. JC423.K575 2011 321.8—­dc22  2011003909 British Library Cataloging-­in-­Publication Data is available This book has been composed in Electra Printed on acid-­free paper. ∞ Printed in the United States of America 1 3 5 7 9 10 8 6 4 2

For Benjamin Brown Knight and for August Johnson, Jeffrey Milano-­Johnson, and Douglas Milano-­Johnson

This page intentionally left blank

Contents Prefaceix

Part One Chapter 1  •  Preliminaries   Chapter 2  •  Pragmatism and the Problem   of Institutional Design Chapter 3  •  The Appeal of Decentralization

1

25 51

Part Two Chapter 4  •  The Priority of Democracy and   the Burden of Justification

93

Chapter 5  •  Reconsidering the Role   of Political Argument in Democratic Politics

128

Chapter 6  •  Refining Reflexivity

167

Part THree Chapter 7  •  Formal Conditions:   Institutionalizing Liberal Guarantees Chapter 8  •  Substantive Conditions:   Pragmatism and Effectiveness  Chapter 9  •  Conclusion

193

222 256

References287 Index307

This page intentionally left blank

Preface

In 1927 John Dewey published The Public and Its Problems in which he complained: “Optimism about democracy is to-­day under a cloud.” This dispirited view, in large part, reflected widespread doubt—­doubt Dewey himself did not share—­about the capacity of regular men and women to act as democratic citizens. In part, skepticism about democracy reflected general concern for large-­scale demographic, technological, and economic change in the late nineteenth and early twentieth century. The despair also had several more proximate causes: a credulous American public (Dewey included) had been persuaded to support a brutal war aimed putatively at securing democracy abroad; the same public endeavored to constitutionalize moral strictures, giving rise to the protracted, ultimately insuperable, tasks of legal monitoring and enforcement that accompanied prohibition; and the same public displayed widespread skepticism—­animated to a large extent by religious fundamentalism—­not just about this or that scientific finding but about the enterprise of science itself and how it was taught in the public schools. In the face of wartime propaganda, prohibition, and fundamentalist creationism, critics of democracy no doubt felt well warranted in their skepticism. One might well say that in very broad strokes, the American political landscape has not changed terribly much in the intervening eight decades. Like Dewey, we confront ongoing, poorly justified military adventures overseas; economic dislocation and rapid technological change; extreme and growing disparities of wealth and income; judicially sanctioned infiltration of those financial resources into politics; and a poorly informed, credulous electorate confronting complex, large-­scale, entrenched social, economic, and environmental problems. Like Dewey, we have witnessed the effectiveness of wartime propaganda, the persistence of moralism in politics with its devastating consequences, and a widespread skepticism about science and its findings. There is little in all that to inspire political optimism. Yet, like The Public and Its Problems, this book is a defense of democratic politics. Indeed it is, we think, a robust defense of democracy. While Dewey regularly proclaimed his “faith” in democracy, we do not embrace that term. It is not that we lack faith in the capacity of regular people to engage in self-­government. But we recognize that in order to effectively exercise their capacity, ordinary men and women must be able to rely on the existence of certain material and institutional conditions. In that, too, we take up and extend Dewey’s pragmatist commitment to

P re fa c e

democratic politics. Although in our day, like his, optimism about democracy remains “under a cloud,” the proper response, we believe, is not to shrink from but rather to renew our commitment to and engagement in democratic politics. Reasonable people may differ about what they think a commitment to democratic politics entails. We emphasize here what we believe to be a fundamental aspect of the democratic project—­political and legal institutions that instantiate the practice of democratic decision-­making. This follows, in part, from our more general concern with the central importance of institutions, formal and informal, in social life. As pragmatists, we are concerned about the consequences of our choices and actions. Social institutions are one of the most important determinants of those consequences. They coordinate all aspects of our social lives. Without them we would fail in the pursuit of our goals—­not only our collective goals but also, to the extent that they rely on the actions of others, our individual life plans. With them we may achieve success, but only when we remain attentive to the various ways in which they affect us and, perhaps more important, to the conditions necessary for us to enjoy the positive benefits of those institutions. Other advocates of democracy may place their primary emphasis elsewhere, but we would argue that any compelling attempt to justify democracy must give pride of place to the formal institutions that facilitate democratic governance. As we argue, the formal institutional rules set the terms and conditions of democratic participation and practice and, in doing so, vitally influence the quality and effectiveness of our collective decision making. When, in our efforts to defend democratic politics, we neglect to account for how we will actually implement our most basic values and commitments, for how we will institutionalize the abstract democratic ideal, then we fail to make the most persuasive case for the priority of democracy. Too often we focus on the question of why we should act democratically at the expense of the equally important explanation of how we should do so. One final matter bears comment. While our argument is pragmatist, we rely at key junctures on the apparatus of rational choice theory. This confluence will no doubt puzzle more traditional pragmatists and aficionados of rational choice alike. We have two reasons—­one critical, one constructive—­ for proceeding as we do. On the critical side, political theorists typically have marshaled the results of rational choice theories to cast what they take to be grave doubt upon the viability of democracy. Their arguments trade upon particular, contestable interpretations not only of democracy but of what various families of rational choice models are useful for and what they tell us. One principle subtext of this book is a more or less systematic challenge to

x

P re fa c e

those interpretations. We believe that we can develop that challenge best by operating on intellectual terrain that such skeptics themselves will find conducive. In other words, we wish to assume the burden of argument and show why skeptics about democratic politics and institutions place inordinate faith in the results of rational choice theory. This is not because we find rational choice analysis problematic, but because we interpret it in ways that, while nonstandard, are, we think, persuasive. On a more constructive note, our own interpretation of rational choice models itself is inspired by Dewey’s concerns in the opening chapter of The Public and Its Problems. There, Dewey discusses the difficulties involved in discerning the consequences, especially the indirect consequences, of interdependent action, and he does so in the context of reflections on observation, causality, and interpretation. The difficulties that preoccupy Dewey confront both political agents and social scientists. He concludes by holding out the need for “a method which proceeds on the basis of the interrelations of observable acts and their results.” From our perspective, the related families of rational choice models afford just such a method. They provide an instrument that is useful for analyzing the interdependencies that call publics into existence—­interdependencies that, especially when they go unrecognized or unexamined, generate the problems that publics rely on democratic politics to surmount. ❏ It has taken us a long time to write this book. Over the course of many years we have, jointly and individually, incurred many debts. We would like to acknowledge those debts and thank those who have supported us in our work. We first acknowledge some joint debts and then take turns extending some individual appreciation. Joint Acknowledgments. We have, singly or together, presented material from this book to a large number of audiences in various departmental seminars and conference panels. These opportunities almost invariably generated healthy skepticism among members of the audience and, thereby, useful comments, questions, and suggestions. We will not try to single out particular individuals but instead simply acknowledge the extent to which we have benefited from engagement with colleagues over the years. We spent the 2002–­2003 academic year as Visiting Scholars at the Russell Sage Foundation. Eric Lomazoff provided extremely helpful research assistance during our year at Russell Sage. By the end of that time, we had a nearly complete draft of the manuscript. We both also realized that the draft was

xi

P re fa c e

inadequate and required substantial reorganization. We were right. But the resulting delays as we reworked the manuscript have tried patience all around. We are grateful to the foundation for funding our work. And we thank Suzanne Nichols and Eric Wanner for their forbearance as we worked things out. At Princeton University Press, Chuck Myers has been an efficient, encouraging, and patient editor. He has overseen the entire process, from coaxing us to complete the manuscript through the production of the book in what we can only describe as a masterful way. We published a condensed version of some arguments from the book  in “On the Priority of Democracy: A Pragmatist Approach to Political-­Economic Institutions and the Burden of Justification” (American Political Science Review, 2007). We thank the American Political Science Association and Cambridge University Press for permission to use reworked versions of that material here. Jack Knight. When working on a project that takes as long as this one has, one inevitably taxes the patience and goodwill of many friends. Or, at least, that is my excuse for many contrarian moments over the last several years. Some deserve special mention for their good nature and support: James Bohman, John Bowen, Randy Calvert, Ingrid Creppell, Alfred Darnell, Lee Epstein, Henry Farrell, Wayne Fields, Marilyn Friedman, Mitu Gulati, Clarissa Hayward, Mark Hornung, Ellen Kennedy, Margaret Levi, Susan Liebell, Frank Lovett, Ian MacMullen, Larry May, Gary Miller, Douglass North, Sunita Parikh, Andrew Rehfeld, Norman Schofield, Melissa Schwartzberg, Itai Sened, Kit Wellman, and James Wertsch. I also want to thank many people in the administration and staff at both Washington University in St. Louis and Duke University for helping me with support and assistance during the time that I was working on this project. I met Jim Johnson in the fall of 1979 in Hyde Park. We began a conversation about pragmatism and democracy that was challenging and stimulating in an uncommon way. Anyone would be lucky to have had that intellectual experience for a short period of time. I have been fortunate enough to have had it for over thirty years. This book is, in an important way, merely a statement of where we stand in that discussion as of 2011. My greatest personal debt is to Margaret Louise Brown. As a pragmatist I can only imagine what the consequences would have been for me if I had not settled down and finally finished this book. Even the most tolerant person has her limits. She has always believed in the value of this project, even on those days on which I questioned it myself. Without her support and encouragement I am sure that I would have set this project aside long before it was

xii

P re fa c e

finished. Margaret Louise and I have been true partners in every endeavor we have undertaken over the last twenty years. I look forward to many more. I started this book in the years before my son, Benjamin Brown Knight, was born. He is now seven. During those seven years, the social, political, and economic events that have occurred in the world would challenge even the strongest commitments to democratic self-­government. But my commitment has not diminished even if the contours of that commitment may have changed. I am still not sure about that. What I am clear about is that my perspective on the underlying questions in this book has changed significantly since his birth. I guess that it is trivial to say that as a parent I now assess the implications of social events in terms of the future implications for his life as opposed to mine. There are days in which I have serious concerns about that future, but they only arise in the context of more general concerns about the prospects for real and vibrant democratic politics in the United States and around the world. I began this book for me, but I finished it for Benjamin. Jim Johnson. My longest-­standing debt is to my parents, Roy and Janet Johnson, who long ago put aside their qualms and supported my, to them nearly incomprehensible, decision to pursue an academic career. I thank them for that. My sister Cindy, her husband Fred, and their now-­grown children Matthew and Amelia have tolerated my inscrutability and preoccupation. Seeing this book might make them wonder about me even more. But hopefully it will help them see what I think is important. By the time this book appears, I will have taught at the University of Rochester for two decades. I thank two deans—­Thomas LeBlanc and Peter Lennie—­ for funds that have supported my research on the book. During my time at Rochester, the staff of the Political Science Department—­Pamm Ferguson, Lisa Zimmerman, Donna Smith, and Kelly Corbett Wicks—­have been unfailingly helpful. I thank them for the many ways they have made it easier to do my job. For four years, I also had the good fortune to work with Linda Lindenfelser. Her hard work, good humor, and kind words did not just make the task of being a journal editor easier; they made it possible. Finally, at different times Caroline Kobick and Heather Winegrad provided welcome help with research and with preparing the manuscript. During those two decades, I also have experienced considerable mayhem in my personal life. This is not the place to either recount the details or identify the cast of characters. I mention those troubles only because, for a significant portion of that time Susan Horwitz proved to be a terrific “shrink.” Susan listened carefully, told me when I needed to cut the bullshit and take responsibility, pushed hard to help me see the very many points where responsibility

xiii

P re fa c e

rested with others—­usually places where I had neither any control nor hope of it—­and helped me figure out how to navigate all that. She also laughed easily, talked about the travails of juggling research with seeing clients, bragged about her students and her own grown children, and gushed about her baby grandsons. I am sure she would be pleased to see this book finally appear in print. Susan died suddenly in the Fall of 2009. So, while my thanks come too late, they are no less heartfelt. I could not have kept working on this project without Susan’s insight and support. Three other people in particular deserve my special thanks. The first is Robert Westbrook, my colleague in the Department of History at Rochester. He has encouraged my interest in pragmatism and, twice, has kindly allowed me to sit in on his graduate seminar on the topic. It is impossible to overstate my respect for Robb or the influence he has had on my thinking. The second person is my friend and coconspirator, Jack Knight. From our first days in Pick Hall to the present, Jack has been not just an unfailing friend but a model intellectual colleague. Over the years I often worried aloud to him that perhaps he and I are the only living creatures who might agree about some point, political or theoretical. At this juncture I worry more when we disagree! This book is, from my perspective, a joint project in the true sense. There are very few sentences in what follows that might be attributed to either of us alone. The third person I want to single out is Susan Orr. I will say only that we share our lives. She has been steadfast through the cold weather, the hard work, and the mayhem. For that I am grateful. By the time this book appears, we hopefully will have set off on a whole new, exciting adventure together. I look forward to that with pleasure. When you look at the big picture, I am a lucky man. I have three sons. Nearly anything I say about my boys and their profound effect on me will be inadequate. Simply put, they are the love of my life. And they have shown me how important it is to keep one’s eye on the consequences and how to tell the difference between those that are important and those that are not. The worst of my personal mayhem, by far, came in April 2007 when my middle son Jeffrey died at age 14. Jeff was as smart and talented and good-­natured a son as a father could ask for. Sometimes his goofiness drove me crazy. I miss him every single day and wish he were here to hold this book in his hands. My youngest son, August, was born in February 2006. He is a sweet, bright, mischievous little boy who, to my amazement, wakes each morning with a grin. At this point, at least, he rightly is oblivious to my obsessions with the politics and theory of democracy. I am sure that will change soon enough. Watching August grow is a continuing delight. Finally, my oldest son Douglas has grown to be a decent,

xiv

P re fa c e

accomplished young man. He has navigated the foibles of adults in his life (me included) and endured his brother’s death with remarkable strength and maturity. I admire him more than he can imagine. I also am pretty sure he figured he’d never see the day that this book would appear. I dedicate this book to my three wonderful sons.

xv

This page intentionally left blank

Part One

This page intentionally left blank

Chapter 1 Preliminaries

I.  Preliminaries Politics, in large part, is a response to diversity. It reflects a seemingly incontrovertible condition—­any imaginable human population is heterogeneous across multiple, overlapping dimensions, including material interests, moral and ethical commitments, and cultural attachments. The most important implication of this diversity is that disagreement and conflict are unavoidable. This is, in part, not only because the individuals and groups who constitute any population are diverse in the ways just suggested but because that diversity is irreducible. There simply is no neutral institutional arrangement that will accommodate their competing demands and projects without leaving some remainder over which they still disagree. But the inevitability of disagreement and conflict also partly reflects the fact that precisely as members of a relevant population, those individuals and groups are commonly, as it were, stuck with one another. Their fates are highly, irrevocably interdependent. Thus, despite their diversity and the discord to which it gives rise, they require some means of coordinating their ongoing social and economic interactions. For this, they need, most importantly, social institutions. More specifically, in their efforts to coordinate in mutually beneficial ways, they require a set of institutional arrangements consisting of everything from the most decentralized institutional mechanisms, like informal norms, practices, and conventions, to a wide range of decentralized (e.g., markets) and centralized (e.g., government) formal institutions. Under these circumstances of discord and interdependence, politics largely consists in deep, persistent contests over the contours and distributive implications of these shared institutional arrangements.1 As Michael Walzer rightly suggests “opposition and conflict, disagreement and struggle where the stakes are high—­that’s what politics is. ” He returns to this theme shortly thereafter, refining his point: “[C]ompetition for power is the primary form of political life . . . it can take different forms. The basic point is that without groups in conflict, there would be no politics at all, or nothing that we would recognize as politics” (Walzer 2004, 117–­18, 128 [emphasis added]). From a considerably different point of departure, Raymond Geuss arrives at much the same destination. “Politics” he suggests” is about conflict and disagreement, and this means not only 1  

C h ap te r 1

Institutions are sets of rules that emerge from and subsequently structure social and political interaction. They are persistent means of coordinating ongoing social, economic, and political activity. And typically they have a systemic quality such that in any particular circumstances, what we call an institutional arrangement will hang together in a more or less coherent, if more or less arbitrary, and more or less contested fashion. Institutional rules often demand that individuals act in ways that run, wholly or partly, counter to their immediate or even longer-­term interests, commitments, and attachments. They therefore must specify not simply what can be done, by and to whom, for what purposes, and when but also what happens when the rules are breached and who decides whether a breach has occurred. On this account, institutions ultimately must be self-­enforcing—­that is, they must rest on the mutual expectations of the relevant participants.2 For most forms of social interaction, there will typically be several, perhaps many, alternative and feasible ways to institutionalize that interaction. For example, if we want to establish an institutional framework to coordinate the multiplicity of social interactions involved in the production and distribution of health care, there are several types of arrangements that could accomplish it. Each will differ in the degree to which they rely on decentralized, as opposed to centralized, mechanisms. In such situations, we encounter further unavoidable sources of disagreement. Institutional arrangements are indeterminate in at least two ways. They are indeterminate in the sense that they represent arbitrary outcomes of strategic interactions. Given that there are alternative ways of institutionalizing social and economic affairs, relevant individuals and groups will endorse arrangements that they expect to operate over time in ways that favor their own interests, commitments, and attachments. They also are indeterminate in their operation insofar as the individuals and groups for whom they are relevant will differ over time about what the rules mean, whether and when they are being followed or breached, who is to decide such matters, and so on. Here again, involved parties will seek to resolve such interpretive disputes to their own advantage.3 that parties will disagree but also that they will have a motivation to exploit existing conflicts or ambiguities in shared beliefs and values.” One implication is “that the idea of a community in which people are clear and in agreement about what they want is not an ideal of a political system: politics is about getting things done when people do not agree” (Geuss 2001, 5–­6, 117). 2   Because there typically will be several, perhaps many, feasible, alternative ways to institutionalize social and political interaction, any given institutional form and the larger arrangements of which it is a part represent large-­scale coordination equilibria in the game-­theoretic sense. See Calvert (1995a; 1995b). 3   Epstein and Knight (1998); Calvert and Johnson (1999).

2

P re limin a rie s

The view we sketch here clearly converges with the way others portray “the circumstances of politics.”4 It just as clearly departs from the views of others who hope to distance themselves from the sort of “political theory that presumes conflict and competition as characteristic modes of interaction.”5 That said, people confronting the “circumstances of politics” on our view share a judgment or recognition or assessment that even as violence, coercion, and force constitute the ubiquitous backdrop to politics, they do not afford effective means of coordinating ongoing interactions. In this sense, these circumstances are themselves something of a tenuous achievement in the absence of which social actors might adopt one of (at least) two broad alternative strategies. The first would be exit, whether individual (via emigration) or collective (via secession). The second is violence (e.g., “ethnic cleansing,” genocide) aimed at eliminating or expelling others whose interests, commitments, and attachments animate projects that appear threatening or even just inconvenient. At this juncture, a skeptic might respond by noting that the historical record is replete with instances in which people adopted either of these two alternatives. We concede the point but suggest the views we advance hardly are unique in being susceptible to the challenge posed by the well-­documented willingness of individuals and groups to fundamentally abandon politics in favor of exit or violence. The skeptical view is a challenge to any general argument about the normative significance of political-­economic institutional arrangements. It raises a general question: What might persuade people to eschew direct recourse to violence and coercion and to accept institutionalized ways of resolving conflict and achieving cooperation? Our answer to this question is that democratic politics, and especially reliance on democratic institutional arrangements, afford the best prospect for doing so. John Dewey’s perspective on the relationship between the use of force and the use of democratic arrangements is especially informative in this regard. Dewey acknowledged that force was a fundamental characteristic of social activity, asserting that “[n]o ends are accomplished without the use of force.”6 From this premise, he argued that the important social questions involved how force is marshaled and employed in the pursuit of social cooperation. The relevant criterion was “the relative efficiency and economy of the expenditure of force as a means to an end.” Thus, “what is justly objected to as

Waldron (1999, 102–­8); Weale (1999, 8–­13). Young (1990, 228). 6   Dewey, (1980, 248); the quoted passage is from a 1916 essay entitled “Force and Coercion.” 4   5  

3

C h ap te r 1

violence or undue coercion is a reliance upon wasteful and destructive means of accomplishing results.”7 For Dewey, institutional arrangements like democracy and the rule of law offered a practical and collectively beneficial alternative to the use of violence.8 “If law or rule is simply a device for securing such a distribution of forces as keeps them from conflicting with one another, the discovery of a new social arrangement is the first step in substituting law for war.”9 In fact, the benefits of institutional solutions to conflict that avoid the use of violence play a fundamental role in Dewey’s argument in support of democracy: Democracy is the belief that even when needs and ends or consequences are different for each individual, the habit of amicable cooperation—­ which may include, as in sport, rivalry and competition—­is itself a priceless addition to life. To take as far as possible every conflict which arises—­and they are bound to arise—­out of the atmosphere and medium of force, of violence as a means of settlement into that of discussion and of intelligence is to treat those who disagree—­even profoundly—­with us as those from whom we may learn, and in so far, as friends. A genuinely democratic faith is faith in the possibility of conducting disputes, controversies and conflicts as cooperative undertakings in which both parties learn by giving the other a chance to express itself, instead of having one party conquer by forceful suppression of the other—­a suppression which is none the less one of violence when it takes place by psychological means of ridicule, abuse, intimidation, instead of by overt imprisonment or in concentration camps.10 Yet, few if any have followed Dewey’s lead in adopting this relatively unromantic view of democratic politics. Normative political theorists tend to ignore or assume away the challenge posed by those who would repudiate politics.11 There tends to be a common assumption that our debates about the normative legitimacy of democratic decision making, as well as other types Dewey (1980, 249); the quoted passage is from a 1916 essay entitled “Force and Coercion.” Recall, that Dewey (1927, 207) agrees with Walter Lippmann that “the ballot is, as often said, a substitute for bullets.” He simply, and rightly, thinks that democracy involves more than casting and counting votes. 9   Dewey (1980, 214); the quoted phrase is from an essay entitled “Force, Violence and Law.” The clear contrast to Dewey’s reference to friendship here is Carl Schmidt, for whom the category of “enemy” constitutes the realm of politics. It also is clear that Schmidt’s own views carried him over to “political” positions that, in fact, eschewed politics in favor of violence. 10   Dewey (1939/1993, 243) 11   For exceptions, see Plotke (2006) and Geuss (2008). 7   8  

4

P re limin a rie s

of institutions, take place exclusively among participants already committed to the nonviolent resolution of disagreements and differences. And even here we admit that this challenge will not be the primary focus of our argument. Nonetheless, we feel that our emphasis on the persistence and ubiquity of disagreement and conflict gives politics more room in our discussion than is to be found in most normative theories. And it will allow us to return later, in the spirit of Dewey’s analysis, to the question of whether the institutional argument we advance—­if successfully implemented—­would diminish the likelihood that significant actors would opt for nonpolitical means and, crucially, what would happen if it did not.

II.  Institutional Pluralism As we have said, any population inhabiting the circumstances of politics as we have sketched them can avail itself of a plurality or range of feasible institutional forms. In addition to markets and democratic decision-­making procedures, this includes, but is not limited to, bureaucracy, adjudication through courts, private associations, economic hierarchies, and social norms. And, of course, these institutional mechanisms, which themselves come in different varieties or forms, also need to be combined in various ways into what we will call “institutional arrangements.” Once we recognize this plurality, the task of choosing which institutions would best coordinate our social interactions in any particular setting appears quite daunting. This in turn raises the difficult task of discerning how, in such circumstances, any heterogeneous constituency, with its diverse and often conflicting interests, values, and commitments, might determine which array of institutional forms to use to coordinate ongoing interactions across various social domains. We offer some examples in the next section. It is, in any case, likely that members of the population in question will disagree pointedly and legitimately on this matter. In other words, the choice of which institutional form or combination of forms to rely on in which domains is a political problem whose resolution in any given circumstances will be contested. Determining which institutional arrangement is appropriate for which purposes turns out to be a very complex task. As standard textbook accounts make clear, in the case of markets, for example, any form of institution will operate effectively and thus generate normatively attractive outcomes only under particular (and, in principle, specifiable) conditions. Not only will the effectiveness with which any institution operates be a function of the extent to which these conditions in fact obtain (e.g., the structural and participatory conditions

5

C h ap te r 1

necessary for effective market competition), it may well be a function of social and cultural norms that characterize any particular society.12 Thus, in any particular context, the task of institutional creation involves not only the selection of one among several available institutional forms but also the identification of mechanisms needed to monitor whether the conditions necessary for effective institutional performance are being adequately fostered and sustained. This point is general. It does not apply only to the selection or design of economic institutions. The effectiveness of any type of institution will be largely a function of the conditions in which it operates. This remains obscured because social scientists have undertaken considerably less systematic analysis of the preconditions for the effective performance of institutions of democratic, bureaucratic, or judicial decision making than has been undertaken on economic transactions in markets. Nonetheless, the task of institutionalizing more centralized arrangements requires that we make a commitment to establishing and maintaining our best understanding of the conditions that make bureaucracy or democratic governance or courts effective in achieving the goals we impute to them. This obviously places a considerable burden on the members of any society in terms of information and knowledge about institutional performance and effects. On our account, this is where experimentalism emerges as central to politics. Institutional experimentation is a useful instrument for generating knowledge about the effectiveness of institutions in various social contexts.13 Theoretical analysis can make important contributions to our understanding of issues of institutional performance, but it alone cannot provide definitive answers to questions regarding the actual effects of adopting one or another institution. Such knowledge will reliably emerge only from the cumulative experience of using for various purposes the various institutions at our disposal. We thus need to be able to rely on the lessons such experience affords us. In the end, we will trust or value experimental outcomes insofar as they emerge under proper conditions, and we must therefore expend considerable effort in establishing and monitoring those conditions. The importance of “proper conditions” highlights the fact that experimentalism is itself an institutional choice. Institutional experimentalism will generate useful knowledge only insofar as the conditions are in place to let the experimental process itself operate in effective ways. To foster and maintain these conditions, we require some kind of institutional framework of monitoring and For a compelling argument to this effect in regard to economic institutions and development, see Rodrik (2007). See also Satz (2010) and Lindblom (2001). 13   On this general point, see Ostrom (1990); Rodrik (2007); Unger (2005; 2007). 12  

6

P re limin a rie s

assessment. Here the logic of institutional experimentation starts to resemble an infinite regress of institutions and conditions and meta-­institutions and meta-­conditions, and so forth. Yet, in this way, it nonetheless clarifies the task before us: to avoid the regress, we must identify some institutional arrangement (or set of arrangements) that can serve as a mechanism for (1) coordinating effective institutional experimentation, (2) monitoring and assessing effective institutional performance for the range of institutions available in any society, and, most importantly, (3) monitoring and assessing its own ongoing performance. Our argument, unsurprisingly, is that we should accord priority to democracy precisely to the extent to which it is adept at performing these tasks.

III.  Examples and Comparisons We speak throughout this book of institutional pluralism or diversity. On the view that we advance, institutions consist of sets of rules that emerge from and subsequently structure social and political interaction. They are persistent, systematic means of coordinating ongoing social, economic, and political activity. Institutions specify what can be done, by and to whom, for what purposes, and when, as well as what happens when the rules are breached and who decides whether they are. Recognizing institutional pluralism simply means grasping that in any given instance, there may well be various ways to structure ongoing social, economic, or political interaction. Moreover, there typically is no general criterion for deciding among the sets of institutional rules on offer. Given the problem of establishing, monitoring, and assessing ways of coordinating ongoing interactions in various domains, consider just a small handful of the institutional arrangements we might adopt for just a limited number of purposes. Economic Exchange. In many instances, a population will confront the choice between trying to regularize commercial transactions by creating and sustaining a market for particular goods and services or trying to block such transactions. For instance, the population might confront the problem of dealing with the selling and buying of votes. It might create a market in votes, or it might adopt the secret ballot as an institutional mechanism that can effectively block such exchanges by preventing voters from revealing to potential buyers how they cast their vote.14 Schelling (1960, 19) offers this interpretation of the secret ballot. Of course, the same population might opt, for other reasons, for public voting. Brennan and Pettit (1990) make a case for the latter. 14  

7

C h ap te r 1

The primary point of the secret ballot is to prohibit trafficking in the “good” altogether.15 In other instances, a population may want to facilitate the availability of some crucially important, scarce good or service and thus face the problem of whether or not to have recourse to markets for that purpose. In other words, it wants to allow exchanges but not allow them to be commercial in the sense that they involve buying and selling. Consider, for instance, human blood.16 A population might opt to allow for voluntary contributions (“gifts”) of blood for medical purposes and establish an official agency to collect, process, and distribute it. Or it might encourage markets in which blood is procured and circulated via commercial transactions and distributed by private entities, whether they be for-­profit business corporations or not-­for-­profit agencies like the Red Cross. Distribution of the Franchise. Any democratic institutional arrangement confronts the problem of how to distribute voting rights. If we focus solely on nation-­states, it turns out that participation is partitioned across multiple dimensions. Virtually all countries exclude individuals under some minimum age (usually, but not always, eighteen) and those with a documented mental disability. But while there is imperfect convergence on those exclusionary rules, there is more or less wide cross-­national divergence on whether voters must be citizens, whether they must reside in the country, whether they must reside in the electoral district in which they vote, and whether prison inmates can vote.17 In short, even if we set aside variation over time, the rules governing inclusion and exclusion from this minimal political right differ considerably. Constitutional Politics. A central feature of many contemporary constitutions is some form of separation of powers among legislative, executive, and judicial branches. Such arrangements inevitably generate problems of what to do in cases of disagreement or conflict among branches. Such conflict might be resolved by force (e.g., a coup) or by any of a range of political gambits (e.g., impeachment or defunding). Either manner of dispute resolution could evolve into an informal norm to govern ongoing interbranch interactions. But such conflict might also be resolved by accommodation. Among the most common forms of the latter is the institution of judicial review, which affords constitutional courts the prerogative of ruling on the legality of acts of the 15   This is a contentious claim. Recall that we are simply illustrating the diversity of institutions. For present purposes we set aside other, critical interpretations of the secret ballot and its uses. See Bertrand, Briquet, and Pels (2006). 16   Titmus (1971/1997); Arrow (1972); Singer (1973). 17   Massicotte, Blais, and Yoshinaka (2004, 15–­39).

8

P re limin a rie s

executive and/or legislature.18 But even if judicial review either emerges informally or is intentionally adopted, a range of possibilities remains; judicial review can itself take different forms.19 Democratic Decision Making. Judicial review is rightly viewed as antimajoritarian and therefore as nondemocratic. We sidestep the ongoing debates on that point. Instead we simply point out that democratic polities often incorporate a range of “submajority” voting rules into their decision-­making institutions. Such a rule “authorizes (i) a predefined numerical minority within a designated voting group (ii) to change the status quo (not merely to prevent change) (iii) regardless of the distribution of other votes” in the body.20 Just in the context of the U.S. federal government, the following examples meet these criteria: The Journal Clause which allows one-­fifth of the legislators present in either House of Congress to call for a roll call vote; The “discharge rule” in the House of Representatives, which (at various points, although not today) has permitted a specified minority of legislators to force bills out of committee for consideration on the floor; Senate Rule XXII, under which a cloture petition is valid when signed by sixteen Senators; The “Seven Member Rule” under which a minority of designated committees in the House and Senate can require the executive branch to divulge information; House Rule XI, which entitles committee minorities to call witnesses at hearings; The famous “Rule of Four” that allows four Justices to grant a writ of certiorari and thereby put a case on the Supreme Court’s agenda.21 So, not only is there, as is commonly recognized, a variety of voting rules that incorporate the principle of majority rule, there are a number of operative voting mechanisms that make no pretense of so doing. Property Rights and Common Pool Resources. Markets are not adept at providing or distributing public goods, that is, goods characterized by joint

18   This paragraph draws on the account of how judicial review emerged in the United States advanced by Knight and Epstein (1996). 19   See Vanberg (1998) for the operation of “abstract” as opposed to “concrete” forms of judicial review. 20   Vermeule (2007, 87). 21   Vermeule (2007, 186–­87). Compare Guinier (1995).

9

C h ap te r 1

supply and nonexcludability.22 Neither are they useful for coordinating interactions on what are called common-­pool resources, which consist in “a natural or man-­made resource system that is sufficiently large as to make it costly (but not impossible) to exclude potential beneficiaries from obtaining benefits from its use.”23 Pertinent examples here are fisheries, irrigation systems, grazing lands, and forests. Given such a resource, a population might want to establish an official, more or less centralized institution to monitor its use and enforce restrictions on those who use it. In some contexts, such a centralized solution may be appropriate, but there are many reasons why a population might deem a more local, less centralized arrangement preferable.24 Having reached that conclusion, however, the population still confronts a choice between relying on some variety of privatization and one or another regime of common property. In other words, a population deciding on how to govern a common pool resource faces a complex problem. “Instead of there being a single solution to a single problem . . . many solutions exist to cope with many different problems.”25 Neither the state nor a regime of private property rights is always and everywhere appropriate to the governance of common-­pool resources. There are other options.26 These are just a few examples of situations where institutional diversity makes itself felt. We might well extend the catalog but believe we have made the basic point. Rather than offer more examples, we now briefly want to situate our approach to the matter of institutional pluralism in the context of some provocative recent research that also addresses the problems it raises. The studies we have in mind are Democracy and Knowledge, by classicist Josiah Ober, and One Economics, Many Recipes, by political economist Dani Rodrik. These works offer robust defenses of democratic practices and institutions in terms of their consequences for macro-­level performance. Ober aims to explain the success of Athens relative to other Greek poleis in terms of Knight (1993). Ostrom (1990, 30). 24   Ostrom (1990); Taylor (1987). 25   Ostrom (1990, 14; 2005). 26   Ostrom (1990, 12–­13, 22, 60–­61), like Rodrik, warns against what the latter calls “property rights reductionism,” a position that “views the formal institutions of property rights protection as the end-­all of development policy” (Rodrik 2007, 184, 155–­57). Both are concerned, by contrast, with the task of establishing effective institutions, which requires looking at local problems in ways that will allow analysts and participants to discover what might count as effective. Both, for instance, countenance the possibility that common property rights can sustain effective control in instances where state authority or privatization might not. 22   23  

10

P re limin a rie s

political-­economic and military competitiveness. Rodrik analyzes the factors that account for economic growth, and thereby political-­economic development, cross-­nationally. Despite their seemingly divergent preoccupations the books share several features. Both Ober and Rodrik recognize the exigencies of inferring from particular evidence—­whether case-­based or statistical studies—­­to broader conclusions regarding political and economic institutions.27 Both promote the importance of experiment in face of the tendency to premise institutional analysis on formal “blueprint” designs.28 And both gesture toward a pragmatist justification for their arguments.29 For both Ober and Rodrik, democratic institutions are crucially important. Both focus on the capacity of democratic institutions to coordinate and communicate socially dispersed or “distributed” knowledge in the face of difficult, persistent, large-­scale problems or social, economic, and political interaction. Ober attributes the relative success of Athens to the ways its democratic arrangements facilitated “the aggregation, alignment and codification of useful knowledge,” while Rodrik argues that economic growth, and hence economic development, requires reliable intuitions for “processing and aggregating local knowledge” and insists that democratic institutions “are the most effective mechanism” for accomplishing that task.30 As will become clear, we find this general line of argument congenial. Yet we think it is, by turns, troubling and incomplete. We find the arguments Ober and Rodrik advance troubling to the extent that they tend toward a functionalist account in which political institutions emerge because they are “efficient” or are “market supporting.”31 Indeed, if institutions—­including markets—­ are themselves public goods, and if we assume in our explanatory account that agents are, if not opportunistically self-­regarding, then not naively other regarding either, surely there is no reason to assume that those agents will be motivated to contribute to the provision or maintenance of that public good.32 If, as Ober reasonably suggests at various points, political agents will avail themselves of the “opportunity to use the system in ways that will promote their own personal and partial interests to the detriment of the interests Ober (2008, 30–­31, 268–­70); Rodrik (2007, 3–­4). Ober (2008, 19, 270–­72); Rodrik (2007, 163–­66). 29   Neither author develops this theme at length; they acknowledge their pragmatist propensities primarily in scattered footnotes. 30   Ober (2008, 3); Rodrik (2007, 8). 31   See Ober (2008, 12) and Rodrik (2007, 156), respectively. 32   Ober (2008, 12) makes clear that he is making fairly narrow motivational assumptions of the sort we describe. That markets are themselves public goods is a commonplace. See, e.g., Ostrom (1990, 15). 27   28  

11

C h ap te r 1

of the less powerful or the general interest” or, conversely, if institutions are susceptible to being “exploited by strategic actors for socially unproductive purposes,” what leads us to believe that in setting up those institutions in the first place, those same agents will set aside narrow preoccupations and the advantages of power to promote the general welfare or collective good?33 We offer a realist account of institutional emergence and change that makes no such presumption and that depicts institutions as by-­products of bargains between self-­interested but differentially situated actors. And we argue that that account raises a serious burden of justification for anyone who embraces it. We find the arguments Ober and Rodrik advance incomplete in the following sense. The claim, which Rodrik makes a running theme, that “democratic institutions . . . can be viewed as metainstitutions that help society make appropriate selections from the available menu of economic institutions” partially anticipates the argument we advance in this book.34 While Ober does not state his case quite so directly, his analysis of Athenian democratic institutions rides heavily on examples of how they enabled Athenians to adopt especially effective political-­economic institutions and practices (say, the owl currency or the physical design of public spaces). Yet—­quite properly, given their own substantive preoccupations—­they offer no systematic account of how democracy operates, in Rodrik’s sense, as a “metainstitution.” That said, if they are to be fully persuasive, the arguments that Ober and Rodrik advance presuppose the case for the priority of democracy we advance in this book or something very much like it. Indeed, we argue that democracy enjoys a second-­order priority precisely because it is uniquely useful in approaching the crucial, complex, and conflictual tasks involved in the ongoing process of selecting, implementing, and maintaining effective institutional arrangements. We sketch this view below and advance a systematic argument in its favor in the chapters that follow. But if we are to generalize from cases like The quoted phrases are from Ober (2008, 97, 213). Ober could respond that agents might be moved by considerations of social efficiency (say, to create institutions that mitigate or minimize transaction costs) due to the “hypercompetitive environment” the Greek poleis inhabited. In fact, he explicitly invokes the analogy of a firm in a competitive market to this effect (see Ober 2008, 80–­84). But markets are competitive and thus generate efficient results just insofar as the parties to exchanges cannot affect one another. So the analogy is flawed at best. Ober would have to establish how the sort of military competition he depicts—­in which defeat means “destruction . . . the temporary or permanent end of the polis as a physical or as a social entity as a result of the sack of the central city and/or the extermination, enslavement, or forced resettlement of the entire population” (Ober 2008, 81)—­resembles the sort of market competition that generates efficient outcomes from self-­interested interactions. 34   Rodrik (2007, 51, 8, 154–­55, 169, 182–­83). 33  

12

P re limin a rie s

the one Ober so cogently constructs, and if we are to fully appreciate the force of Rodrik’s observation that the “plausible variation in institutional setups is larger than is usually presupposed,” then, we suggest, what is called for is a theoretical account of how democracy operates in the ways they suggest. We offer just such an account.

IV.  The Tasks of Political Theory Our brief sketch of politics and institutions poses three interrelated but distinct tasks for political theorists.35 The first analytical task requires that we identify the set of feasible social institutions, examine their respective features, and delineate the conditions necessary for the effective operation of different members of the set. This amounts to an examination of possibilities—­what are the best outcomes that can be produced by a particular institutional arrangement under the various possible sets of conditions? For analytical purposes, we are unconcerned with the question of how, or under what conditions, any given institutional arrangement might actually emerge. That concern animates the second task theorists face. This properly explanatory task is pressed upon us by the first of the indeterminacies we mentioned above. If multiple feasible ways to institutionalize social interaction exist, we must account for how and why actors eventually arrive at a particular institutional arrangement. This is a theoretical rather than an immediately empirical problem. It demands that we specify mechanisms of institutional emergence and change and the conditions under which we can expect them to operate.36 These first two tasks are clearly related. They nevertheless are distinct enterprises. This becomes clear when we see that those pursuing these two “positive” tasks also unavoidably confront a third, normative, task that arises directly from the demands of analysis and explanation. Insofar as the task of analysis involves not simply identifying possible institutions but also inquiring into both their features and the conditions under which they operate effectively, theorists are involved in the comparative assessment of institutions. Thereby, we hope to be able to discriminate among the normative features of competing institutions and, in particular, to establish expectations about when such institutional arrangements will operate in more or less attractive ways. But notice the plural here. It makes the intrusion of distributive considerations We are indebted to Schelling (1978, 25–­27) for this point. On the importance of understanding explanation in terms of mechanisms and the conditions under which they operate see Knight (1995) and Johnson (2006). 35   36  

13

C h ap te r 1

unavoidable. Given that there almost always are different ways to institutionalize social interaction, and given that these will sustain often dramatically divergent distributive outcomes, we must confront the demand to explain why a relevant population arrives at one among the possible set of institutional arrangements. This raises, unavoidably, if tacitly, what we call the burden of justification. This normative challenge arises because the different explanatory accounts on offer invoke different causal mechanisms, some of which noticeably increase the burden of justification while others tacitly diminish it. Given the strategic exigencies involved in common processes of institutional emergence and change, the mere fact that an institution emerges as the stable outcome of the process in no way implies that it is justified. Indeed, those exigencies, minimally including power differentials among actors and unforeseen contingencies, often pose a heavy burden of justification. This problem commonly manifests itself in the significant discrepancy between how an institution actually emerges and operates in a particular society and the theoretical ideal of how that institution operates, as advocated by its proponents. Faced with such discrepancies, it is quite easy to misplace the burden of justification. But efforts to justify the choice of one institutional form over another cannot avoid attention to these discrepancies. We can only appreciate this difficulty and accurately assess the extent to which competing theoretical positions surmount it by attending explicitly to normative questions. This, it turns out, is a complicated matter.

V.  Situating Our Argument Given our view regarding the tasks of political theory, it turns out that the argument we present fits uneasily within existing categories. In particular, the view we just sketched leads us to believe that the now conventional dichotomy between “ideal” and “nonideal” theories is especially unhelpful. The tasks of political theory, as we sketch them, are considerably more complicated than the dichotomy allows. In the end—­indeed, well before the end—­the dichotomy between ideal and nonideal theories turns out to be a diversion that we should set aside in order to get on to important questions. One might distinguish between ideal and nonideal in more or less expansive ways. Some seek to differentiate the specification of principle from any consideration of “the facts” very broadly construed.37 We focus on the way Cohen (2008).

37  

14

P re limin a rie s

Rawls draws the distinction because it is familiar and influential.38 For Rawls, an ideal theory is one that assumes that “(nearly) everyone complies with, and so abides by, the principles of justice.” By depicting what he calls a “realistic utopia,” Rawls believes that we can establish the criteria by which to assess various forms of injustice and how they might be remedied. He presents this task as necessarily prior to any assessment of whether it might operate effectively in nonideal circumstances. We set aside the question of whether such assessments presuppose specification of an “ideal theory.”39 Instead, we concentrate on matters Rawls assumes. In particular, he assumes that matters of implementation are of secondary concern. He assumes, in other words, that given what he calls “the circumstances of justice,” it is possible that “a perfectly just, or nearly just, constitutional regime” can both be generated and sustained. This is hardly an innocent assumption. Indeed, it truncates the tasks of political theory in untenable ways. Specifically, it presumes that concern for institutions—­which, after all, is what a constitutional regime consists in—­is derivative of or parasitic on the specification of ideals. This complaint may seem misguided. After all, Rawls proposes justice as fairness as a set of principles to govern “the structure of political and social institutions.” The difficulty appears in the way proponents of ideal theory approach institutions.40 They proceed by articulating principles and then postulating a set of institutions that might best approximate what those principles demand. That, we believe, is a disabling approach to the tasks political theorists confront. Even when they disagree vigorously, political theorists acknowledge that in their joint enterprise, they confront multiple tasks.41 Indeed, disagreements among theorists consist to a significant degree in conflicting visions of what they are, or ought to be, up to. Here we wish only to reiterate that the various Unless otherwise stated, all the quoted snippets in the remainder of this section are from Rawls (2001, 12–­14). It hopefully is clear that we “pick on” Rawls and his followers here because we take his work to be enduringly important. 39   That seems doubtful. For reasons, see Geuss (2008) and Sen (2009). 40   Again, see Sen (2009), who contrasts a transcendental and a comparative approach to institutions, attributing the former to proponents of “ideal theory” while endorsing the latter. 41   Rawls (2001, 1–­5), for instance, identifies four such tasks: the “practical” task of addressing the problem of order; the task of orienting us “in the (conceptual) space, say, of all possible ends, individual and associational, political and social”; the task of “reconciliation,” by which we can come to see our current institutions and their development as “rational”; and, finally, a “realistically utopian” task of exploring the bounds of political possibility. For a contrasting list, see Geuss (2008, 37–­50). 38  

15

C h ap te r 1

tasks any theorist takes up discipline one another.42 The commitments a theorist makes when addressing any task clearly constrain the options available to her as she pursues others. Given the conception of institutions we adopt, we suggested that any political theorist tacitly confronts at least three tasks: she must account for both how institutions persist and how institutions emerge and change, and she must offer criteria for assessing institutional performance. How a theorist approaches each of those tasks—­we called them analytical, explanatory, and normative—­ places constraints on how she might plausibly approach the others. Throughout the chapters that follow, we seek to keep those constraints in mind both as we discuss the work of others and as we lay out our own argument. Here, ideal theory of the sort Rawls advances serves as a useful counterpoint. Recall that Rawls is concerned with formulating principles of justice while setting aside the tasks of showing how any institutional arrangement governed by those principles could emerge or sustain itself. In short, he neglects almost completely the explanatory and analytical tasks we sketched in the last section. It may be unfair to Rawls to suggest that he dismisses those other tasks as unimportant. Nevertheless, it remains the case that in the various reformulations of his theory that he crafted, he nowhere addresses them in a systematic way. Yet Rawlsians cannot defer indefinitely the explanatory and analytical tasks we identify. For even if we follow Rawls and adopt the reasonable rather than the rational as a standard of moral action, this in no way eliminates the sorts of strategic adjustment that demand an equilibrium settlement or something like it. Rawls himself acknowledges: “Reasonable persons also understand that they are to honor these principles, even at the expense of their own interests as circumstances may require, provided others likewise may be expected to honor them.”43 In other words, beyond the obvious point that it is not narrowly rational for one to accept the principles of justice, the question of whether it is reasonable to do so will in each instance depend on one’s expectations regarding how others will behave. How deep and pervasive the gap between the reasonable and the rational turns out to be is a crucial question. We will not explore it here. How susceptible Rawls’s conception of justice is to the potentially destabilizing pressures of strategic interdependence is one that seems much more important. It therefore seems appropriate and sensible to ask whether Rawls, having defended his principles of justice for the basic structure of society as ones that How, for instance, might a political theorist, on Rawls’s view, both promote reconciliation and explore utopian possibilities? 43   Rawls (2001, 7 [emphasis added]). 42  

16

P re limin a rie s

reasonable agents must accept, is able to plausibly explain how institutional arrangements that embody those principles might emerge or to analyze how they might be sustained in equilibrium. In short, does Rawls do more than bracket or set aside analytical and explanatory concerns? Does the way he constructs his normative argument—­as an ideal theory unconcerned with matters of implementation—­preclude addressing them in a persuasive way? It seems that Rawls—­or, now, his progeny—­is under considerable burden to argue otherwise. This is a task that, to date, Rawlsians seem to have almost wholly neglected.44 And it is a task whose importance, paradoxically enough, Rawls underscores with his tacit demand that reasonable expectations must be self-­enforcing. Our point here is simple and burden shifting. Instead of defending our argument as an exercise in nonideal theory, we have been concerned with whether ideal theory, at least as Rawls characterizes it, is a plausible undertaking. We have not claimed that ideal theory is impossible. We merely have suggested that in its Rawlsian variant, ideal theory both (1) acknowledges analytic and explanatory tasks and (2) pursues a normative project in ways that seemingly foreclose the possibility that theorists who subscribe to the latter undertaking can credibly prosecute those tasks. With those objections in mind, we proceed by setting aside the distinction between ideal and nonideal theory. Indeed, one distinctive aspect of our argument is the explicit way we not only acknowledge the multiple tasks that political theorists confront but also acknowledge how the ways we take up those tasks constrain one another in our own argument.

VI.  A Note on Method A final caveat is in order here. Several brands of institutionalism have emerged in recent years.45 We focus on rational-­choice models of institutions for three reasons. First, these models allow us to keep distinct the analytical, explanatory, and normative tasks that political theorists confront.46 In that sense, they provide a useful vehicle for our argument.

See Simmons (2010) for a defense of ideal theory that takes this observation as a premise. Hall and Taylor (1996). 46   In this regard, we hope it will be clear that we interpret the usefulness of formal models in a way quite different from those who advocate “positive political theory.” See Johnson (2010). 44   45  

17

C h ap te r 1

Second, both critics and advocates of rational-­choice analysis commonly and mistakenly suppose that it sustains substantive political conclusions quite divergent from those we defend. In particular, both presume that the results of social choice and game theoretic models underwrite a robust challenge to democratic theory.47 We consider that presumption largely misguided. So, by relying on rational-­choice models ourselves, we tacitly assume a rather substantial burden of argument. In short, we argue for the priority of democracy on theoretical terrain that many will think especially inauspicious. Finally, as our title suggests, we endorse a pragmatist approach to the tasks of social and political theory. Rational-­choice accounts of institutional emergence raise suspicions about any strategy that grounds justification in the processes of institutional genesis. Such accounts thus are perhaps surprisingly congenial to our pragmatist leanings insofar as these lead us to seek justification in a brand of consequentialism that is constrained or tempered by awareness of the conditions under which consequences are produced. Combined, these considerations amount to the following: while we find rational-­choice models useful, we offer a distinct interpretation of the enterprise as a whole.48 In that sense, we aim to place pressure on advocates of rational-­choice approaches to explore more carefully the ways their analytical proofs, their explanatory claims, and their normative pronouncements hang together.

VII.  Précis In these preliminary comments, we have characterized the circumstances of politics as we think they exist in the contemporary world. This is the terrain on which political theorists must advance and defend their analytical, explanatory, and normative claims. Diversity is a “social fact.” One of the main implications of diversity is that conflict and disagreement are unavoidable. At the same time, our lives are unavoidably interdependent in significant ways. Thus, we must find ways to coordinate our actions and resolve our differences. Because institutions are the primary mechanism for doing so, politics primarily revolves around problems of institutional choice. We develop our argument about the normative significance of democracy against this backdrop. Democracy is a set of institutions. It has an important Riker (1982) is typically taken as the exemplar here. We elaborate different aspects of this interpretation in several earlier papers. See Knight (1995); Johnson (2010); Knight and Johnson (1994; 1999; 2007). 47   48  

18

P re limin a rie s

priority among the available institutional alternatives. This priority is not of a first-­order variety: democracy is not the best way for us to coordinate all of our ongoing social, economic, and political interactions. We are often better served by relying on other institutional forms for that purpose. Rather, the priority we accord to democracy reflects its usefulness in approaching the crucial second-­order tasks involved in the ongoing process of selecting, implementing, and maintaining effective institutional arrangements. These second-­order tasks of achieving effective institutional performance involve monitoring and assessing the performance of the various social institutions—­ including democratic decision making itself—­that any particular society relies upon to coordinate its ongoing interactions. As it turns out, these tasks involve establishing and maintaining an experimental environment that can enhance our knowledge of the relationship between different institutional forms and the conditions under which they produce good consequences. We argue that democracy is especially, almost uniquely, conducive to these undertakings. The priority of democracy derives from what, on our view, are its fundamental features. We highlight three such features—­voting, argument, and reflexivity—­each of which relates to the positive effects of democratic processes on collective decision making. We argue that these effects distinguish democracy from other ways of coordinating ongoing social interaction. These qualities lend democratic arrangements presumptive priority of a particular sort. In any effort to negotiate unavoidable social disagreement over institutional arrangements, democracy enjoys a second-­order priority precisely because it operates in ways that potentially meet a heavy burden of justification. This claim will perhaps startle those familiar with the burgeoning literature on social institutions. As we demonstrate, much of that literature tacitly and improperly privileges a quite different component of our institutional arrangements, namely markets. It does so insofar as it presumes that the proper way to explain and assess other institutional forms is in light of the extent to which they are functional to the operation of markets. We argue that once the premises and argumentative strategies common to the institutionalist literature are clarified, it simply is not possible to sustain the privilege it accords to markets. In fact, we argue that the analytical models institutionalists deploy, and the most robust explanatory strategies available to them actually sustain, our case for the priority of democracy. As with other institutional arrangements, the mechanisms of democratic decision making will not operate effectively (i.e., in terms of producing positive

19

C h ap te r 1

effects on collective choice) absent a set of specific conditions.49 In the case of democracy, these conditions relate to the nature and terms of political participation. On our account, effective democratic decision making requires that each participant enjoy equal opportunity of political influence. This criterion places a substantial burden on any democratic society, as it has significant implications for how that society institutionalizes freedom and equality for its citizens. One of the interesting features of our account is that it provides a distinctive justification for freedom and equality in a democracy, a justification grounded in the positive effects such participation has on the collective outcomes of the democratic process.50 The beneficial consequences of employing democracy as a means of collective governance are substantial. Standard justifications of democracy rest their case on what we call the first-­order effects of an institution: the way in which an institutional arrangement coordinates substantive social interactions. On such an account, democracy is beneficial primarily for the task of facilitating collective decision making about substantive policy questions. The emphasis is commonly placed on the extent to which democracy can bring individuals together in the pursuit, if not of a common good, then at least of generally shared goals. The pragmatist justification that we offer here identifies a different source for the beneficial effects of democracy. On our account, the important benefits of democracy are derived from its second-­order effects: the way in which democratic institutional arrangements facilitate effective institutional choice. This shift in focus highlights the important role that democracy plays, not in achieving consensus or commonality, but rather in addressing the ongoing conflict that exists in modern society. Significant beneficial consequences accrue to societies that rely on democratic arrangements to accommodate tensions that in any event do not lend themselves to any principled and lasting resolution. Democracy provides an effective means for a society to address these ongoing tensions and to decide in any particular moment how best to resolve specific problems and issues. Our pragmatist justification of democracy supports the claim that democracy has a normative priority over other social institutions in regard to the fundamental task of effective institutional choice. A remaining question asks about the role this second-­order priority plays in arguments about (1) the 49   As will become obvious, our argument diverges from the empirical literature—­running from Lipset to Przeworski—­that probes the social and economic preconditions of democracy. 50   Put most starkly, we defend particular conceptions of freedom and equality insofar as they are required for democratic decision-­making rather than defending democracy because it is instrumental to attaining the goals of freedom and equality.

20

P re limin a rie s

legitimate authority of democratic governance or (2) the political obligation generated by democratic decisions. Stated otherwise, is there a pragmatic answer to questions of legitimacy and obligation? If there is such an answer, it must, we think, acknowledge the inescapably political nature of such questions. And, in doing so, it will also require us to rethink the emphasis on the moral dimension of state action. The key to a plausible pragmatic alternative rests in its underlying commitment to tempered consequentialism and in the role effective institutional performance plays in everyday social life. We divide our argument into three parts, each consisting of three chapters. In addition to this introductory chapter, the first section contains chapter 2, where we set out our conception of pragmatism and discuss its implications, and chapter 3, where we begin the comparative analysis of institutions by considering the strengths and weaknesses of decentralization. In chapter 2, we provide a basis for the claim we make in our subtitle, namely that we advance a pragmatist account of democracy. We specify three features—­fallibilism, anti-­skepticism, and consequentialism—­as central to pragmatism understood as a philosophical position. In that regard, our view is fairly standard. But we also make two further claims that are distinctive. We first argue that pragmatism has important political consequences and that those sustain a commitment to robust democratic politics. We then argue that insofar as pragmatists are committed to democratic politics, they necessarily are committed not just to an ethos or ideal but to the analysis of democratic institutions. Each of these arguments will come as a surprise not just to those who come to pragmatism from the outside but to many who consider themselves pragmatists. In chapter 3, we launch our comparative analysis of political-­economic institutions in what may seem like a counterintuitive manner. We take up what arguably is the default presumption in this domain, namely that the task of establishing institutional arrangements should accord priority to markets and other decentralized mechanisms. We offer an extended critical analysis of the case for markets along with an assessment of several other decentralized mechanisms for social cooperation. We argue that because the commonly acknowledged initial conditions that markets presuppose if they are to operate efficiently are quite restrictive and because markets themselves offer no mechanism for monitoring and maintaining those conditions, there is no clear reason to bestow first-­or second-­order priority on markets as a default mechanism of social coordination. We recognize that this critical argument, by itself, does not afford grounds for according priority to more centralized institutions. We therefore consider three further decentralized institutional mechanisms (Coasian bargaining, community and technocratic incentive

21

C h ap te r 1

compatibility schemes) and advance analogous reasons for withholding priority from them (singly or jointly) as well. In part, but only in part, the argument in chapter 3 is ground clearing. After all, if some combination of decentralized institutional forms can adequately address the problems of coordination that the circumstances of politics generate, then our argument for democracy might be considered redundant. Our critical assessment shows that any such supposition, however common, is mistaken. In the end, however, we also acknowledge that something like our case against granting special priority to decentralized institutional mechanisms applies to more centralized institutional alternatives as well. We consequently conclude that democracy must meet a quite substantial burden of justification if we hope to sustain our claim that it deserves priority in debates about institutional design and reform. In that sense, chapter 3 plays a constructive and fundamental role in our argument: it establishes the threshold that our own account must meet if it is to be persuasive. In the three chapters that comprise part 2, we directly take up the case for the priority of democracy. In chapters 4 and 5, we analyze the two primary mechanisms that animate democratic decision making: argument and voting. As a starting point, we assess the challenges that social choice theory brings to normative claims about democracy. Social choice theorists commonly critique democratic decision making on the grounds that voting is susceptible to unavoidable pathologies and that insofar as voting is essential to democracy, those pathologies subvert the normative legitimacy of democratic outcomes. After assessing the implications of these critiques, we consider three ways that political argument can affect democratic decision making and, thus, significantly mitigate the force of the social choice challenge. First, we analyze the potential effects of political debate or argument on the assumptions that drive social choice results. We contend that by engaging in political argument, relevant agents can settle the dimensions that, in any instance, structure their disagreements. This causal effect not only dampens the prospects that collective decision making will generate cyclical outcomes, it thereby reduces the opportunities for strategic manipulation that such instability presents. In this fashion, we show how the effects of argument, under the proper conditions, can significantly diminish the challenge that social choice theory commonly is held to pose for any robust normative commitment to democracy. In the process, we change the subject of democratic theory from showing that democracy can induce agreement to recognizing that when it operates effectively, it serves to structure disagreement. Once the analytical argument has established the possibility that voting, augmented by argument, could produce normatively legitimate decisions, we

22

P re limin a rie s

consider two ways in which democratic argument can enhance the quality of such decisions. We first present an epistemological argument about the beneficial effects of diversity. We then provide a structural argument about the beneficial effects of a central feature of democratic decision making: reflexivity. We conclude chapter 5 by bringing the three possible effects together to satisfy the burden of justification for democracy and, thus, to justify its priority among the available institutional alternatives. In chapter 6, we clarify our views on reflexivity and how it operates in democratic arrangements by considering a set of possible objections to our argument. We first address the potential objection that we have underestimated the capacity of decentralized markets. We directly compare the relative claims about democracy and markets. In doing so, we highlight the ways in which competition operates in the different environments and the relative importance of reflexivity for the two institutional alternatives. We then take up a second potential objection: that we have failed to consider other, more centralized institutional arrangements that might embody reflexivity. We consider three such alternatives: courts and judicial decision making, bureaucracy, and a hybrid form that combines informal norms within formal institutional arrangements. Finally, drawing on the analysis of the effects of social norms on formal decision making, we consider whether the positive effects of social norms might, in fact, be most likely to emerge in an environment of democratic decision making. In part 3, we change focus. We argue, to this point, that democracy is due second-­order priority because it embodies a reflexivity that renders it uniquely adept at the experimental task of determining which institutional arrangements to rely on across different domains. Our argument presumes that democratic arrangements operate effectively only under specifiable initial conditions and that democratic institutions are able to reflexively monitor whether their own preconditions obtain. In our final three chapters, we focus directly on those conditions. In chapters 7 and 8, we offer an analysis of the necessary conditions for democracy to produce the beneficial effects we attribute to it. This involves an extended discussion of the implications of particular conceptions of freedom and equality for encouraging and sustaining political participation. More specifically, we address what effective democratic performance requires in terms of free and equal participation and consider what this actually entails in terms of institutional guarantees and public policy. We conclude with a discussion of the inevitably political nature of these considerations. In chapter 9, we consider three questions that emerge from our extended argument for the priority of democracy. First, to what extent does our argument

23

C h ap te r 1

for the normative priority of democracy provide support for a more general pragmatist theory of legitimate authority and political obligation? Second, to what extent might our argument offer greater appeal to reluctant participants who might otherwise adopt violent or coercive strategies? Third, to what extent is our conception of pragmatist democracy practical, and to what extent is it utopian? Our discussion in this concluding chapter is meant to situate our argument for democracy on the broader terrain that pragmatist political theorists must navigate. In other words, our aim is to conclude not on a defensive note but with a look at some of the broader consequences that follow from our pragmatist argument for the priority of democracy.

24

Chapter 2 Pragmatism and the Problem of Institutional Design

I.  Pragmatism: Philosophical Commitments and Political Consequences As the title of the book announces, the account of democracy we advance is pragmatist. Having said that, however, is not to have said nearly enough. In this chapter, we explain what we have in mind when we embrace the label “pragmatist.” The commitments we sketch here emerge more fully in the argument we present in subsequent chapters. Pragmatism “narrowly conceived” is a philosophical view that consists of “a 1 set of arguments about knowledge, meaning and truth.” That said, pragmatists are an unruly lot. They disagree among themselves about various philo2 sophical and political issues. We are uninterested here in adjudicating these disputes. Our concern is less with policing pragmatism and its boundaries than with advancing a conception of democracy that derives considerable motivation from its pragmatist bearings but that will prove compelling whether or not one embraces pragmatism. We consequently distinguish between the general, relatively ambiguous concept of pragmatism and the particular conception of pragmatism that we endorse. In so doing, we embrace what Dewey calls “the pragmatic rule” which holds that “in order to discover the meaning of [an] idea ask for its consequences.”3 As we formulate and defend our Westbrook (1993, 1). “American Pragmatism has always been less a coherent philosophical school or movement than a philosophical family—­often a contentious family—­of thinkers holding distinct if related positions on the ‘workmanlike’ nature of knowledge, meaning and truth. Pragmatists, that is, have been as inclined to squabble among themselves as to do battle with their nonpragmatic adversaries” (Westbrook 2005, 1). In addition to Westbrook, see the surveys by, for instance, Bernstein (1992; 2010) and Kloppenberg (1998). This unruliness has prompted one pragmatist philosopher to assert that “the wholesale appeals to, or defenses and criticisms of, ‘pragmatism’ that are in currency today are misplaced. There is no pragmatism per se or in general. An appeal to the term should be accompanied by a suitably nuanced qualification indicating broadly the kind of commitment one takes pragmatism to be” (Talisse 2007, 21). That is the task we take up here. 3   Dewey (1920/1948, 163). 1   2  

C h ap te r 2

conception of pragmatism, we are especially concerned with highlighting its consequences for thinking about political institutions in general and democratic political arrangements in particular. Our characterization of pragmatism explicitly is reconstructive. We do not claim that this or that classical or contemporary pragmatist actually adopts a position of the sort we outline. We do, however, draw upon the resources that those classical and contemporary pragmatists offer as we formulate our own defensible pragmatist position. Our enterprise is distinctive in large part because other pragmatists either have not made the sorts of argument we advance, have derived different implications from positions we might share, or would resist the more general position that we reconstruct from particular arguments. We are, in any case, interested here less in engaging in a full-­scale debate with these writers than in relying upon them, as foil or as warrant, for the conception that we reconstruct. At the most abstract level, one might depict pragmatism as one variety of instrumentalism in that it conceives of beliefs, conceptions, theories, and so 4 forth as tools humans use to navigate the natural and social world. This portrait, while perhaps accurate, is too general to be useful. In order to discern the consequences of pragmatism, we need a more fine-­grained conception. Here we understand three commitments as central to pragmatism. 5 First, pragmatists are committed to fallibilism. This commitment has two implications. It means that pragmatists recognize that even those beliefs about which we are most convinced may turn out to be false. It also means that whatever certainty we may attain requires that we actively organize experience in such a way that our beliefs are challenged.6 This requirement is not nearly so onerous as it might at first seem. In a sense, we have no other option. For pragmatists insist that the social and natural world is marked by con7 siderable indeterminacy. As a result, individuals and groups necessarily will confront unforeseen contingencies that will challenge their ostensibly settled beliefs and judgments.8 Fallibilism simply requires that when faced with such See, e.g., Dewey (1920/1948, 145, 156) and, more generally, Menand (2001, xi–­xii). Bernstein (1992, 814); Putnam (1995a, 20–­21, 71); Festenstein (1997, 3–­8; 2003, 551); and Posner (2003, 6–­9, 101). 6   “If truth were to be the belief which would best fit with experience and argument, then if we want true beliefs, we should expose them to experience and argument which might overturn them” (Misak 2000, 57). 7   Hence, for example, Dewey (1929/1958, 127) locates “man in the midst of an uncertain, incomplete, precarious universe.” 8   On the concept of “unforeseen contingencies,” see Kreps (1990a) and Dekel, Lipman, and Rustichini (1998). Such contingencies are events that are literally unexpected—­they are not 4   5  

26

P rag ma tism an d In stitu tion a l D e sig n

problematic circumstances, we prove ready and willing to revise our beliefs should they prove mistaken or erroneous. 9 Second, pragmatists are committed to anti-­skepticism. While any of our beliefs may prove false, it is not possible for us to suppose that they all are false. As Peirce insists, “complete doubt” is impossible, a mere philosophical pretension. Peirce did, however, countenance the sort of “real and living doubt” that is generated by problems we encounter in experience and that prompts us to try to settle our beliefs. This is “real doubt” precisely insofar as we have reason to ques10 tion our beliefs. For pragmatists, then, both doubt and belief must be justified. Finally, pragmatists are committed to consequentialism. This comes through clearly in the “pragmatic rule” just mentioned. There, Dewey distills the advice of Peirce, who urges pragmatists to “consider what effects, which might conceivably have practical bearings, we conceive our conception to have. Then, our conception of these effects is the whole of our conception of 11 the object.” The classical pragmatists arguably considered this advice to be general and took its relevance to the philosophical problems surrounding the 12 concept of “truth” only as one specific instance of its broader applicability. Pragmatists also commonly endorse holism in the sense that they see beliefs and conceptions as forming interconnected, more or less coherent webs. This view applies to pragmatism itself. Fallibilism, anti-­skepticism, and consequentialism are themselves interrelated. Start with fallibilism. It is clear that if any of even our most confidently held beliefs might be false in the pragmatist sense of being useless or self-­defeating, they also must be revisable. This raises the issue of what bounds, if any, constrain our ability to question and revise our beliefs. One temptation here is thoroughgoing skepticism or relativism.13 Pragmatists resist this temptation. Instead they embrace an anti-­skeptical position that limits both the nature and scope of doubt. It is plausible to see this combination of theo14 retical commitments as among pragmatists’ unique intellectual contributions. events that we have considered but to which we attach a very, very low probability, but rather events whose occurrence we have literally not considered at all. 9   Bernstein (1992, 814); Putnam (1994, 152; 1995a, 20–­21, 68); Festenstein (1997, 3–­8; 2003, 551); and Posner (2003, 5–­9). 10   Peirce (1958, 40, 98–­101). 11   Peirce (1958, 124, 192). See also Posner (2003, 6, 63–­65). Talisse (2007, 4–­12) traces the vicissitudes of this maxim as Peirce’s formulation was taken up by William James, then by Dewey. 12   Putnam (1995b, 291–­93). 13   If relativism consists in the view that any belief or practice is as good as any other, pragmatists are in no way relativist. The commitments we sketch here afford us with pragmatist criteria by which to assess beliefs and desires. 14   “That one can be both fallibilistic and antisceptical is perhaps the basic insight of American Pragmatism” (Putnam 1994, 152; 1995a, 21).

27

C h ap te r 2

Fallibilism and anti-­skepticism each sustain the pragmatist commitment to consequentialism. Pragmatists see the social and natural world as fraught with contingency. As a result, even our most fundamental beliefs inevitably will be called into question and potentially proven false. What is important is how individuals and communities respond to the resulting tensions and strains, to the real doubt that unforeseen contingencies generate. Peirce famously identified several ways of settling beliefs—­the methods of tenacity and of authority, the a priori method, and the method of science. And while he himself insisted that the majority of mankind would forever resolve doubt via the method of authority, he insisted that the experimentalism central to science afforded the 15 surest way to fix or settle belief. For experimentation represents the most reliable way of generating solutions to problematic situations. And we assess the results of properly conducted experiments in terms of their actual or anticipated consequences. We take this commitment to be general in the sense that it applies, for example, not just to theoretical concepts, but to public policies and to social, economic, and political institutions as well.16 This claim may seem surprising, given the propensity of some prominent contemporary pragmatists to insist regularly and loudly that pragmatist philosophical commitments have no identifiable political consequences. Indeed, something of an odd, imperfect, and ultimately uncomfortable convergence has 15   Peirce (1958, 101–­11). Note that Peirce is here talking about beliefs that have been become unsettled in some way. He therefore is not contradicting his anti-­skepticism, which holds that we always assume an expansive network of background beliefs. In other words, Peirce does not fall foul, for instance, of the view that Hardin (2009) articulates concerning the “economics of everyday knowledge.” 16   In this regard, we follow Dewey (1927, 169): “Men have got used to an experimental method in physical and technical matters. They are still afraid of it in human concerns.” Dewey issues this complaint explicitly with regard to the creation and assessment of political institutions. He is equally frank in the context of his views on “inquiry”: “Every measure of policy put into operation is, logically, and should be actually of the nature of an experiment” (Dewey 1938/1980, 502). Subsequent pragmatists also have embraced this emphasis on experimentalism and extended it to the social and political domain. See, for instance, the passages from Dewey quoted earlier in the text and, more recently, Rorty who depicts a large part of the pragmatists’ task as consisting “in getting our fellow citizens to rely less on tradition, and to be more willing to experiment with new customs and institutions” (Rorty 1999, 168). See also Posner (2003, 9); Unger (2007). That said, Rorty also insists that it is a mistake to place much stock in the notion of a “scientific method” (e.g., Rorty 1999, 35, 95). He concludes that since no such method exists, pragmatists therefore cannot differentiate in any meaningful way between ways of settling belief. This is a major issue well beyond the scope of our argument. Suffice it to say that we think such claims are highly exaggerated if only because “the fact that we cannot reduce scientific inquiry . . . to an algorithm . . . does not mean that we don’t know anything about how to conduct inquiry” (Putnam 1995a, 69). See also Putnam (1994, 463–­91).

28

P rag ma tism an d In stitu tion a l D e sig n

emerged on this point.17 Richard Posner, for instance, holds that “philosophical pragmatism . . . lacks any political valence and thus is equally compatible with reactionary and revolutionary social visions.”18 Similarly, Richard Rorty intones: “Philosophy and politics are not that tightly linked. There will always be room for a lot of philosophical disagreement between people who share the same politics, and for diametrically opposed political views among philosophers of the same school. In particular, there is no reason why a fascist could not be a pragmatist in the sense of agreeing with pretty much everything Dewey said about the nature of truth, knowledge, rationality and morality.”19 In what follows, we focus on the role that interpretations of the philosophy and politics of Martin Heidegger play in sustaining what we refer to as the Posner-­Rorty consensus. The conception of pragmatism we advance allows us to demonstrate in a straightforwardly pragmatist way that the Posner-­Rorty consensus is mistaken. For while those who subscribe to the consensus hold that pragmatism has no important political consequences, they base that conclusion largely on a considerably more expansive conception of pragmatism than the one we have sketched. On our view, the consequences of embracing so expansive a conception are dire, disabling any effort to discriminate between pragmatists and others. Although we remain uninterested in the task of policing the boundaries of pragmatism, it remains the case that it must have some boundaries. If we cast our conception of pragmatism at too abstract a level, we risk draining the word of meaning in any pragmatic sense. If nearly anyone can be reinterpreted as a pragmatist, the concept verges on being useless for purposes of navigating the intellectual terrain.20 The ­Posner-Rorty consensus is merely a specific manifestation of this larger shortcoming. It is precisely because the expansive conception of pragmatism on which the consensus is grounded sanctions including Heidegger within the rubric of pragmatism that those who endorse it are able to insist that pragmatism has no political consequences. If Heidegger is both a pragmatist and a Nazi, then our claim that pragmatism has democratic political consequences is jeopardized. Must we concede that Martin Heidegger is a pragmatist? 17   On this issue more generally, see Knight and Johnson (1996). Ultimately, questions regarding the political consequences of pragmatism actually are practical and political as much as a theoretical matter (Isaac 1999). 18   Posner (2003, 55). 19   Rorty (1999, 23). 20   This is the thrust of Westbrook’s (2005) critical assessment of various contemporary thinkers who lay claim to the pragmatist label. Conversely, there are thinkers who, for various reasons, and under various influences, “express pragmatist-­sounding ideas without feeling any sense of commonality with self-­identified pragmatists” (Hacking 2007, 33).

29

C h ap te r 2

Here it is important to be careful. Posner refers to Heidegger as “an authentic pragmatist” and insists that he “was both a pragmatist and, for a time anyway, a Nazi.”21 And he does so largely on Rorty’s authority. Rorty acknowledges, in passing, an intellectual debt to the “early pragmatist Heidegger.”22 However, except in this shorthand way, he does not actually characterize Heidegger as a pragmatist in any unambiguous sense. Posner, it seems, takes Rorty to be making a stronger claim than he actually makes in order to support the consensus to which they both subscribe. The underlying difficulties here are conceptual. They begin if, following Rorty, we characterize pragmatism simply as another species of “anti-­ representationalism.”23 The justification for such a view runs roughly as follows: because on a pragmatist view, concepts, theories, and so on are tools that we use to navigate and manipulate the natural and social environment, it does not make sense to think of them as representing the world. While a hammer or a screwdriver, for instance, will be useful insofar as it fits the world (and not otherwise), it is in no way obvious that either represents the world. The problem is that Heidegger subscribes to a version of this view. It is, in other words, plausible to paint both Heidegger and pragmatists as anti-­representationalist in the sense we have just sketched.24 But the fact that Rorty can find ways in which Posner (2003, 9, 45). Rorty (1999, 190–­91). One way to subvert the claim that pragmatist philosophical commitments are consistent with fascist or Nazi politics seems obvious. One could simply point out that it was the “later Heidegger” who was the Nazi, and that he had by then relinquished the “qualified sympathy for pragmatism” discernible in his youthful writings and embraced instead a view of “[l]anguage­as if it were a quasi-­agent, a brooding presence, something that stands over and against human beings” (Rorty 1991a, 3, 32). Whether this gambit might work would require considerably more familiarity with Heidegger’s texts and politics than we can claim. But it would subvert the tendentious claim that one can simultaneously and consistently be both pragmatist and fascist precisely by depriving those who make of it their poster child. It would compel Posner to revise the statement we have quoted in the text. If our speculations are correct, Heidegger was “for a time anyway” a pragmatist and he was “for a time anyway” a Nazi. The problem for Posner is that he was not both simultaneously. 23   Rorty (1991b, 1, 65). 24   Rorty is not alone in lumping Heidegger among the pragmatists in this way. For example, in distinguishing his own expressivist approach to concept use, which he characterizes as a variety of pragmatism, Brandom (2000, 11) invokes as a point of contrast the work of “prominent theorists who are pragmatists in the sense of subscribing to use theorists [sic] of meaning such as Dewey, Heidegger, Wittgenstein, Dummett, and Quine.” For Brandom, what is distinctive about our capacity to use concepts is that we embed them in practices of giving and accepting reasons that allow us to render explicit the role they can play as premises or conclusion of our inferences. We do not wish to discuss Brandom or his work here. We simply want to suggest that Heidegger is a pragmatist only from the vantage point of this fairly coarse-­grained categorization in which any 21   22  

30

P rag ma tism an d In stitu tion a l D e sig n

Heidegger and pragmatists overlap or converge hardly means that we should simply ignore the ways in which they diverge or remain resolutely distinct.25 Consider Rorty’s essays on Heidegger. Even if we accept for purposes of argument the way Rorty depicts pragmatism (which, in many ways, Posner surely does not), it is clear that Rorty goes to at least as great a length differentiating Heidegger’s views from pragmatism as he does assimilating the two. Thus we repeatedly encounter phrases like “whereas Heidegger thinks that . . . the pragmatist thinks that . . .” and “the crucial difference between the Heideggerian and pragmatist view . . .” and so forth.26 Moreover, Rorty acknowledges that in assimilating Heidegger to his own brand of pragmatism, he is putting various “Heideggerian notions . . . to un-­Heideggerian purposes.”27 This admission will hardly surprise the many critics who complain when Rorty seeks to recruit others, like Dewey or Wittgenstein, to purposes that are un-­ Deweyan or un-­Wittgensteinian. To a considerable extent, what Rorty offers in these essays is what he himself calls “a possible Heidegger” distilled from the changes and permutations that mark the actual philosopher’s writings and preoccupations over time.28 We are not prepared to ascertain the extent to which this possible philosopher is related to any actual Martin Heidegger. The simple point is that, on examination, the grounds for asserting that Heidegger is a pragmatist appear remarkably shaky. A larger point clearly follows: if Heidegger was not a pragmatist in any relevant sense, then consideration of his political predilections—­or of whether and how they relate to his philosophical views—­is simply not germane to consideration of the political consequences of pragmatism. Simply put, Rorty never establishes that Heidegger is a pragmatist in a sense relevant to our claim that pragmatist philosophical commitments (anti-­skepticism, fallibilism, consequentialism) have distinctive implications for our thinking about politics and, specifically, about political-­economic institutions. So insofar as the Posner-­Rorty consensus requires its defenders to invoke Heidegger’s theorist who understands language as a tool we use to navigate or perhaps “constitute” the world instead of as a means for representing it is a pragmatist. Such a categorization may be useful for Brandom’s purposes, but it is not obviously useful for thinking about the political consequences of pragmatism. Brandom’s categories were designed for other purposes. There is no reason to think they will prove useful for the one at hand. The same goes for Rorty. 25   Bernstein (2010, 15–­22) offers a succinct, useful overview of the common preoccupations and significant differences between pragmatists and other anti-­Cartesian thinkers such as Heidegger and Wittgenstein. 26   Rorty 1991a, 17, 18. 27   Rorty 1991a, 19. 28   Rorty 1991a, 60.

31

C h ap te r 2

Nazi politics to show that pragmatism has no political consequences, we find it wholly unpersuasive. To the best of our knowledge, Rorty never seeks to establish whether Heidegger is committed to any of the three components of our narrower conception of pragmatism. This is unsurprising since those components hardly feature in Rorty’s own thinking or in the portrait he paints of the pragmatist tradition.29 Yet, as a historical matter, it is their commitment to fallibilism, anti-­skepticism, and consequentialism that distinguishes the classical pragmatists and their more orthodox progeny from other anti-­representationalists with whom they share some family resemblance. We think it a considerable virtue of our conception of pragmatism that it more accurately captures these historical features of pragmatist thought than do competing, more expansive conceptions. Our concern, however, is not solely with historical fidelity. We find our conception of pragmatism useful especially insofar as it excludes thinkers such as Heidegger and so deflates the claims of those who subscribe to the Posner-­Rorty consensus that pragmatism lacks political consequences. If we find the Posner-­Rorty consensus unpersuasive, we have yet to establish the affirmative claim that pragmatism indeed has political consequences. Pragmatists, just insofar as they are committed to experimentalism, would generally and properly concede that, in any given case, they are not commit30 ted in advance to a specific moral stance, political action, or social policy. We insist, however, that this very same commitment to experimentalism has quite determinate implications for the design and justification of institutions of democratic decision making. This last assertion requires substantial defense. One way to approach this task is to notice that although contemporary pragmatists disagree vigorously on the matter, Dewey was quite explicit that his philosophical commitments were relevant only insofar as they have political implications. In a relatively early, properly philosophical essay, he reflexively applies “the pragmatic rule” to the pragmatist conception of truth. He concludes: If the pragmatic ideal of truth has itself any pragmatic worth, it is because it stands for carrying the experimental notion of truth that reigns among This is not true, however, of Posner (2003), who suggests that pragmatists precisely are committed to anti-­skepticism, fallibilism, and some brand of consequentialism. It is difficult to see how Posner squares these commitments with the claim that Heidegger is a pragmatist. The perplexity is one consequence of his embrace of “everyday pragmatism” at the expense of “philosophical” concerns. 30   Dewey, for example, insists that the sort of social inquiry that pragmatists endorse is open-­ ended in the sense that “no particular end is set up in advance so as to shut in the activities of observation, forming of ideas, and application” (Dewey 1920/1948, 146). 29  

32

P rag ma tism an d In stitu tion a l D e sig n

the sciences, technically viewed, over into political and moral practices, humanly viewed.31 He subsequently (and repeatedly) remarks that the pragmatic conception of truth . . . places upon men the responsibility for surrendering political and moral dogmas, and subjecting to the test of consequences their most cherished prejudices. Such a charge involves a great change in the seat of authority and the methods of decision in society.32 When we pursue the implications of such observations, it becomes clear that pragmatism implies some kind of priority for democracy. Before proceeding, however, we must address another group of pragmatists, this one consisting of those who might wonder just what we are going on about. These more or less orthodox pragmatists will not find it terribly contentious to accord priority to democracy since they insist that their philosophical views have radically 33 democratic implications. Regardless of whether one accepts that label, we agree that pragmatism has democratic implications even if we specify them in ways that such theorists might not embrace.

II.  Democratic Commitments Presuppose Institutional Embodiment Posner and Rorty attempt in implausible ways to sequester politics from philosophy and vice versa. One reason their approach has attracted a degree of notoriety is that it departs from the “radical democratic” views that many other pragmatists presume follow from their own philosophical stance. Richard Bernstein is a prominent proponent of this latter position. He interprets pragmatist politics in light of a distinction Dewey notoriously draws between “democracy as a social idea and political democracy as a system of government.”34 Whereas the latter constitute the “mechanism,” “forms,” “devices,” or “machinery” that embody the democratic idea, Bernstein and other contemporary pragmatists plausibly assert that Dewey accords priority to how the “idea” of democracy animates such institutional arrangements. Hence Bernstein asserts that for Dewey, democracy “was not primarily a set of institutions, formal Dewey (1911, 110). Dewey (1920/1948, 159–­60). 33   See, e.g., Bernstein (1992, 815; 2010) who insists that pragmatists are committed to “radical democratic” politics. 34   Dewey (1927, 143, 83). 31   32  

33

C h ap te r 2

procedures or even legal guarantees.”35 This inference seems correct, but it is important to be clear about just what we are inferring. For Dewey in no way dismisses the importance of political institutions. Indeed, the opposite is the case. While he stresses the importance of democracy as a “social idea.” Dewey hardly discounts the importance of “governmental institutions” or “political machinery.”36 He insists only that when undertaking the essential task of critically assessing democracy, we resist the temptation to restrict it narrowly to any domain or identify it with any particular institutional manifestation. This is a crucially important but ultimately quite limited claim. And there is a range of good reasons why contemporary pragmatists, especially those who believe their philosophical views entail radical democracy, ought to avoid inferring anything more from the underlying distinction. Put otherwise, while we may well need to differentiate democracy as a social or political “ideal” and democracy as a “system of government,” we should be careful not to inflate that analytical distinction into a dichotomy, especially one that hinders our ability to examine how ideals and institutions are entangled in politics. Dewey rightly insisted that democracy, like any regulative ideal, presupposes some institutional embodiment, some means of existing in the world.37 He recognized, too, that this theoretical observation has important political implications. Indeed, he was adamant more generally about the dangers that beset those who treat ideas in a disembodied manner. It thus turns out that the old, old dread and dislike of matter as something opposed to mind and threatening it, to be kept within the narrowest bounds of recognition; something to be denied so far as possible lest it encroach upon ideal purposes and finally exclude them from the real world, is as absurd practically as it was impotent intellectually. Judged from the only scientific standpoint, what it does and how it functions, matter means conditions. To respect matter means to respect the conditions of Bernstein (1986, 261). Dewey (1927, 143–­44). 37   When we refer to democracy as a regulative ideal, we mean that it is something that might guide critical assessment of actual institutions, not as something that we might in fact achieve in practice. “Regarded as an idea, democracy . . . is an ideal in the only intelligible sense of an ideal: namely, the tendency and movement of some thing which exists carried to its final limit, viewed as completed, perfected. Since things do not attain such fulfillment but are in actuality distracted and interfered with, democracy in this sense is not a fact and never will be” (Dewey 1927, 148). In this sense, Dewey distances himself from a sort of neo-­Hegelianism that might see history as the unfolding of democracy. Indeed, he explicitly dismisses the notion that the idea of democracy caused the emergence of its political manifestations (Dewey 1927, 144–­46). 35   36  

34

P rag ma tism an d In stitu tion a l D e sig n

achievement; conditions which hinder and obstruct and which have to be changed, conditions which help and further and which can be used to modify obstructions and attain ends. Only as men have learned to pay sincere and persistent regard to matter, to the conditions upon which depends negatively and positively the success of all endeavor, have they shown sincere and fruitful respect for ends and purposes. To profess to have an aim and then neglect the means of its execution is self-­delusion of the most dangerous sort. Education and morals will begin to find themselves on the same road of advance that say chemical industry and medicine have found for themselves when they too learn fully the lesson of wholehearted and unremitting attention to means and conditions—­ that is, to what mankind so long despised as material and mechanical. When we take means for ends we indeed fall into moral materialism. But when we take ends without regard to means we degenerate into sentimentalism. In the name of the ideal we fall back upon mere luck and chance and magic or exhortation and preaching; or else upon a fanaticism that will force the realization of preconceived ends at any cost.38 The warning could not be more explicit. But Dewey’s case for treating institutional arrangements seriously is not solely a negative one. Concern for the institutional “conditions” necessary for democracy is a central theme throughout Dewey’s properly political writings. In The Public and Its Problems, for instance, he was preoccupied with “conditions,” namely those that are “indispensable” if the public is to surmount its inchoate, uncoordinated, and therefore ineffectual state and assert control over the forces that bring it into being.39 The problem, more specifically, is that given the complexity and size of modern societies, our inherited political institutions are inadequate to the task of coordinating a democratic public.40 Likewise, in “Creative Democracy—­The Task Before Us,” the essay that inspires contemporary pragmatists like Bernstein to profess “faith” in the project of “radical democracy,” Dewey famously wrote: Democracy is a way of personal life controlled not merely by faith in human nature in general but by faith in the capacity of human beings for intelligent judgment and action if proper conditions are furnished. . . . For what is the faith of democracy in the role of consultation, of conference, Dewey (1920/1948, 72–­73). Compare Dewey (1927, 149, 173) for similar but less fully elaborated observations. 39   Dewey (1927, 157, 166, 208). 40   Dewey (1927, 113–­14). 38  

35

C h ap te r 2

of persuasion, of discussion, in formation of public opinion, which in the long run is self-­corrective, except faith in the capacity of the intelligence of the common man to respond with commonsense to the free play of facts and ideas which are secured by effective guarantees of free inquiry, free assembly and free communication?41 Regardless of whether one endorses the characterization of democracy as a way of life, here too, “conditions”—­the mechanisms, machinery, devices, or forms provided by political democracy—­play an obvious, crucial role. On Dewey’s account, the capacity for intelligent action is not sufficient, in and of itself, to support faith in democracy. That faith presupposes very much the same sustaining scaffolding of institutional mechanisms required by a public if it is to recognize its common, defining interests.42 A plausible argument to the effect that pragmatism has “radical democratic” consequences, in other words, requires that we offer a critical analysis of political-­economic institutions.

III.  Institutional Implications of Pragmatism Pragmatists typically are suspicious of accounts of politics that start from at43 omistic individuals in some original presocial state. We share this suspicion, and none of what we argue presumes such fictions. Yet we also recognize that even in a world constituted by the sorts of inevitable diversity, disagreement, and strategic interdependence that we have termed the circumstances of politics, individuals and groups confront a basic problem of how to coordinate their ongoing interactions. Large populations who occupy the circumstances of politics thus confront the need for a set of social institutions. The problem, of course, is that individuals and groups will disagree about how best to institutionalize their ongoing affairs. From a pragmatist perspective, such disagreement is understandable given the plurality of institutional forms available for social coordination. Markets, common-­law adjudication, Dewey (1939/1993, 242). It arguably is just his preoccupation with “conditions” that allows Dewey to distance himself not just from the emergent “realist” theorists of democracy like Walter Lippmann (1927) but also from his curmudgeonly fellow pragmatist, Charles Peirce. The latter quipped that “the mass of mankind” are by disposition forever consigned to rely on the “method of authority” rather than availing themselves of the “method of science” as a way of “fixing belief.” Peirce was characteristically dismissive of the masses: “If it is their highest impulse to be intellectual slaves, then slaves they ought to remain” (Peirce 1958, 104–­5). 43   For example, Dewey (1927) explicitly disavows any view of social or political order presented “in terms of causal origins.” 41   42  

36

P rag ma tism an d In stitu tion a l D e sig n

social convention, and so forth, might well be appropriate for different classes of interaction under different circumstances. Pragmatists themselves are agnostic about the substantive choice of institution that might govern any sort of interaction. Given their fallibilist commitments, they advocate experimentalism on this issue.44 Such an experimental posture exploits the virtues of diversity and promotes decentralized control to the extent that various constituencies or jurisdictions might well adopt different institutional mechanisms to cope with the same or very similar coordination problems. However, pragmatists also recognize that both diversity and decentralization are susceptible to the baleful effects of asymmetrical distributions of power and so insist that experimentalism be constrained by conditions of freedom and equality of the sort we discuss at length in chapters 7 and 8.45 The importance of these conditions becomes apparent when we consider the linkages between pragmatist commitments to experimentalism and the conception of community that they endorse. For pragmatists like Dewey, democracy and community are intimately linked. This creates a pair of dangers. On the one hand, it invites what we take to be dubious efforts of contemporary communitarians to recruit Dewey specifically and pragmatism more generally to their cause.46 On the other hand, it underwrites the sort of skepticism that Posner voices regarding what he considers the naive or utopian character of Dewey’s conception of democracy.47 We think it is important to avoid both temptations. More constructively, the view we sketch here allows us to state what we consider among the defining features of our enterprise. While the pragmatist account we offer is clearly normative, the value we ascribe to democracy is as much epistemic as moral or ethical.48 In this sense, our argument converges In support of this view, consider Dewey’s characterization of the process by which we should determine where to draw the line on appropriate state activity: “The trial process may go on with diverse degrees of blindness and accident, and at the cost of unregulated procedures of cut and try, of fumbling and groping, without insight into what men are after or clear knowledge of a good state even when it is achieved. Or it may proceed more intelligently, because guided by knowledge of the conditions that must be fulfilled. But it is still experimental. And since conditions of action and of inquiry and knowledge are always changing, the experiment must always be retried; the State must always be rediscovered” (Dewey 1927, 33–­34). 45   On the dangers that attend efforts to implement schemes of decentralization that neglect extant asymmetries of power at the local level, see Bardhan (2005). 46   See, e.g., Sandel (1996). 47   Posner (2003). 48   We invoke the term “epistemic” in the same sense as Putnam (2002, 31–­33) does. We do so advisedly, as we do not subscribe to what has commonly become known as the epistemic view of democracy endorsed by Cohen (1986a) and Estlund (1993). Whereas such views presume that 44  

37

C h ap te r 2

with a provocative, if still somewhat inchoate line of argument among contemporary pragmatists (and sympathetic interlocutors) who adopt just that stance.49 These pragmatists do not agree about everything, and we make no effort here to sort out the overlapping allegiances and disagreements among them. But they do converge, in ways we find congenial, in seeing democratic institutions and practices as performing a broadly cognitive task in our ongoing task of coordinating social, economic, and political interactions. In the following much quoted passage, Dewey asserts what he takes to be an intimate link between democracy and community. Regarded as an idea, democracy is not an alternative to other principles of associated life. It is the idea of community life itself. It is an ideal in the only intelligible sense of an ideal: namely the tendency and movement of some thing which exists carried to its final limit, viewed as complete, perfected. . . . [D]emocracy in this sense is not a fact and never will be. But neither is there or has there ever been anything which is a community in its full measure.50 Community consists in decentralized, informal arrangements (including rules, roles, values, and identities) for coordinating ongoing interactions. Dewey makes it clear here that he has in mind not just this or that extant community but a regulative ideal, a normative vision of what any existing arrangement might become. He goes on to explain: Associated activity needs no explanation; things are made that way. But no amount of aggregated collective action itself constitutes a community. For beings who observe and think, and whose ideas are absorbed by impulses and become sentiments and interests, ‘we’ is as inevitable as ‘I.’ But ‘we’ and ‘our’ exist only when the consequences of combined action are perceived and become an object of desire and effort, just as ‘I’ and ‘mine’ appear on the scene only when a distinctive share in mutual action is consciously asserted or claimed. . . . Interactions, transactions, occur de facto and the results of interdependence follow. But participation in democratic institutional mechanisms are in some sense truth tracking, where truth is posited as existing independently of the democratic processes, we do not presume that in politics there is a truth of the matter waiting to be discovered. This hopefully becomes clear in subsequent chapters. This does not mean that we discount the notion of truth per se, only that we doubt it offers much purchase in thinking about politics. 49   See, e.g., Putnam (1991); Honneth (1998); Misak (2000); Westbrook (1998; 2005); Kitcher (2006); Anderson (2006); and Talisse (2011). 50   Dewey (1927, 148–­49).

38

P rag ma tism an d In stitu tion a l D e sig n

activities and sharing results are additive concerns. They demand communication as a prerequisite.51 Community, then, consists not just in coordinated interaction but requires the reflexive recognition of participation and its consequences. Dewey is clear that the communication that sustains communication is not—­as in, say, Arendt, Habermas, or Oakeshott—­what we might term purposeless. Instead, communication of the sort he deems integral to community provides the medium for instrumental action: “Recollection and foresight are possible; the new medium facilitates calculation, planning, and a new kind of action which intervenes in what happens to direct its course in the interest of what is foreseen and desired.”52 Informal communities, on this view, like more formal institutions and practices, are instrumental to our coordinated efforts to solve problems that emerge as we navigate the natural and social world. As a result, it is no surprise that that the basic commitments to fallibilism, anti-­skepticism and consequentialism carry over into pragmatist conceptions of community.53 We can appreciate some of the implications of this conception of community by shifting our focus from Dewey to Peirce. The latter was perhaps primarily responsible for the pragmatist preoccupation with community, insofar as he advanced a conception of belief, knowledge, and truth as grounded in the activities not of individuals but of communities of inquiry. This raises two crucial themes: (1) conflict and disagreement within such communities and (2) the institutional structures necessary for them to operate. While Dewey has been criticized for neglecting the pervasive conflict that rends communities, the same cannot be said of Peirce. The dire consequence of pursuing truth by one means rather than another, or of holding to or advancing dissident views, is a pervasive theme in his essay “The Fixation of Belief.” Those consequences are not, on his view, pretty. Of the method of tenacity, he says: “When complete agreement could not otherwise be reached, a general massacre of all who have not thought in a certain way has proved a very effective means of settling opinion in a country.” And of the method of authority: “Certain non-­conformities are permitted; certain others (considered unsafe) are forbidden. These are different in different countries. And in different ages; but, wherever you are, let it be known that you seriously hold a tabooed belief, Dewey (1927, 151–­52). Dewey (1927, 151–­53). Stated otherwise, Dewey understands politics as “the cognitive medium with whose help society attempts, experimentally, to explore, process, and solve its own problems with the coordination of social action” and, more specifically, that he “wishes to understand democracy as a reflexive form of community cooperation” (Honneth 1998, 765, 775). 53   Bernstein (1998). 51   52  

39

C h ap te r 2

and you may be perfectly sure of being treated with a cruelty less brutal but more refined than hunting you like a wolf.” Given that he believed that most men most of the time are subject to this method of settling belief, it is safe to say that Peirce’s assessment is damning.54 It is safe to say, too, that this hardly amounts to a naive view of how knowledge operates in society. Peirce, of course, advocated what he called the “method of science” even if he despaired of the prospects that it would come to govern social relations generally. But the communities of inquiry that do embrace this method of settling belief hardly are diffuse congeries of individuals engaged in uncoordinated endeavors. They inhabit institutions of various sorts that can be assessed in terms of effective performance in generating, disseminating, and verifying knowledge.55 In this respect, communities of inquiry resemble communities of other sorts in that they are subject to the same vagaries of disagreement and substandard performance. So, while it may not be possible or plausible to identify democratic politics with an idealized conception of a community of inquiry, it is a mistake to dismiss the task of ascertaining how the two actually and potentially interact.56 That, of course, is the problem that Dewey explores at length in The Public and Its Problems. That his analysis is in some ways incomplete or naive hardly obviates the task he took up. From our pragmatist perspective, the themes we just have sketched come together in ways that underwrite our argument for democracy. First, since knowledge and information are a function of experience, the diversity of experience generates what might in contemporary terms be called “distributed intelligence.”57 Community establishes boundaries that partition any such distribution and that thereby constrain and channel the possibilities for the subsequent development of information and knowledge. Second, community on a pragmatist account is informed by the basic commitment to fallibilism. If any of the beliefs to which an individual subscribes may be erroneous, this will be clearest just insofar as there are others who are able and willing to point this out. This is a point that goes back at least to Peirce.58 For a pragmatist, therefore, community is characterized by reflexivity just to the extent that it is a locus of disagreement and conflict. Finally, we later argue at length Peirce (1958, 110). On this point, see Kitcher (2006), who, invoking Dewey, explores the exigencies of establishing and maintaining what he calls an “Inquiry-­and-­Information System.” 56   Bernstein (1998, 147). 57   See Posner (2003, 102), who notes that Dewey and Hayek both recognize the distributed nature of social knowledge even as they draw opposite policy prescriptions from that observation. See also Knight (2001). 58   See, e.g., Peirce (1958, 40). 54   55  

40

P rag ma tism an d In stitu tion a l D e sig n

that institutions generally emerge as the by-­product of distributional conflicts marked by systematic asymmetries of resources and hence of power. Since they are simply sets of informal institutional arrangements (e.g., norms, conventions, and so forth), we explain the emergence of communities in precisely the same manner. They are, in other words, the more or less arbitrary product of power relations. And this, on our view, has implications for how communities operate. On our view, then, pragmatists should not presume that communities partition knowledge or structure disagreement in ways that are neutral, efficient, or otherwise collectively beneficial. This characterization deflates the concerns of those who find the pragmatists’ analogy between communities of scientific inquiry and political commu59 nities too rarified. Like communities of other sorts, communities of inquiry are, on our view, more or less thoroughly politicized. And, like institutions more generally, they confront a severe burden of justification in the sense that, absent specifiable initial conditions, a community will not effectively achieve whatever goals or aims its members set for it. This highlights the need for institutional mechanisms that allow community members to monitor and enforce the sorts of conditions that mitigate the effects of power asymmetries. As we make clear below, this imperative holds both for particular institutional forms and for the process of institutional experimentalism more generally. This discussion of community underscores the institutional implications of pragmatism. For pragmatists do not just insist on the need for institutions to coordinate ongoing social and political interactions. That, after all, would be a platitude. Pragmatists also understand that we can assess any institution that performs those first-­order coordinating functions and judge whether it is operating well in both absolute and relative terms. Regardless of whether relevant parties seek to establish whether some institutional mechanism is coordinating interactions in a given domain well on its own terms or compared with how some alternative mechanism might perform the same task, they will, from a pragmatist perspective, recognize two things. First, different institutional mechanisms will be judged by quite different criteria of effectiveness. And second, institutional mechanisms will operate effectively only if particular initial conditions obtain. A pragmatist on our view, then, will further insist upon the need for some institutional mechanism that will enable relevant parties to monitor the existence of those initial conditions, propose remedies when the relevant conditions do not actually obtain, and assess the effectiveness with which particular institutions, in fact, coordinate ongoing interactions across various domains. The task, in other words, is not simply to coordinate ongoing Posner (2003).

59  

41

C h ap te r 2

social, economic, and political interactions, but to do so—­as Dewey would say—­“intelligently.” This motivates the pragmatists’ search for institutional arrangements that facilitate this second-­order task. Here it becomes apparent not only why, on our account, pragmatists accord priority to democracy but the specific sense in which they do so. For pragmatists rightly hold out scant hope that even under idealized conditions, alternative institutional mechanisms (e.g., markets, bureaucracies, legal adjudication, social norms) might effectively accomplish the second order tasks of monitoring and assessment. Pragmatists thus accord democratic institutional mechanisms strong priority just insofar as they are able to perform tasks necessary if those other forms—­and indeed, if democratic arrangements themselves—­are to accomplish effectively the first-­order task of coordinating ongoing social and political interactions. One of our aims is to present a detailed argument to this effect. Before turning to that argument, however, we want to stress that we here are working at the shifting boundary between the analytical and normative tasks of political theory. For in assessing the effectiveness of institutional forms, whether singly or as part of some more encompassing arrangement, we are engaged in a normative task that explicitly raises issues of justification.

IV.  Normative Implications of Pragmatism: Diversity, Consequences, and Conditions We have focused thus far on how the notion of community underwrites pragmatist thinking about social and political institutions. Pragmatist conceptions of community, as we portray them, stand at considerable distance from common 60 views that presume homogeneity and value consensus. Pragmatists embrace community because it affords an institutional framework for experimentalism, which, in turn, presupposes diversity—­both within particular communities 61 and across them—­and the sorts of disagreement that follow from it. The importance pragmatists accord to diversity allows us to highlight the distinctive normative framework to which their basic theoretical commitments give rise. Diversity is, on our view of the circumstances of politics, a simple social fact. From a pragmatist perspective, this fact takes on a normative significance 60   Contemporary pragmatists themselves do not always or consistently avoid such assumptions. However, close examination of writings by early pragmatists suggests that they did not identify community with homogeneity (Bernstein 1998; Campbell 1998). Such interpretations render dubious the attempts by contemporary communitarians to appropriate pragmatism (e.g., Sandel 1996). 61   See Posner (2003, 9, 101–­103).

42

P rag ma tism an d In stitu tion a l D e sig n

that derives from the contribution diversity makes to processes of experimentation. First, insofar as experimentation consists in processes of problem solving by trial and error, diversity of inputs is in some sense definitional. Experimentation simply presupposes an array of proposed solutions to any given problem, and diversity of perspectives will enhance that range. Clearly, the relations between the extent of social diversity, the range of proposed solutions it might generate, and the outcome of social and political experimentation need not be linear. At some point the potential benefits from considering or even proposing yet another solution to some problem will start to diminish.62 A pragmatist will acknowledge that there is no nonpolitical way to identify the bounds on how many solutions—­or even which of those that are proposed—­ deserve consideration. But she also will insist that the threat to collective decision making lies less in the risk of excessive diversity than in overly heavy a reliance on factors like tradition and authority that too rapidly or too narrowly 63 or too unreflectively constrain the range of considered alternatives. Second, diversity contributes to experimentation in ways other than by generating proposals. It also is necessary to the sort of robust criticism and challenge that experimental testing demands. Diverse perspectives will help generate a range of objections that in turn will compel relevant actors to reconsider or further justify the solutions they favor for any given problem. This, of course, is a dynamic process in which those who press objections will be compelled to rethink their own views by the responses their criticisms elicit. In this sense, diversity should increase the confidence with which decision makers either accept or repudiate any of the particular solutions on offer. This brings us to the normative point. In an interdependent, indeterminate world where even our most firmly held beliefs might prove mistaken or misguided in the face of unforeseen contingencies, institutional arrangements that can exploit the virtues of diversity will be more likely to generate better social and political outcomes than arrangements that cannot. This is simply the manner in which the pragmatist’s consequentialist commitments manifest themselves in the assessment of institutions and policies. It obviously leaves open the question of what we might mean by “better.” In contrast to utilitarians, devotees of law and economics, “epistemic” democrats, theorists of social or cultural evolution, and others who each advance one or another nonpolitical and putatively definitive criterion for Rodrik (2007, 165) indeed suggests that it can be difficult to distinguish between productive experimentalism and policies that are simply rationalized in those terms as a subterfuge because delay serves entrenched interests. 63   See Knight and Johnson (1996). 62  

43

C h ap te r 2

differentiating better from worse consequences, pragmatists insist that any such distinction is necessarily political. Pragmatists insist that there exists no independent metric for making such determinations.64 And thus, they rightly could claim that the political nature of decision making is likely to be even more pronounced in the design and implementation of institutions. There is no way to avoid the conclusion that in collective decision making, the criteria of assessment are as much at issue as are solutions to the substantive problems that prompt it in the first place. Pragmatists insist that politics preempts any recourse to foundational criteria. So they recognize that any understanding of what counts as “better” consequences will emerge from the politics of collective choice. We return to this matter later in this chapter. Pragmatists do not value either diversity or consequences per se. Rather, they are committed to what we call tempered consequentialism—­a concern with outcomes tempered by attention to the conditions under which those outcomes are generated.65 Pragmatists assess institutions and policies in terms In particular, it is important not to identify pragmatism with some version of utilitarianism. This would be a mistake. In fact, pragmatists and utilitarians adopt quite different approaches to problems of assessment and justification. Pragmatists are nonfoundationalists and so deny that some nonpolitical vantage point exists from which we might render and assess political judgments. Utilitarians, by contrast, hold out hope that “utility” can provide just such a foundation. This is not the place to explore this difference in intellectual strategy. It is enough to indicate that on their own terms, utilitarians encounter grave difficulties once they acknowledge that there are broad expanses of politics in which interpersonal comparisons of utility are unreliable or impossible. For in such circumstances, endorsing utility maximization simply amounts to endorsing democratic processes and attending to the conditions that might serve to justify the outcomes that such processes produce (Riley 1990). Political theorists have, of course, made efforts to “rehabilitate” (Goodin 1995) or, more ambitiously, to systematically restore, repair and renovate (Braybrooke 2004) utilitarianism as a tool for assessing policies and institutions. These efforts typically involve an attempt to finesse the problem of interpersonal comparisons of utility by invoking as the utilitarian maximand some other putatively foundational criterion such as “needs” (Braybrooke) or “welfare” or “interests” (Goodin). The difficulty, of course, is that those concepts are thoroughly subject to contestation and debate. In short they are political and hence, should we adopt some such metric, we would require some set of institutions to structure our debates and disagreements over just what counts when we talk about say, the needs or interests whose satisfaction we are seeking to maximize. While we will not make the argument here, it is plausible, too, to suggest that Posner embraces pragmatism just insofar as he recognizes that wealth maximization—­the basic criterion of the law and economics movement—­suffers from difficulties analogous to those we posit with respect to utilitarianism. 65   Critics may find this reliance on what we call tempered consequentialism somewhat fishy. In response, we simply point out that we hardly are alone in claiming that when it comes to matters of justification, political theorists must keep an eye on both consequences and procedural factors. For others in this company, see, e.g., Christiano (2004) and Sen (2003, 280–­1). 64  

44

P rag ma tism an d In stitu tion a l D e sig n

of their anticipated consequences, and they claim that those consequences are more likely to be normatively attractive to the extent that they reflect a diversity of perspectives. But they also insist that even consequences that emerge in circumstances of diversity must be further constrained by initial conditions of freedom and equality in both procedural and substantive terms. These conditions are intended to increase the opportunity for diverse voices to be heard in processes of collective decision making and thereby increase the likelihood that the views they express and interests they articulate will influence the resulting outcomes. This tempered consequentialism reflects the basic pragmatist commitment to fallibilism and is at the core of pragmatist experimentalism in politics. Thus, the normative implications of pragmatism are complex. They also are simultaneously modest and insistent. They are modest insofar as pragmatists extend no guarantees. They are insistent to the extent that pragmatists propose freedom and equality as necessary to the normative justification of democratic politics. But since pragmatists seek to dispense with foundationalist commitments, it is, on their interpretation, the commitment to democratic politics that animates their commitment to freedom and equality, rather than the reverse. We defend this claim at length in chapters 7 and 8.

V.  Political Implications: Institutional Experimentalism As we have been emphasizing, given the fact of institutional pluralism, the selection of appropriate institutions for the coordination of interaction across various domains is a complex undertaking. From a pragmatist perspective, political decision making involves not only the task of identifying appropriate institutions but also ascertaining with a reasonable degree of confidence that something approximating the requisite initial conditions presupposed if our chosen institutions are to operate effectively are likely, in fact, to obtain. Dewey sketches, in a general way, what this view of inquiry entails: “observation of the detailed makeup of the situation; analysis into its diverse factors; clarification of what is obscure; discounting of the more insistent and vivid traits; tracing the consequences of the various modes of action that suggest themselves; regarding the decision reached as hypothetical and tentative until the anticipated or supposed consequences which led to its adoption have been squared with actual consequences.”66

Dewey (1920/1948, 164).

66  

45

C h ap te r 2

As Dewey emphasizes, the results of such an inquiry are scientific just to the extent that they are the product of appropriate methods or procedures.67 But unlike many characterizations of the process of experimental inquiry, the ultimate resolution of the problem of institutional choice is not guided by some value-­neutral metric or decision rule. Pragmatists understand that despite our best efforts to establish and maintain the appropriate conditions to make the process trustworthy and scientific, the experimental process remains, on some dimensions, inherently political—­even when successful. Pragmatists view the results of experimental inquiry as hypotheses revisable in light of social experience and political debate. Dewey understands the scope of experimentalism broadly, emphasizing “the place in experiment of comprehensive social ideas as working hypotheses in direction of action.”68 Such inquiry presupposes not only impartial investigation but also free and extensive communication of results. As Dewey remarks, “knowledge is communication as well as understanding.”69 For a pragmatist, the results of social inquiry cannot be applied mechanically to social and political problems. Pragmatists are not committed to social engineering. They insist instead that the results of inquiry must enter into social and political deliberation and debate rather than be merely applied to problems.70 This is a broad and somewhat vague demand. However, it captures at least two important constraints on social inquiry. The first constraint is that the results of social inquiry typically do not unambiguously support any specific policy proposal or course of political action. They require interpretation and discussion and, in that sense, inquiry, and especially what relevant agents might infer from the results of inquiry leaves considerable room for disagreement.71 The second constraint is that improved understanding of practices and institutions can enter reflexively into the ways that social and political agents participate in those practices. Institutional choice itself is a function of the outcomes produced by the social interactions that take place within the context of institutional rules. On our pragmatist view, social relations consist simply in ongoing interactions that need to be coordinated in an effective way; and we appraise and compare institutions in light of our expectations regarding which will most effectively facilitate that task, where the criteria of effectiveness, in turn, are specified in terms of social consequences. For a pragmatist, the criterion of “best” depends Dewey (1927, 163). Dewey (1987, 32). 69   Dewey (1927, 176). 70   Dewey (1927, 123–­25, 174, 207–­9). 71   Dewey (1927, 178, 203). 67   68  

46

P rag ma tism an d In stitu tion a l D e sig n

upon the specific circumstances of the particular interaction. This is where the inherently political nature of social inquiry and experimentation comes into focus. For, unlike utilitarianism or other variants of consequentialism grounded in some foundational criterion of assessment, pragmatism rejects the notion that there is one single metric that forms the basis of institutional choice. Rather, a number of factors might come into play in assessing the consequences of such a choice. Some participants might favor efficiency, while others might give priority to principles of equality or fairness or distributive justice. Some might place the emphasis on long-­term stability, while others might advocate the benefits of fluidity and change. And some might insist on the importance of the effects on freedom of choice or empowerment, while others might place greater emphasis on the collective benefits derived from constraining choice in particular circumstances. And so on. Only rarely, moreover, is it likely that any participant will wholly discount the importance of any of the criteria we have just mentioned. In the face of what they see as reasonable disagreement, pragmatists believe that there is no ex ante way to resolve these differences and tensions. And so they conclude that the criteria of institutional choice must be determined by those who are involved in the relevant social interaction in accordance with their own interests, attachments, and commitments. That is, the participants themselves must determine what constitutes the “best” consequences in any particular situation. In doing so, they must balance the various factors by which they assess the relevant consequences and derive a workable standard for particular cases, hence the political nature of the experimental process. The main implication here is that processes of institutional experimentation themselves require an institutional framework to help resolve disagreements over what counts as best. This is an important aspect of the more general task of institutionalizing processes of institutional experimentalism, a primary concern for pragmatists. Here it is important to note that on a pragmatist view, processes of institutional experimentation are inherently political not only because the criteria of assessment are contestable. Politics enters into any process of social experimentation on a second, and important, dimension insofar as we attend to the costs of inquiry. Although pragmatism gives priority to experimentation and inquiry, it also requires us to take account of other factors, including the consequences of experimentation itself. Any form of social experimentation brings with it costs. Pragmatists understand that they should be sensitive to these costs and consider them in assessing the merits of any new policy proposal. The consequences of experimentation can extend beyond the circumstances of a particular social activity and affect social relationships

47

C h ap te r 2

in other related areas. The costs might be substantial, and they could take many forms. We can see what is at stake here by briefly considering two such costs that introduce potential obstacles to this approach to experimentation. On the account we present here, pragmatism undergirds a specific approach to social decision making. Thus, a commitment to pragmatism entails not only a willingness to participate in the process of inquiry but also a basic commitment to abide by the policy implications of that process. But what will enter into the individual decision to participate and comply? The first issue underscores the importance of stable expectations and, as such, is essential to pragmatism’s own concern about the value of individual autonomy. Central to the pragmatist vision is the idea that individuals should have the autonomy, whenever possible, to realize their self-­chosen life plans. One of the necessary conditions for such self-­realization is stability in our expectations as to how the world around us works, including most importantly how others will behave in their interactions with us. The stability of social expectations is always in tension with experimentation, since experimentation creates social uncertainty and challenges, at least temporarily, our existing beliefs. Put this way, experimentation at the societal level appears to complicate self-­realization at the individual level. This highlights one of the ongoing tensions societies face when they set out to address questions of institutional choice. There is no principled way of permanently resolving the question of experimentation and change versus stability. It is, rather, a factor that must be addressed in the decision-­making process itself. For it forces pragmatists to confront the basic question of how much experimentation, with its attendant uncertainty and potential instability, individual members of society will be willing to accept?72 The second issue relates to the distribution of the costs of experimentation among the individual members of society. Pragmatists understand that institutional experimentalism is justified primarily in terms of its value for social decision making and emphasize both the collective and mutual benefits of such inquiry. But they must also take account of the personal calculations of costs and benefits in which individuals might engage before making a decision about their willingness to participate. And thus the distribution of costs among the members of society becomes a relevant consideration. There are good 72   Contemporary pragmatists approach this tension in divergent ways. Rorty notoriously uses a stark public-­private distinction to partition matters of social and political coordination from matters of individual self-­realization. Unger (2005; 2009) sees the two as inextricably entangled and endorses political strategies that reflect that intimate relationship.

48

P rag ma tism an d In stitu tion a l D e sig n

reasons to believe that left unchecked, the costs will be borne disproportionately by relatively disadvantaged participants. One important reason is that in any diverse society, existing resource asymmetries will tend to disproportionately distribute both the ability and the willingness of differentially situated participants to bear the costs of such experimentation. A second significant reason is that the very nature of experimentation in societies with substantial populations will likely generate a range of relatively small-­scale experiments, each of which is aimed, in part, to produce lessons for the society at large. This structure of experimentation forms the basis for many of the normative justifications of federalism as a governance mechanism. In such circumstances, some will succeed and some will fail, and the costs will fall on participants in the failed undertakings. This is the nature of experimentalism. Any asymmetrical distribution of costs or burdens clearly raises normative concerns regarding equity. Given the primary emphasis on the social nature of institutional experimentation, basic principles of equity suggest no particular social group should disproportionately bear the risks associated with experimentation. In addition to such normative concerns, there are the practical implications related to the importance of diverse participation in the experimental process. As we have suggested, experimentation relies on diversity both in inputs and in assessment of quality. Costs affect the incentives for individual participation. Therefore, pragmatists will be concerned with maintaining the incentives necessary for diverse participation throughout the process. For these reasons, the distribution of costs is another factor that pragmatists will incorporate into the experimental equation. One of the virtues of the approach to experimentalism that we defend in this book is that it recognizes the central importance of establishing an institutional framework that allows those who will bear the potential costs of experimentation to monitor the conditions under which it takes place.

VI.  Summary: Experimentation and Politics Pragmatists contend that in modern diverse societies, institutional choice is most appropriately seen as an ongoing process of inquiry and experimentation. The bottom line, of course, is that the task of deciding exactly which institutions work best, where, and to what end is, for a pragmatist, context-­specific and a subject for ongoing experimentation and contestation. As Dewey reminds us, the act of establishing boundaries “is still experimental. And since conditions of action and of inquiry and knowledge are always changing, the

49

C h ap te r 2

experiment must always be retried.”73 But the very process of experimentation is, for its own success, contingent on the maintenance of the appropriate method of social inquiry. This requires policies, practices, and, most important, institutions that allow relevant individuals and groups to foster, monitor, and maintain the kinds of social conditions under which experimentalism will thrive in the society at large.74 It nevertheless is easy to see how this process of experimentation and institutional choice will remain unavoidably political. There is simply no getting around it. In the choice of any institution, there is room, at least, for disagreement over both the quality of the outcomes we might anticipate it will generate and the initial conditions under which we can expect it to operate effectively. Social actors cannot reasonably expect a neutral, apolitical metric that might be applied to such decisions. Even under ideal conditions, there is no reason to expect consensus on such matters as how to weight competing normative criteria or how to judge whether empirical conditions sufficiently approach those necessary to make likely, let alone ensure, normatively attractive outcomes. The fact of pluralism we recognized in the introduction suggests that there will always be space for legitimate disagreement about such matters.

Dewey (1927, 33–­34). On this point, see Kitcher (2006).

73   74  

50

Chapter 3 The Appeal of Decentralization

I.  Decentralization Before we turn to our account of the priority of democracy, we need first to address an alternative approach to questions of institutional choice, an approach that has come to dominate contemporary thinking on these questions. Demands for decentralization have long been central to debates over the design and performance of political and economic institutions. They form the basis for justifications of economic institutions such as the market as well as other market-­mimicking mechanisms. The appeal of decentralization is, as we will see, among the defining elements of what are perhaps the most common justifications for markets. It influences discussions of the many salutary effects attributed to markets, from the ways in which they structure individual choice to their ability to coordinate complex systems of small-­n economic exchange. Decentralization also forms the basis for justifications of political institutions such as federalism and other varieties of multijurisdictional governance structures. In regard to these multijurisdictional systems, the argument for decentralization is somewhat more complicated, given the fact that any political governance structure involves some degree of centralization of decision making within the system. Nonetheless, justifications for such institutional arrangements tend to emphasize the value of breaking down the decision-­making structure into smaller components. The underlying idea is that government decision making will be more responsive to citizen interests as well as more accountable to citizen concerns if decision-­making responsibility is brought closer to the individual participants who will be most affected by the process. For many, decentralization seems to hold a nearly irresistible intuitive appeal. It taps into concerns about individual autonomy, emphasizing our capacity to make decisions for ourselves without coercion from others. And it similarly taps into concerns about efficiency and institutional effectiveness, grounded in a belief that things will work better if people are allowed to make their own decisions and to control their own affairs. In this sense, decentralization seems to embody the following notion: who knows better than I what

C h ap te r 3

I want and what would be good for me, and who, thus, is better situated to make it happen? The appeal of decentralization has been so pronounced in contemporary discussions of politics and institutional choice that, as we will show, decentralized institutional arrangements like the market have come to enjoy a default priority among the available institutional alternatives. Given this pervasive view—­that we should rely primarily on the market and other complementary decentralized institutions to coordinate social interactions—­we believe that it is necessary to assess the merits of the decentralization argument as the first step in our argument for the priority of democracy. Pragmatists believe that all institutional choices, including the initial choice of the institutional mechanism for making other future choices, are comparative in nature. The choice of a particular institutional arrangement is always a function of a comparison of the relative merits of the different institutional alternatives. The argument for the priority of democracy is, in significant part, based on a claim that democratic decision making will do a better job of facilitating the various tasks of institutional choice than will decentralized mechanisms like the market. To support this claim, we will make two arguments about the decentralized alternative. The first we take up in this chapter. Here we assess the strengths and weaknesses of decentralized institutional mechanisms and ask the basic question: do such mechanisms adequately address on their own terms the important tasks of effective institutional choice? If we were to conclude that they do so, then this would lend support to the default priority presently enjoyed by decentralized alternatives. If we were to conclude that they fail to do so, this would undermine the default priority and reorient contemporary discussions in the direction of a comparative assessment of the relative merits of the various institutional alternatives. For the various reasons that we set out in this chapter, we find that on their own terms, a familiar array of decentralized institutional mechanisms fail to satisfy the claims made on their behalf regarding effective institutional performance. In chapter 6, we take up a second argument about decentralized mechanisms. There we assess decentralized institutions in explicitly comparative terms, asking whether decentralized mechanisms can do as good a job or better than democratic decision making in terms of the criteria by which we justify the priority of democracy. We argue there that they do not. But here we begin with an assessment of the standard justification for the priority of decentralized social institutions. To properly motivate this discussion, we begin with a brief review of Hayek’s argument for the priority of the market. Hayek’s argument is an important one because it forms the theoretical basis for most contemporary arguments about the priority of decentralization.

52

Th e A p p e al o f D e c e n tra liz a tion

For this reason alone, it would be the appropriate place to begin the assessment of justifications of decentralized institutional arrangements. But there is a second important reason for pragmatists to engage seriously with Hayek. That is because of the significance Hayek properly accords to the problems of knowledge and information aggregation. Like pragmatists, Hayek highlights the importance of distributed knowledge in society. And, like them too, he believes that resolving the difficulties that arise from the unavoidably distributed character of knowledge and information poses a central challenge to those engaged in selecting or reforming institutions. Hayek notes that the original justification for markets was that they were a set of institutions that produced collectively beneficial outcomes through the contributions of self-­interested individuals: “The chief concern of the great individualist writers was indeed to find a set of institutions by which man could be induced, by his own choice and from the motives which determine his ordinary conduct, to contribute as much as possible to the need of all others; and their discovery was that the system of private property did provide such inducements to a much greater extent than had yet been understood.”1 On this account, markets produce these benefits through the salutary effects of competition. The most important of these effects involves the ways in which competition solves problems of effectively acquiring information and using knowledge. Hayek esteems “competition as a procedure for the discovery of such facts as, without resort to it, would not be known to anyone, or at least would not be utilised. . . . [C]ompetition is valuable only because, and so far as, its results are unpredictable and on the whole different from those which anyone has, or could have, deliberately aimed at.”2 Hayek argues that the most significant problem for society related to economic activity is “one of rapid adaptation to changes in the particular circumstances of time and place.”3 In order to adequately solve this problem, “it would seem to follow that the ultimate decisions must be left to the people who are familiar with the circumstances, who know directly of the relevant changes and of the resources immediately available to meet them. . . . We need decentralization because only thus can we insure that the knowledge of the particular circumstances of time and place will be promptly used.”4 The argument here is that decentralized decision making offers the best way of aggregating all of the local knowledge necessary to achieve an effective societal response to the problem. Hayek (1948, 12–­13). Hayek (1984, 255). 3   Hayek (1948, 83). 4   Hayek (1948, 83–­84). 1   2  

53

C h ap te r 3

Against that background, Hayek argues that markets are superior to systems of centralized planning.5 In characterizing the market, he places special emphasis on the more efficacious ways that the decentralized framework handles individual choice and coordination. On his account, decentralization draws out the benefits of individual choice by attending to two factors that significantly affect that choice: the information/knowledge factor and the incentive factor. In regard to knowledge, decentralized markets structure individual choice so as to allow individual actors to make those decisions about which they have the most knowledge, decisions about their own preferences over goods and services. In regard to incentives, so the argument goes, decentralization places the responsibility for the individual effects of the market on the individuals who make those decisions. Thus, it creates an incentive for each individual participant to take the time and resources necessary to get the decision right, in the sense that the resulting basket of goods and services matches their preferences better than any alternative. For all of this to work effectively at the individual level, there must be an institutional mechanism that effectively coordinates the myriad of individual choices.6 For Hayek, the relevant mechanisms are competition and the price system that it produces: “The price system is just one of those formations which man has learned to use (though he is still very far from having learned to make the best use of it) after he had stumbled upon it without understanding it. Through it not only a division of labor but also a coordinated utilization of resources based on an equally divided knowledge has become possible.”7 From Hayek’s account, we can see what is intuitively appealing about decentralization as well as what is really at stake for the problem of institutional choice.8 The appeal flows from claims about the benefits of and the effects Note that in our own analysis, we do not cast the comparison as simply one of decentralized markets versus centralized planning authorities. Rather, we are trying to assess a continuum that runs from decentralized coordination of small-­n interactions at one end to centralized large-­n decision-­making on the other. There are varying degrees of decentralization and centralization involved in the feasible set of institutional alternatives faced by any real society. 6   As Schelling (1978) rightly points out, and as we stress later, absent such a mechanism, patterns of decentralized exchange do not constitute a market and so often lack the salutary characteristics we ascribe to market exchange. On the role markets play as coordinating devices in contrast to common characterizations of them as the locus of competition, see Lindblom (2001). 7   Hayek (1948, 88). We add the emphasis for “equally divided knowledge” to highlight the importance that Hayek places on the equality condition for market participation. We come back to this point later in this chapter and in chapter 6. 8   As Hayek refined through the years his argument about economic decentralization, he developed a similar theme about political centralization. The basic idea is captured in the claim that “[t]o split or decentralize power is necessarily to reduce the absolute amount of power” (Hayek 5  

54

Th e A p p e al o f D e c e n tra liz a tion

on individual choice. The enjoyment of individual freedom and the pursuit of autonomous choice is a readily understandable benefit for people of many different ideological and intellectual persuasions. Moreover, anyone concerned about our ability to draw on the distributed knowledge existing in society will understand the importance of crafting institutions that structure institutional choice to do so. But in the end, the benefits of decentralized choice at both the individual and collective levels are contingent on the presence of institutional mechanisms that can effectively coordinate all of the small-­n exchanges animated by such choices. For advocates of any decentralized institutional arrangement, the burden of justification rests on claims about the consequences their preferred arrangements will produce. These claims, in turn, rest on prior claims about mechanisms; the conditions under which those mechanisms operate in predictable, attractive ways; and our confidence that those conditions are relatively robust, in the sense that we can plausibly expect that they, in fact, will regularly or commonly obtain in the world. Hayek sought to satisfy the task of identifying the relevant institutional mechanisms in the case of the market with his characterization of competition, the structure of individual choice, and the price system. Yet, assuming the mechanisms he identifies are plausible, several important questions remain. Under what conditions will these mechanisms work effectively? And how robust are those conditions? That is, to what extent will the positive consequences Hayek attributes to his preferred institutional arrangement persist as we deviate from these optimal conditions? More generally, and perhaps most crucially, what mechanisms are available to us for purposes of monitoring and calibrating the conditions of effective institutional performance? It is to these questions about the market that we now turn.

II.  Privileging Markets I: The Textbook Account Markets are institutions in just the sense we have in mind—­mechanisms for coordinating ongoing social interactions. This is a truism.9 Yet markets occupy an unusual place—­indeed, one that is unusually and, we contend, unjustifiably privileged—­in extant theories of how institutions emerge, operate, and change. This becomes clear when we inquire what it might mean to claim that markets are institutions. We address this issue in this section and the next. 1944, 166). For a subsequent argument to the effect that political decentralization diminishes the power of government authorities, see Hayek (1976). 9   Kreps (1990b, 5–­6, 187–­98); Lindblom (2001).

55

C h ap te r 3

In the introduction, we depicted institutions as sets of rules that more or less persistently and systematically coordinate ongoing social, economic, and political interactions. We suggest that since institutions often require individuals to act in counterpreferential ways, they must establish both first-­order and second-­order expectations. They must, in other words, establish not simply mutual expectations concerning individual behavior but also expectations about what will occur if some individual or group deviates from those expectations. By this, we mean simply that institutional rules establish a generalized expectation that third parties, even when they themselves are not directly involved in some interaction, will react to the noncooperative behavior of the parties who are directly involved in it.10 It is important to distinguish at the outset among three concepts that economic and political theorists too often conflate. These are exchange, competition, and markets. Exchange is pervasive in social life. It occurs whenever two or more individuals transfer goods and/or services among themselves. Exchange, however, does not automatically imply market interactions. The latter emerge from the need to predictably coordinate exchange in multilateral, increasingly impersonal settings.11 Only once exchange starts to occur under certain specifiable conditions do we observe the institutionalized interactions and, importantly, the attractive outcomes characteristic of markets. These conditions operate in various ways to establish and sustain varying degrees 12 of competition among parties to exchange relations. One virtue of standard microeconomic analysis is that it clearly identifies these conditions. In standard textbook versions, microeconomic analysis shows that under specifiable conditions, (1) an equilibrium generated by market competition always yields a Pareto-­efficient allocation, and (2) any such allocation consti13 tutes an equilibrium. These results, roughly speaking, are the fundamental 10   It is, of course, a crucial part of any institution to identify these third parties. Informal social norms may require all or most members of a community to react to noncooperation. More formal, centralized institutions, by contrast, typically will assign this task to particular officials. This is unimportant for present purposes. 11   Lindblom (2001, 37). 12   Here we have in mind something like what economists take as the “the natural meaning of ‘competition,’ that each participant in the market is so small that he will have no effect on the behavior of others” (Stiglitz 2002, 467). Put otherwise, to the extent they exist, the conditions of competitive exchange operate to mitigate asymmetries of power. 13   For present purposes we rely almost exclusively on a single recent textbook account (Kreps 1990b). This textbook is especially valuable for our purposes because it is both written by a highly esteemed economist and is admittedly “chatty” (p. xvi). It consequently is more explicit on matters that concern us here than are many texts. For other clear presentations of these issues, see Myles (1995) and Przeworski (2003).

56

Th e A p p e al o f D e c e n tra liz a tion

theorems of welfare economics.14 They establish the existence and attractiveness of market equilibrium. This is a remarkable convergence. As Kreps comments: “You should now be hearing choirs of angels and choruses of trumpets. The ‘invisible hand’ of the price mechanism produces equilibria that cannot be improved upon.”15 But Kreps also prompts us to ask what it actually means to say economists can establish these things. Here it is helpful to raise two subsidiary questions. What, more precisely, are economists doing? And how robust are the conditions under which their analyses hold? By raising these questions, we can highlight why each of the three tasks of institutional analysis is important for political and economic theorists. Consider the first question. What economists offer is an analysis of the existence and qualities of equilibrium outcomes that does not explain how markets actually generate such nice equilibria. In other words this demonstration “doesn’t provide . . . any sense of how markets operate. There is no model here of who sets prices, or what gets exchanged for what, when, and where.”16 This is because what economists offer “is a reduced form solution” that “describes what we imagine will be the outcome of some underlying and unmodeled process.”17 Standard microeconomic accounts offer little understanding of 18 precisely how “market/exchange mechanisms” actually operate. Kreps thus admits that the claim that economic agents will find their way to equilibrium in a decentralized process is a “rather heroic assertion.” He suggests that it “seems natural to think that we could increase (or decrease) our faith in the concept of Walrasian equilibrium if we had some sense of how markets really do operate.” Progress on this task can be made “only if we are more specific about the institutional mechanisms involved” in market interactions.19 Economists, to be sure, have generated some good work on this score. But as Kreps asserts, the “exploration of more realistic models of markets is in relative infancy.”20 And, moreover, that work encounters important difficulties. Kreps (1990b, 199–­202, 286–­89). Kreps (1990b, 200). As Przeworski (2003, 28) rightly reminds us: “This is a miracle, not the real world, not a description of any real economy (although some people who believe in miracles take it for such) but only a statement of the exacting conditions that are necessary for such a miracle to occur.” 16   Kreps (1990b, 195–­98). 17   Kreps (1990b, 195, 187). 18   Kreps (1990b, 195, 190). In other words, “why the economy should actually reach the equilibrium is not entirely clear” (Myles 1995, 25–­26). While this is true for standard microeconomics, it clearly is not the case for recent work by sociologists (Swedberg 1994) or economic historians (Greif 1997). 19   Kreps (1990b, 187, 190). 20   Kreps (1990b, 190, 195, 197). 14   15  

57

C h ap te r 3

On the one hand, economists face difficulties in actually explaining how the price mechanism generates equilibrium outcomes. On the other hand, they face the challenge of persuasively demonstrating how equilibria attained in particular settings retain the attractive normative properties of competitive 21 equilibria that economists can establish in the general case. In order to appreciate the difficulties here, we need to turn to our second question. Economists can establish the existence of normatively attractive equilibria only under reasonably restrictive conditions. Some of these conditions are 22 purely technical, and we set them aside. Others revolve around questions 23 about the adequacy of efficiency as a normative criterion. While efficiency is a rather anemic normative criterion, we grant for purposes of the present argument that market institutions can be justified in terms of their efficiency-­ enhancing effects. We thus set these conditions aside as well. What we do care about at the moment are several further conditions necessary to the argument that market interactions generate normatively attractive equilibria. These conditions essentially are meant to ensure that no It is important, in other words, not to presume that the demand for realistic models is satisfied simply by a move from general to partial equilibrium analysis. The latter simply recapitulates both the explanatory shortcomings and normative character of more general approaches. For example, what if we want to try to explain how the price mechanism generates an equilibrium in particular cases? First, moving to “partial” equilibrium analysis is not terribly helpful insofar as it, too, neglects underlying institutional mechanisms (Kreps 1990b, 263–­64, 286). In essence, it simply amounts to shifting the markets surrounding those under scrutiny into the same domain as what Kreps (1990b, 263–­64) identifies as the “larger social and political environment which general equilibrium takes as given.” Second, we cannot assume that the particular case will have the attractive normative properties that the general equilibrium argument establishes. That, however, is pretty much what partial equilibrium analysis does. Since it focuses on some particular realm of exchange, it takes all others as given. It thus allows the analyst to assume that the interactions under consideration in fact meet the restrictive conditions presupposed by “nice” equilibrium outcomes and to not ask, one way or the other, about all the other interactions that a general equilibrium argument would encompass (Kreps 1990b, 279–­84). But from the vantage point of general equilibrium analysis, this process of extraction clearly is artificial and thus is a theoretical move for which the analyst must offer a plausible defense. Otherwise, there is no reason to expect that whatever equilibrium emerges in particular cases is normatively attractive in even efficiency terms. 22   For instance, the general equilibrium results presume a set of technical assumptions about consumer preferences—­they are “convex, continuous, nondecreasing, and locally insatiable” (Kreps 1990b, 199–­200, 287). 23   The second welfare theorem is sensitive to the initial distribution of endowments. Depending on how endowments are distributed (or redistributed) at the outset, market exchange can generate efficient but highly inequitable outcomes. This observation often generates “normative” arguments about the degree and feasibility of establishing such initial equity (Kreps 1990b, 200). We want to avoid debating the relative importance of equity and efficiency as normative criteria. 21  

58

Th e A p p e al o f D e c e n tra liz a tion

producer’s or consumer’s choice exerts significant influence on the choices of other agents. In other words, standard microeconomic models are populated by agents who are free and equal.24 The conditions necessary to ensure this are significantly less easy to relax than those we have set aside thus far.25 They fall into two categories. The first category concerns the role of information and time in market exchange.26 These conditions can, of course, be relaxed with some strain if we incorporate futures markets and increasingly clairvoyant agents.27 Awareness of the constraints that information and time impose on economic exchange prompts economists to focus on the “transaction costs” involved in processes of search and monitoring. This, in turn, provided the impetus for a large literature on economic institutions.28 This literature largely is concerned with the ways that information and time constraints produce inefficiencies. What is less often noted is that while transaction costs affect all parties to economic exchange, their burdens typically are distributed unequally. For example, information imperfections typically confer advantage on one among the parties to an exchange in ways that clearly subvert the equal status of others, thereby threatening the normative attractiveness of the resulting outcomes. As we establish in subsequent sections, the recent literature on political and economic institutions typically implies that institutions function to mitigate such asymmetries. To the extent that it does so, this literature tacitly but systematically intermingles its normative and explanatory tasks. The second category of conditions that concern us is meant to establish the sort of competitive environment needed to ensure that individuals choose independently and, so, to preclude such things as externalities, monopolies, and Here, it is helpful to remember Hayek’s emphasis on the equal treatment of bearers of knowledge in his justification of the market. 25   Myles 1995, 48–­9. 26   Kreps 1990b, 193–­5. 27   Kreps (1990c 216f); Myles (1995, 48–­9). Economists disagree sharply regarding the extent of the strain. Some, like Stiglitz (2002), claim that even small amounts of imperfect information derail standard microeconomic arguments. As he says, even if we acknowledge the fact that information is costly, “economies with information imperfections would not be Pareto efficient” and there will always be feasible government “interventions in the market that could make all parties better off (Stiglitz 2002, 477–­78). In this regard, two points are important. First, in a sense, this argument clearly sustains our own case against the priority of markets. But second, it is not as radical as the case we advance in the next section, for Stiglitz clearly conceives what he calls “the economic role of the state” as market enhancing and so tacitly retains the presumption we seek to displace. 28   See, e.g., Williamson (1985); North (1990a); and Eggertsson (1990). Kreps (1990b) clearly traces the connections between standard microeconomic analysis and the transaction-­cost economics that underlies much of the recent rational-­choice literature on institutions. 24  

59

C h ap te r 3

public goods. These conditions, in other words, are meant to ensure that agents who populate these models are parametric in the sense that they all are more or less equally situated price takers.29 The difficulty is that absent such conditions, not only will market competition potentially generate an inequitable distribution along the efficiency frontier (the normative problem we set aside above), but it generally cannot be relied upon to generate efficient outcomes at all.30 Both categories of conditions are necessary for markets to operate in something resembling “textbook” fashion. Thus, we do not question whether they must hold in order for economists to generate their analytical results. We simply wish to highlight the quite restrictive character of those results and to suggest how awareness of the conditions needed to sustain normatively attractive equilibria prompt us to ask two significant questions. First, how are these restrictive conditions embodied in actual economic institutions? And second, do those institutions operate to shape economic exchange into competitive markets, and if so, how? Likewise, when we question whether these conditions obtain, we are not challenging the economist’s preferred normative criterion—­efficiency—­with some other criterion (e.g., equity) but rather questioning whether the economists’ normative claims are warranted on their own terms. We then, of course, also need to ask whether there exist alternative means of coordinating ongoing interactions that are preferable in normative terms—­whether those be terms of efficiency or otherwise. Here we return to the claim that microeconomic analysis presumes further inquiry into the “institutional framework” of market interaction.31 It is important to notice an ambiguity in this claim. On the one hand, standard accounts readily acknowledge that the proper operation of markets presupposes an exogenous institutional framework, including well-­defined property rights and enforceable contracts.32 On the other hand, it less commonly is acknowledged that because they offer only reduced-­form analyses, microeconomic accounts do not properly address the ways an informal institutional framework of roles, norms, and prerogatives defines the scope and operation of markets. It is a truism that markets cannot themselves produce institutional arrangements of the first, “external,” sort. And as we argue in what follows, there is little reason to expect that they can generate institutions of the second, “internal,” sort. But we cannot understand how markets in fact operate without specifying how 29   Kreps (1990b, 202, 264–­5); Myles (1995, 20). On the conceptual distinction between parametric and strategic rationality see Elster (1979). 30   Kreps (1990b, 202–­4, 288–­92). 31   Kreps (1990b, 190). 32   Kreps (1990b, 263); Myles (1995, 20, 48–­49).

60

Th e A p p e al o f D e c e n tra liz a tion

informal institutions and decentralized exchange interpenetrate to establish the sort of “trading environment” that allows us to treat markets as institutions.33 At this point, a skeptical reader might well object. She might suggest that what we really have established is a virtue of standard microeconomic analysis. For such analysis prompts us to seriously and systematically consider the role institutions of various sorts play in structuring economic exchange into more or less competitive markets. In one sense, such a skeptical reader would be correct. Standard microeconomic accounts privilege markets in a relatively benign sense insofar as they treat other institutional forms primarily as responses to imperfections in or failures of markets. On this view, markets serve as the default mechanism for coordinating social interactions, while various alternative institutional mechanisms represent “means of achieving the benefits of collective action in situations where the price 34 mechanism fails.” As a purely analytical move, this raises little concern. A very significant problem arises, however, when theorists unself-­consciously transport this relatively benign privilege from its native analytical environment into their properly explanatory and normative enterprises. This is clearest when theorists characterize a range of social and political institutions—­ families; the rule of law generally or enforceable contracts and property rights in particular; various constitutional provisions, such as federal arrangements or separation of powers; and even democracy—­as “market preserving”35 or “market supporting.”36 For this characterization sustains efforts to explain and, tacitly, to justify institutions primarily in terms of how they contribute functionally to the operation of markets. It is this way of privileging markets that we find objectionable. We now turn to that topic.

III.  Privileging Markets II: Explaining Institutional Emergence 37

Consider Calvert’s abstract game theoretic treatment of institutions. On his view, institutions consist of sets of rules that are sustained as an equilibrium Swedberg (1994); Lindblom (2001). Arrow (1974, 33). Arrow actually refers to “organizations” but, in fact, also uses the term “institutions” more or less interchangeably. For an effort to differentiate institutions and organizations, see North (1990a, 4–­5). Most importantly for our present purposes, Arrow explicitly refers to just the sorts of entity—­such things as governments, business firms, moral norms, and markets—­ that we consider institutions. 35   Weingast (1995). 36   See McMillan (2002, 11) and Rodrik (2007, 156). 37   Calvert (1995a; 1995b). Calvert’s papers are especially useful insofar as he aims to articulate the theoretical structure that informs several influential historical accounts that claim political 33   34  

61

C h ap te r 3 38

(and so are self-­enforcing) of some underlying repeated game. He represents the basic underlying game as an indefinitely repeated N-­person prisoner’s dilemma where players are randomly matched for pairwise interactions. He then models a set of increasingly complex versions of this game in each of which he identifies threshold values for the discount parameters of players, the size of N, and eventually the costs of communication that sustain a variety of cooperative equilibrium outcomes, both noninstitutional and institutional. These outcomes all involve some more or less elaborate variation on conditional cooperation in which players opt to cooperate today in light of the credible threats others can make to retaliate against noncooperators in future interactions. They range from what we might call specific reciprocity, in which mutual cooperation remains noninstitutional, through the sorts of generalized reciprocity characteristic of informal social norms and on through increasingly formalized institutions that incorporate first multilateral communication, then centralized communication, and finally centralized communication coupled with protections against higher-­order malfeasance.39 This is in many ways a remarkable demonstration. But notice that, as Calvert himself acknowledges, he does not actually explain the emergence of 40 any social or political institutions. His analytical exercise generates an existence result that, by identifying the conditions under which they can emerge, institutions emerge and change to facilitate the creation and stabilization of markets. See North and Weingast (1989); Milgrom, North, and Weingast (1990); and Greif, Milgrom, and Weingast (1994). 38   Calvert (1995a, 228–­33; 1995b). Calvert actually is somewhat ambiguous on this point. He pointedly suggests that “Institution is just a name we give to certain parts of certain kinds of equilibria.” In that sense, he tacitly seeks to dispense with any reference to the rule-­like features of institutions. Yet he also characterizes such equilibrium institutions as “no more than a regular behavior pattern sustained by mutual expectations about the actions that others will take when anyone violates the rule—­whether the violation be defection, lying, nonpayment of fees, extortion, or malfeasance in office. The institution is just an equilibrium” (Calvert 1995b, 74, 73). And here, his definition of institution clearly is parasitic on the recognition that institutions are composed of rules that on any conception must amount to more than a mere behavioral regularity. This last observation is in keeping with Calvert’s own claim that there is “no better” way to conceive of institutions than as rules (Calvert 1995a, 217). 39   Specific reciprocity obtains when the strategy any given player adopts is conditional only on whether that particular other player cooperates or not. Generalized reciprocity consists in a strategy profile for all players in which it is equilibrium play for each to sanction any player who fails to cooperate, regardless of whether that player is the one with whom they themselves are matched in the current round. 40   Calvert (1995b, 80–­82) remains agnostic regarding the variety of accounts that might explain the emergence of equilibrium institutions. We return to this issue at length below and are decidedly not agnostic. For present purposes, we wish merely to highlight the analogy between

62

Th e A p p e al o f D e c e n tra liz a tion

establishes the possibility of equilibrium institutions. The difficulty, of course, is that as with nearly any repeated game in which players discount the future, the games Calvert examines generate multiple equilibria. This unavoidable coordination problem sets the basic explanatory task for institutional analysis. Which, if any, of the available equilibrium institutions actually will emerge from social and political interaction? How might we best explain the way institutions are generated? Under what conditions are such institutions likely to be stable, and under what conditions are they likely to change? These are crucial questions if we aspire to fulfill the explanatory ambitions of institutional theories.41 They are questions that Calvert explicitly leaves unanswered.42 For present purposes, these questions are important because they highlight choice accounts of institutional emergence the way that various rational-­ privilege markets. There is no single rational-­ choice explanation of institutional genesis. While rational-­choice theorists typically depict institutions as emerging from the decentralized efforts of rational actors to resolve recurrent strategic problems, they advance several competing accounts of this process. These accounts differ in (1) the ways in which the theorists conceive and model the basic social interactions that produce institutions and (2) the features of the social context that they invoke in order to resolve the strategic problems inherent in these interactions. There exist at least three distinct rational-­choice explanations for the de43 centralized emergence of social institutions. The first is a theory of arbitrary coordination on social conventions.44 The second is a market-­based theory of exchange and selection through competition.45 The third is a bargaining theory that explains the emergence of institutions in terms of the asymmetries of power in a society.46 On each account, social institutions operate to resolve the strategic problems inherent in social interactions characterized by multiple equilibria: any of the available equilibria might coordinate social interaction, but the actors need to identify a common equilibrium to pursue. Each explanatory account starts from the micro level, arguing that social institutions emerge as equilibrium outcomes in repeated interactions among Calvert’s analytical exercise and standard microeconomic analysis. The crucial point is that neither is explanatory. 41   Calvert (1995a, 228). 42   Calvert (1995b, 70, 80–­82). 43   Knight (1995) offers a detailed discussion of the basic features of the three approaches. 44   Schotter (1981). 45   North (1990a). 46   Knight (1992).

63

C h ap te r 3

small numbers of rational actors. The process of institutional development culminates when one set of rules is established as the common set of institutions for that population. Over time, the maintenance and stability of that institutional arrangement is a function of the persistence of the conditions under which it was originally created. While the three approaches share the view that social institutions can emerge from decentralized interactions, they identify very different mechanisms to explain institutional emergence, maintenance, and change. The social conventions approach is animated by the idea that institutions 47 coordinate beliefs on mutually beneficial outcomes. This explanation of institutional genesis rests on an account of how conventions are recognized and coordination is achieved that builds on Thomas Schelling’s conception of “salience.”48 The causal mechanism central to convention accounts is the existence of focal points that coordinate social expectations. Some feature of the social context in which the relevant interaction occurs establishes the salience of a particular cooperative outcome. Convergence on the focal point leads to the establishment of a convention among members of the group that coordinates their ongoing interactions. The key explanatory claim is that institutional rules evolve arbitrarily. The coordination process is unconsciously driven by whatever salient factor attracts and focuses the attention of the actors. The nature of the final institutional arrangement is not determined by interests of any particular actor or group of actors, and so the explanation of institutions that are generated does not invoke distributional differences pro49 duced by different rules. Advocates of the convention approach endorse quite specific substantive conclusions about institutional genesis, conclusions that focus on the collectively beneficial nature of social institutions. They hold that the emergence 47   This approach represents one of the earliest efforts to employ formal game-­theoretic models to develop a general theory of decentralized emergence and change (Schotter 1981). It has long been the foundation for efforts to explain the emergence of specific social institutions, including language (Lewis 1969), social norms (Ullman-­Margalit 1978), and property rights and morality (Sugden 1986). 48   Schelling (1960). 49   This description by Schotter (1981, 79) captures the underlying nature of the process: “The important point to be made, and the one that makes this approach unique, is that the social institution that is actually created . . . is a stochastic event, and that if history could be repeated, a totally different convention could be established for the identical situation. The point is that the set of institutions existing at any point in time is really an accident of history and that what exists today could have evolved in a very different manner.” All of that said, this approach adopts an unjustifiably naive view of how salience emerges and enters into strategic interactions (Johnson 2000; 2001; 2002). We do not pursue this point here.

64

Th e A p p e al o f D e c e n tra liz a tion

of institutional rules produces an outcome that is Pareto-­superior to uncoordinated interaction. Here the emphasis on collective benefit is crucial because they are comparing the aggregate benefits produced by different social outcomes. Many advocates of the convention approach commonly acknowledge that coordination processes will not necessarily lead to Pareto-­optimal institutional arrangements and offer various reasons for this failure. What is crucial to note here, however, is that such theories do predict that changes in existing rules will involve Pareto improvements over the status quo. They justify this restriction by reasoning that rational actors will not respond to a movement toward a Pareto-­inferior alternative. 50 A variety of theorists endorse the contract and selection account. Eggertsson succinctly characterizes its essential features: Changes in contractual forms may often be a long process, particularly when there is a lack of experience with arrangements that would be best suited to a new situation. Once a successful experiment has been made, the forces of competition establish new equilibrium contracts. It is also reasonable to expect that a community that has a very long experience with stable technology and a stable range of relative prices has settled on contractual forms that minimize costs for each branch of production, given the state of knowledge about contractual arrangements and the basic structure of property rights.51 The contract and selection account consists of two main components: individual exchange and the competition over alternatives. Like the convention approach, this approach too is grounded in the logic of mutual benefit. The motivation for and consequences of exchange are quite familiar. Any two social actors who perceive that they can benefit from mutual exchange must agree on the terms of that exchange. In other words, they must come to terms on a contract that will govern their actions over the course of their interaction. In the context of the theory, these contracts constitute social institutions that social actors intentionally produce to facilitate the achievement of socially beneficial outcomes. The main mechanism that explains the selection among possible contracts is voluntary agreement. But these contracts can take many forms and serve merely as the set of possible institutional forms in a society. As a result, those who endorse the contract and selection approach constrain the mechanism of voluntary agreement with an assumption See, e.g., Brennan and Buchanan (1985); Heckathorn and Maser (1987); Williamson (1975; 1985); North (1990a); Ostrom (1990). 51   Eggertsson (1990, 55 [emphasis in original]). 50  

65

C h ap te r 3

regarding the kinds of contracts social actors will try to produce. The assumption is grounded in Coase’s insights about the effects of transaction costs.52 For example, North argues that economic actors will contract for rights, which minimize the transaction costs involved in their ongoing exchange, subject to a set of constraints on their ability to identify the most efficient contract (e.g., ideology, lack of knowledge, lack of capacity.)53 They minimize transaction costs in order to maximize the benefits gained from the property. Here it is important to note that this is merely a discussion of the motivations of the actors in an individual exchange. As Eggertsson clearly points out, the more sophisticated versions of the transaction-­cost approach invoke more than the intentions of the actors in an individual exchange to explain the emergence of social institutions. Individual exchanges merely produce the variety of possible institutional forms; the second mechanism invoked by the contract approach is selection via competition. Many explanations of institutional development and change situate the decision to establish social institutions in the context of a market or a market-­like environment. The main influence of the competitive environment on the choice of institutional form rests in the pressure it supposedly exerts on the institutionalization process. For these explanations, competition serves as a dynamic selection mechanism that determines the survival of various institutional forms on grounds of reproductive fitness. This is the logic behind the Alchian model of evolutionary competition employed by most economic analyses of institutional genesis.54 On this account, competition selects the institutional form that maximizes the aggregate benefits of the exchange. In doing so, it neutralizes arbitrary asymmetries of power among the actors, asymmetries that some actors might use to alter the institutional selection process in the direction of alternative outcomes that they would find more desirable in distributional terms. By invoking competition as a selection mechanism, advocates of this approach draw on a particular conception of the social context in which these social interactions take place. The extent to which competitive pressure will effectively select for efficient institutions depends on the degree to which the social context embodies the empirical conditions necessary for competition to operate. The bargaining approach explains institutional genesis primarily in terms of 55 distributional consequences. On this account, social institutions emerge as Coase (1960). We offer a detailed analysis of Coasian bargaining below. North (1990a, 37–­43). 54   Alchian (1950). 55   See, e.g., Heckathorn and Maser (1987); Bates (1989); Ensminger (1992); Knight (1992); G. Miller (1992). 52   53  

66

Th e A p p e al o f D e c e n tra liz a tion

by-­products of strategic conflict over substantive social outcomes. The mechanism of selection among the possible institutional alternatives is bargaining among the actors over the various alternatives. Institutional development is a contest among actors to establish rules that structure outcomes to those equilibria most favorable for them. As in any bargaining situation, some factors differentiate actors and thus influence the bargaining outcome in favor of one of the parties. These factors are what we usually refer to when we speak of bargaining power. In earlier work, Knight developed a theory that identifies the resource asymmetries characteristic of a society as the main factor explaining the resolution of bargaining over social institutions.56 Because it emphasizes resource asymmetries, the bargaining approach incorporates a mechanism of institutional creation that reflects an important feature of many social interactions: social actors suffer significant costs if they fail to coordinate on equilibrium outcomes, yet those costs need not be suffered uniformly. When social actors are aware of these differentials, this awareness can influence the credibility of certain strategies. The main thrust of the argument is that asymmetries in resource ownership affect the willingness of rational self-­interested actors to accept the bargaining demands of other actors. In explaining how particular social institutions are established, bargaining consists in attempts by players to commit to particular rules of behavior where the credibility of any such commitment is conditional on asymmetries in the resources available to the bargainers. On this account, social actors who have fewer resources or less advantageous alternatives will be more inclined to acquiesce to commitments made by those who are better advantaged. In this way, the existence of resource asymmetries in a society can significantly influence the selection of equilibrium institutions. If institutions arise out of bargaining interactions, we would anticipate the following emergence process. Bargainers will resolve their particular interactions on terms that reflect the commitment strategies of those whose relative resource advantage lends them credibility. Through a series of interactions with various members of the group, actors with similar resources will establish a pattern of successful action in particular types of interaction. As others recognize that they are interacting with one of the actors who possess these resources, they will adjust their strategies to achieve their best outcome given the anticipated commitments of others. Over time, rational actors will continue to adjust their strategies until an equilibrium is reached. As this becomes recognized as the socially expected combination of strategies, a self-­enforcing Knight (1992).

56  

67

C h ap te r 3

social institution—­one that operates to the relative benefit of advantaged actors—­will be established. We now have sketched three competing rational-­choice accounts of institutional emergence. The plausibility of the explanation these competing accounts offer will depend, in large part, on whether the underlying conditions assumed by the respective models actually obtain. The coordination approach will perform best when all actors derive equal benefit from the institutional rule. The contract and selection approach will offer the best explanation when highly competitive conditions obtain. The bargaining approach will produce the best account when actors have asymmetrical bargaining power and stand to derive unequal benefits from whatever institution emerges from their interactions. The more fully actual empirical or historical circumstances approximate one or another of these sets of conditions, the greater the explanatory power we would accord to the respective account. Casual observation suggests that the conditions presupposed by the bargaining approach are more prevalent than those assumed by the other two 57 theories. As a result, we contend that the bargaining approach is more robust than its competitors. Notice that we do not claim that the convention or contract-­competitive selection approaches are false across all cases, only that they are likely to produce plausible explanations of institutional genesis in a considerably narrower range of cases. In other words, we simply anticipate that the bargaining account will accommodate the assumption of rational, self-­interested behavior for a broader range of circumstances. The relative robustness of the bargaining account has important implications for both explanatory and normative claims about the priority of markets as an institution of social coordination. These implications become clear when we differentiate two ways that market institutions enter into competing explanations of institutional genesis. Markets, or market-­like mechanisms, enter into the convention and contract/selection accounts as part of the explanation to the extent that each account invokes competitive pressures that select among the institutional alternatives. Markets also enter into those accounts as part of the explanandum, as one of the range of possible institutional alternatives for coordinating social interactions. Those who argue for the normative priority of markets relative to other institutional forms presuppose Empirical studies in various comparative settings support this conclusion, e.g., Allio et al. (1997); Binder and Smith (1997). That said, brute induction is unlikely to settle this matter in any definitive way. For some indication of the persistent theoretical difficulties, see Knight and North (1997). 57  

68

Th e A p p e al o f D e c e n tra liz a tion

that the following positive claims about these two features of the explanation actually hold. First, the conditions for the existence of market-­like competitive pressure are generally prevalent empirically and readily reproducible in decentralized social environments. Second, impartial market institutions that neutralize the effects of power asymmetries are the types most likely to be produced by a decentralized process of institutional emergence and change. The relative robustness of the bargaining account suggests that both of these claims are quite dubious. When we compare alternative accounts of decentralized institutional emergence, it is important to recall that the conditions necessary for the existence of market-­like competitive pressure are extremely restrictive. Consequently, we should anticipate that such competitive pressures only rarely will govern the decentralized emergence of social institutions. Moreover, to the extent that actual conditions depart from those that market competition demands—­levels of individual freedom and equality sufficient to sustain parametric action—­we will more and more closely approximate the conditions of asymmetrically distributed resources that the bargaining approach presupposes. All of this leads us to an important expectation: the types of institutions most likely to emerge from such a decentralized process will systematically favor the interests of relatively advantaged social groups rather than neutralizing the effects of resource asymmetries in ways that would facilitate effective market competition. In the last section, we highlighted the fact that markets generate normatively attractive outcomes only under highly restrictive conditions. Our confidence in markets as a means of coordinating social interaction rests, therefore, on our estimation of the extent to which those conditions actually obtain or might be brought about in particular circumstances. This, of course, poses substantial problems of creating and sustaining institutions that will effectively monitor and maintain effective market performance. In this sense, microeconomic analysis of how markets operate naturally leads us to take up the task of explaining how institutions emerge. In this section, we have demonstrated that the most robust account of institutional emergence actually undermines just the sort of confidence required to sustain the privilege frequently accorded to markets in discussion of institutional emergence and change. Here it is important to be clear about the implications of our analysis. In their pure form, the three theories we sketch here characterize one crucial feature of the environment of institutional emergence and change in exactly the same way: the process occurs in a context that is open and unconstrained by formal institutional arrangements. It is what many might have in mind when they talk of a natural environment, and it can be interpreted as the

69

C h ap te r 3

limiting case of a laissez-­faire approach to institutional selection.58 Although the theories explicitly seek to explain institutional emergence and change in such an environment, they also characterize the ongoing operation of these institutions. They describe both the types of institutions that would persist and the conditions under which they would continue to operate, absent some change in those conditions. Our analysis suggests that environments unconstrained by formal institutional arrangements are most likely to be characterized by the conditions the bargaining account presupposes. And these are the conditions least likely to facilitate effective operation of markets. In fact, we expect that in most cases, the institutions, including market institutions, that emerge from decentralized social interactions will embody sorts of asymmetries at considerable variance with the conditions of free and equal participation that the efficient operation of markets presupposes. Therefore, one of the important implications of our analysis for the basic question of institutional choice is that, absent some effort to alter the context in which institutional arrangements are initially selected and subsequently operate, markets will operate, at best, quite imperfectly and, thus, are not a good candidate for an ex ante first-­order priority vis-­à-­vis other institutional forms.59 We suspect that proponents of market priority understand this last point, at least implicitly. It is clear, we hope, that there is a pronounced resemblance between the conditions of unconstrained institutional selection that the coordination and contract/competition accounts each presuppose and the conditions presumed necessary for effective operation of the market. We believe this resemblance reflects a failure to properly differentiate between the analytical, explanatory, and normative tasks that political and economic theorists face. Simply put, those who accord priority to markets commonly import an assumption regarding initial conditions drawn from their analytical account of the effective operation of markets into their explanatory account of how institutions emerge and change. In so doing, they are engaging in sleight of hand with considerable normative implications.60 At a minimum, they owe us some justification for this theoretical move. Absent such justification, it simply is a In other words, we have sketched these three accounts in deliberately abstract form in order to capture crucial differences in their underlying logic. To this end, we envision rational agents interacting in an “institutionally unbound” setting. 59   We set aside for the moment the broader implications for other types of institutional arrangements. In the next chapter, we acknowledge that our analysis has significant implications for democratic institutions as well, establishing a substantial burden of justification for these institutions. 60   Unger (2007, 4n) puts this point differently. With a characteristic rhetorical flourish, he says of the way that practitioners of the “new institutional economics” tend to elide the distinction 58  

70

Th e A p p e al o f D e c e n tra liz a tion

conceptual mistake to conflate, as they do, the different ways in which market mechanisms inform their analyses. At another level, it may suggest an inchoate acknowledgment of the normative problems that follow from an explanatory account grounded in asymmetries of power rather than in market-­like competition. Among the beneficial features of the bargaining account is that it highlights more starkly the burden of justification that confronts proponents of any social institution.

IV.  Nonmarket Decentralized Mechanisms I: Coasian Bargaining One might question whether the limitations of markets exhaust the possibilities for the justification of a first-­order priority for a decentralized mechanism for coordinating social interactions. Might it be possible to identify an alternative decentralized institutional mechanism (or set of mechanisms) that will avoid the daunting problems of monitoring and maintaining effective market performance? Such a mechanism might be designed either as an alternative to markets or as a supplemental remedy to market failure. In this section and the next two, we examine three candidates—­one instantiating the laissez-­faire logic of Coasian bargaining, a second capturing important aspects of the notion of community, the third appealing to technocratic neutrality—­that might serve such a purpose. All three mechanisms appeal to decentralization, to the hope that we might coordinate our ongoing social and political interactions without relying on centralized decision-­making mechanisms. Each prompts theorists to adopt a different analytical strategy. None, it turns out, enable us to evade issues of robustness and the problem of effective monitoring and maintenance. Indeed, given that each of the mechanisms we examine resembles markets in the sense that it operates effectively only under a restrictive set of initial conditions, it seems unlikely that they might—­even in combination—­ provide an effective way of coordinating anything like the broad range of ongoing social, political, and economic interactions in a complex, heterogeneous society. Consider first Coasian bargaining. Coase’s analysis of the treatment of negative externalities in the law is probably one of the most famous contemporary formulations of the idea of the power of mutual exchange.61 His argument is attractive in no small measure because he not only recognizes that there are a between their analytical, explanatory, and normative tasks: “There are few more striking instances of the right-­wing Hegelianism—­the real is rational—­that pervades the social sciences today.” 61   Coase (1960).

71

C h ap te r 3

plurality of institutional mechanisms from which we might choose as we seek to coordinate our ongoing interactions, but he urges us not to prejudge the matter.62 The problem of institutional design or reform, from his perspective, is best approached as a problem for sustained inquiry. The logic of Coase’s argument is deceptively simple; its ambition is extensive and profound. It was promulgated as an analysis of the social welfare implications of tort rules and regulations. It was a response to the then-­extant understanding of externalities. The standard response to the “external effects” of social or economic interactions was either regulation (to prohibit the action that produced the externality) or taxation (to make the action more costly.)63 Both responses clearly rely on centralized authority. Coase aimed to reorient the way we conceptualize the problem of externalities and to introduce an alternative approach grounded in the value of decentralized exchange. His argument formed one of the bases of the “economic analysis of law” movement. Here it is important to note that Coase does not offer an argument about the value of the market per se, even though there are close affinities between his account and standard justifications of the market. His claims are more thoroughgoing in that he argues that decentralized bargaining can produce collectively beneficial outcomes, even without the pressure of market competition, as long as certain minimal conditions about property rights hold.64 As a pure defense of the superiority of laissez-­faire treatment of decentralized bargaining, it is a prototype. A second important thing to note is that Coase’s analysis does not directly address normative questions of fairness or justice. Any such questions that are implicated by the existence of negative externalities are beyond his consideration. Coase adopts the criterion of social efficiency, the maximization of the aggregate value of productive exchange, and seeks to understand how best to institutionalize a framework for voluntary exchange that will achieve this goal. As we argued earlier, the choice of the appropriate criteria for assessing effective institutional performance is a question that must ultimately be addressed by the relevant population itself. We will not weigh in further on this matter 62   He explicitly mentions markets, firms (hierarchies), common law, and state agencies; and, of course, he himself advances the possibility that a regime of decentralized bargaining could coordinate extensive domains of interactions (Coase 1988, 118). 63   For the classic statement of the justification for these policies, see Pigou (1912). 64   Farrell (1987). As McKelvey and Page (1999, 238) put it in their critical assessment: “The Coase Theorem is the competitive model writ large.” Whereas standard arguments for the market recognize that participants must be independent of one another and so choose in a parametric environment, Coase is claiming that decentralized bargaining can generate efficient results even in situations of interdependent choice.

72

Th e A p p e al o f D e c e n tra liz a tion

here. Rather, we take Coase’s criterion and assess how well his analysis holds up on its own terms.65 To see the logic of Coase’s proposal, consider the following example of a negative externality. Two actors, A and B, own adjacent parcels of land in a neighborhood. A builds a factory on his property. He produces goods in the factory that he sells on the market. The production in the factory also produces pollution that regularly blows into B’s property. B owns a house on her property where she and her family live. The pollution from A’s factory seriously affects the family’s daily life and diminishes their enjoyment of their property. The obvious question, from an institutional and/or policy perspective, is what, if anything, should we do about the pollution? Prior to Coase’s analysis, the standard answers to this question were (1) create a rule making A liable for the damages to B caused by the pollution, (2) tax A an amount equal to the damage to B, or (3) establish a zoning rule that would prevent factories that produce pollution from locating in residential neighborhoods.66 Each of the options places the focus of the analysis on A’s behavior. Options (1) and (3) are basically institutional choices to create a rule constraining A’s behavior in significant ways. Coase argued that from a social perspective, this was the wrong way to conceptualize the externality problem and, thus, the broader problem of institutional choice it generates. He insisted, rather, that this was a “problem of a reciprocal nature.” And, thus, the best way to conceptualize the problem was one of mutual potential harm: “To avoid the harm to B would inflict harm on A. The real question that has to be decided is: should A be allowed to harm B or should B be allowed to harm A? The problem is to avoid the more serious harm.”67 Once the issue of externalities is treated in this way, then the criterion of institutional choice changes from one of prohibiting a tortious harm to one of minimizing the social costs of pollution. Adopting this perspective makes the problem a collective one and treats the actions of both parties as part of the cause of the situation. Now, for many, this approach will run counter to our normative intuitions about causal responsibility for harmful behavior. Nonetheless, Coase insists on the view that from a social perspective, we should, in the process of resolving the problem of pollution, be more concerned with the collective effects on the efficiency of production than with our normative intuitions about individual harm. Coase himself recognizes that efficiency is an anemic criterion for assessing any institutional arrangement (1988, 154). 66   Coase (1960, 1–­2). 67   Coase (1960, 2). 65  

73

C h ap te r 3

If we adopt this alternative perspective, he argues, we will achieve greater collective benefits: “If we are to attain an optimum allocation of resources, it is therefore desirable that both parties should take the harmful effect (the nuisance) into account in deciding on their course of action. It is one of the beauties of a smoothly operating pricing system that, as has already been explained, the fall in the value of production due to the harmful effect would be a cost for both parties.”68 Coase expands our conception of the costs of the problem in such cases. In addition to the costs of pollution itself, he highlights the costs of prevention, in terms of the potential loss of production. He concludes, “The economic problem in all cases of harmful effects is how to maximise the value of production.”69 This is precisely where Coase rests his case on the power of mutual exchange. For, according to his argument, the resolution of the problem is simple once it is conceived of in this way. Once the legal rights are clearly established and the parties know what the law allows them to do vis-­à-­vis their preferred activities (in this case, factory production for A and unpolluted family life for B), then they will freely and voluntary bargain about any issue over which they conflict. If the value of production to A exceeds the value of unpolluted enjoyment of property to B, then A will pay B an amount equal to her loss of enjoyment in order to maintain the factory. If, on the other hand, the value to B of no pollution is greater than the value of production to A, then B will pay A not to produce. For Coase it is simply a matter of Pareto-­improving mutual exchange: “It is always possible to modify by transactions on the market the initial legal delimitation of rights. And, of course, if such market transactions are costless, such a rearrangement of rights will always take place if it would lead to an increase in the value of production.”70 Two things are notable about this initial result. First, Coase argues that the outcome of the negotiations between A and B will be socially efficient, regardless of who has the right in the particular case. This is a striking claim, in that it suggests that the outcome of the negotiations will be the same whether A holds the initial right to pollute or B holds the initial right to live in an unpolluted neighborhood.71 On this account, the question of institutional choice is satisfied by the selection and maintenance of clear and commonly understood property rights, with no social welfare implications of which party is granted the substantive right. The Coase logic assumes that the actors are motivated Coase (1960, 13). Coase (1960, 15). 70   Coase (1960, 15). 71   McKelvey and Page (1999) refer to this as Coase’s “invariance claim.” 68   69  

74

Th e A p p e al o f D e c e n tra liz a tion

by their own self-­interest and will be willing to trade the right for money if the exchange increases their net wealth. Second, and more important, this initial claim rests on the key assumption that there are no transaction costs.72 Coase assumes, for the sake of argument, that all of the normal costs of negotiating and enforcing an agreement are set at zero. If transaction costs are zero, then the parties only need to take account of the actual substantive costs of interacting with each other, and as long as they conceive of the problem as one of mutual responsibility, they will bargain to the socially efficient outcome. However, Coase knew, as we all do, that transaction costs always exist and can have a significant effect on bargaining. In the limiting case, if the costs of the transaction exceed the benefits to the parties of the exchange, the parties will rationally reject the bargain. The existence of transaction costs requires Coase to change his focus more explicitly to the question of institutional choice. Coase acknowledges that due to positive transaction costs, A and B might not reach an agreement. The social problem for Coase is not the incompatibility of uses that A and B want to make of their property but the transaction costs that prevent them from achieving a mutually beneficially and socially efficient exchange. The solution for Coase is to create institutional mechanisms (property rights and other rules governing social interactions) that minimize these costs. On Coase’s analysis, then, a decentralized system of property rights accompanied by rules to minimize transaction costs constitutes the best institutional arrangement for resolving problems of externalities, as well as other types of market failure or incompatible activities among citizens. As we previously suggested, Coase’s analysis has had a profound effect on the study of law and institutions. Its appeal is, in part, a function of its simplicity and, in part, of its fundamental importance for justifications of decentralized institutional arrangements. Although Coase formulates his argument with examples of simple two-­person interactions involving a negative externality, the scope of its implications are much more general. He insists that his analysis holds regardless of the type of market failure and regardless of the size of the group.73 He suggests that the appropriate way to think about such problems is a comparative analysis of “how, in practice, the market, firms and

In Coase’s own words, transaction costs arise because “it is necessary to discover who it is one wishes to deal with, to inform people that one wishes to deal and on what terms, to conduct negotiations leading up to a bargain, to draw up a contract, to undertake the inspection needed to make sure that the terms of the contract are being observed, and so on” (1988, 114). He was entirely aware that the assumption of costless transactions is extremely demanding. For a more contemporary discussion of transaction costs, see North (1990a). 73   Coase (1960, 19). 72  

75

C h ap te r 3

governments handle the problem of harmful effects.”74 As for the breadth of his approach, he adds, “It would seem desirable to use a similar approach when dealing with questions of economic policy and to compare the total product yielded by alternative social arrangements. . . . [T]he analysis has been confined, as is usual in this part of economics, to comparisons of the value of production, as measured by the market. But it is, of course, desirable that the choice between different social arrangements for the solution of economic problems should be carried out in broader terms than this and that the total effect of these arrangements in all spheres of life should be taken into account.”75 As we saw in the previous analysis of markets, a crucial question for the process of institutional choice is, “How robust is the claim about the superiority of a particular institutional form?” This is especially relevant in the case of Coasian bargaining because of the breadth of its advocates’ claims. At one level, Coase provides a justification of decentralized bargaining as the preferred institutional arrangement for coordinating a wide range of social interactions. At a second, but related, level, this transaction-­cost account may be seen as a direct solution to the problems we raised above about the robustness of market effectiveness. Coasian bargaining can be interpreted as the solution to failures of competition in market economies. If it is the case that we can negotiate efficient solutions to most of the problems that follow from the failures of competition, then the combination of markets supplemented by Coasian bargaining might be a powerful candidate for a default priority for the question of institutional selection. The problem, of course, is that Coasian bargaining, like claims for markets, is nowhere near as robust as its advocates suggest. Criticisms of the Coase theorem often challenge its normative shortcomings. However, for reasons previously discussed, this is not the approach we want to take here. Rather, for the sake of analysis, we think that it is best to take the account on its own terms and assess how well it stands up when we deviate from the assumptions that form the basis of the theory. This will provide a more telling assessment of how robust we might expect decentralized bargaining to be as a mechanism for coordinating ongoing political and economic interactions. Coasian bargaining does not rely on perfect competition to produce its efficiency results. It relies instead on the power of instrumental rationality and mutual exchange along with a cost-­minimizing institutional framework. The theory draws its persuasive force from its deceptively simple move from the Coase (1960, 18). Coase (1960, 43).

74   75  

76

Th e A p p e al o f D e c e n tra liz a tion

rationality of small-­n exchange to its claims about the efficiency of the aggregate system. The question for the theory becomes, “What coordinates all of the relevant parties to realize and then negotiate all of the available gains from exchange?” Attending to the effects of private information highlights the difficulty of satisfying the requirements of coordination. For our purposes here, we can define private information simply as any information about the preferences or values of an individual participant that is known by that participant but about which the other parties to the interaction remain uninformed. In the presence of private information, participants are required to assess their choices with incomplete information about the opportunities available to them and thus might miss out on potential exchanges otherwise available to them. At the aggregate level, unrealized beneficial exchanges diminish the efficiency of the system. Several studies on the nature of bargaining of the kind envisioned by Coase bear explicitly on the question of the implications of private information. The literature on formal models of noncooperative bargaining offers some general insights into the effects of private information on the efficiency of bargaining.76 The standard conclusion of these studies is that the presence of private information makes bargaining inefficient. In a set of important papers, Farrell applies these results specifically to Coase’s­ claims.77 He sets out the problem for the “Coase theorem” as follows.78 On the one hand, there are conditions under which the strong version of Coase’s claim—­“voluntary negotiation will lead to fully efficient outcomes”—­will hold. Unfortunately for its proponents, the necessary conditions are very restrictive. The strong version holds “when everyone’s tastes and opportunities are common knowledge,” a circumstance in which “people know one another exceptionally well.”79 This is the case when there is no relevant private information For an accessible survey of these results, see Sutton (1986). See, e.g., Farrell (1987); Bolton and Farrell (1990). 78   We resist using the honorific phrase “Coase theorem” because, as McKelvey and Page (1999, 78) make clear, treating Coase’s argument as a theorem is problematic in ways that support our claim that it is not terribly robust. “More specifically, we argue that to obtain the usual conclusions of the Coase theorem (efficiency and invariance to the assignment of property rights), it is necessary to make restrictive assumptions on preferences and use solution concepts that preclude the modeling of transaction costs and incomplete information. On the other hand, if we attempt to relax the assumption of complete information, then the standard conclusions of efficiency and invariance do not hold, even in an environment with no transaction costs. Unlike transaction costs, the inefficiencies and property rights biases due to incomplete information cannot be eliminated by policies designed to make it easier to bargain and negotiate, since these inefficiencies will persist in any game form, even when there are no costs to bargaining.” 79   Farrell (1987, 115). 76   77  

77

C h ap te r 3

in the system. And, as Farrell points out, this is not a very interesting result because it does not distinguish the performance of decentralized bargaining from that of a centralized authority. In the unusual case of complete information, a centralized authority will also be able to achieve efficient coordination.80 When we turn to the more plausible case of a system with private information, inefficiencies unrelated to transaction costs emerge in the Coasian framework.81 The sources of such inefficiency are numerous. Some inefficiencies are the result of the increased time involved in trying to discover the information relevant to complete the exchange. Others involve additional costs that are incurred as the parties try to strategically shift the terms of the bargain in their favor, strategic opportunities that would not be available if all of the information relevant to the bargain were common knowledge. And, perhaps most importantly, private information produces unrealized beneficial exchanges. This last issue epitomizes the problems of coordination in the Coasian framework. As Bolton and Farrell put it, “If agents simply do not know what others are expecting them to do, then coordination problems arise, and these are not well captured by standard notions of externalities and incentives (which describe why agents may knowingly act in a way inconsistent with efficiency).”82 If otherwise willing participants are unable to identify the other actors who represent opportunities for mutually beneficial exchange, then the bargains remain unfilled regardless of the nature of property rights in the system. Farrell argues convincingly that the decentralized institutional framework envisioned by Coase, with its emphasis on the issues of externalities and incentives, focuses too narrowly on the institutional requirements for efficient small-­n exchange. In doing so, it neglects the broader institutional issues of how to coordinate the actors in such a way as to capture all of the gains from exchange and thus to achieve a socially efficient system. Moreover, the problems of coordination only increase as the size of the relevant group increases. The original inefficiency results in bargaining models involve only two actors, the most basic formulation of economic exchange. But it is important to remember that Coase claimed that size was not, in principle, an obstacle to efficient social interactions in his framework. Yet it is clear that size affects the nature of the coordination problem. As Farrell notes, “Looking at things this way suggests that we try to assess the difficulty of coordination and negotiation between the people who must get together to improve Farrell (1987, 116). This conclusion is reinforced by other theoretical and experimental results. See, e.g., McKelvey­and Page (1999; 2000). 82   Bolton and Farrell (1990, 806). 80   81  

78

Th e A p p e al o f D e c e n tra liz a tion

on inefficient outcomes. For instance, if many people must cooperate, it is harder than if only a handful are involved. Getting many people to negotiate successfully about supplying a public good is hard.”83 Having identified a significant challenge to the Coasian framework, Farrell asks whether the coordination problem is, in fact, a good reason for introducing a centralized authority. He suggests that Coase’s argument does not warrant a default preference for laissez-­faire decentralization and that in the spirit of Coase’s own recommendations, what we require instead is a comparative institutional analysis. In his research on these questions, Farrell has undertaken a comparison of sorts, albeit at the level of competing theoretical models. In anticipation of objections from advocates of the decentralization approach, he models the central authority as an uninformed coordinator in a manner consistent with Hayek’s formulation. The results of the comparison are unsurprising, given what we know about the sources of inefficiencies in bargaining: “[T]he welfare comparison is ambiguous: depending on the parameters, the outcome of negotiation may be more or less efficient on average than the bumbling bureaucrat.”84 Decentralization does a better job of information acquisition and use, while the central authority does better in urgent situations when coordination is crucial. The bottom line of the Farrell analysis is that Coase’s framework fails in the crucial task of coordination, undermining normative claims about the broad superiority of decentralized bargaining for achievement of socially efficient outcomes. The comparative analysis reinforces the view—­a view that we share—­that institutional choice is complex and multidimensional. Neither of the institutional alternatives performed in such a way as to justify any default priority status as a first-­order mechanism for coordinating political and economic interactions.

V.  Nonmarket Decentralized Mechanisms II: Community-­based Cooperation The circumstances of politics as we portray them are thoroughly strategic in the sense that they are constituted by irreducible pluralism and the Farrell (1987, 114). Farrell (1987, 124). Note that Farrell retains efficiency as the primary metric for assessing performance. In that sense, his argument parallels the one we advanced earlier in this chapter. He does not question the normative criteria that Coase adopts, he simply shows that the conditions under which Coase’s preferred institutional recommendation operates effectively—­given his own normative criteria—­are quite narrow. 83   84  

79

C h ap te r 3

disagreements attendant on it in conditions of interdependence. In any given instance, not only will members of some population pursue a range of projects animated by diverse interests, commitments, and attachments, but what any of them is able to accomplish will depend in crucial ways on how they expect other members of the population will act. This sort of interdependence makes it highly unlikely that many of the problems we face can be coordinated adequately through market exchange, even exchange that is supplemented by some system of Coasian bargaining. The circumstances of politics thus underscore the unavoidable need for coordinated efforts to, for example, provide public goods, redress or promote externalities, and so forth. Yet strategic interdependence generates well-­known problems of collective action that were, for 85 some time, considered quite insurmountable. Although they reveal themselves in numerous ways, the difficulties posed by strategic interdependence are exemplified by the task of producing goods that 86 are public rather than private, in the technical sense. In the extreme case, the prospects appear quite bleak: when rational actors confront a one-­time “prisoner’s dilemma” (PD)—­or even a repeated PD with a specifiable and commonly known termination point—­each player has a dominant strategy to withhold cooperation. That means that regardless of what she expects others to do, each individual is better off “defecting” by not contributing to the collec87 tive action. This is equilibrium behavior in the game-­theoretic sense, meaning that no individual has any reason to act otherwise given how he anticipates that others will act. The PD, of course, is only an especially extreme case. But while collective action problems often reflect one or another less dire underlying strategic logic, the prospects for voluntary, joint cooperation need not markedly improve. This predicament often is taken as prima facie evidence of the need for centralized political institutions—­“the state”—­and, by extension, for aggregation mechanisms. Not everyone, however, readily concedes this point. This is so even for those who remain suspicious of libertarian enthusiasms for market institutions and schemes of decentralized bargaining. See Hardin (1982); Schofield (1985). Examples of public goods are desirable macroeconomic performance, environmental protection, and the sorts of institutional arrangement (e.g., informal norms as well as formal property rights and enforceable contracts) that are prerequisite to the operation of economic markets. This last point becomes important to our argument in chapter 6. A pure public good has two primary features. It is jointly supplied in the sense that consumption of the good by any individual does not diminish the amount available to others. It is nonexcludable in the sense that once a public good exists, it is impossible to restrict its use solely to those who bear the costs of providing it, hence the risk of free riders. See Knight (1993). 87   Those unfamiliar with the logic of the PD should see Hardin (1982); Taylor (1987). 85   86  

80

Th e A p p e al o f D e c e n tra liz a tion

For present purposes, we focus on one response that seeks to establish the possibility that rational individuals will voluntarily provide public goods in a decentralized manner. It is possible to show that even a population confronting prisoner’s dilemma incentives can converge on cooperative outcomes where each individual, conditional on her expectations about how relevant others will act, rationally 88 contributes to the provision of public goods. Thus, whereas standard analysis of the static PD establishes mutual defection as a unique equilibrium outcome, in specifiable conditions mutual cooperation also emerges as equilibrium behavior. The argument runs roughly as follows: if players confront a repeated PD and if each individual in a relevant population places a large enough value on the returns to future interactions, and if each individual is uncertain about the actual duration of his ongoing interactions with others, then each player will be willing to cooperate so long as those others do so. Cooperation thus takes a conditional form—­I cooperate today in anticipation that others also will, but if those others fail to cooperate today I punish them by defecting in future interactions. The same is true for all other individuals. And this constitutes equilibrium behavior just so long as we all assume that there will be future interactions. Several aspects of this argument are important. The first is uncertainty. If at any point, any player comes to think that the current round of interaction actually is the final one, he rationally will defect for the very reason that there will be no payoff to future interactions and no opportunity for others to punish him. So the possibility of cooperation here depends on a generalized expectation among players that they will face continuing interactions with one another. The time horizon must remain indefinite. Second, each player must 89 value the returns from future interactions at a high enough rate. If I am sufficiently myopic, I will defect in the face of PD incentives even in a repeated setting. Indeed, mutual defection remains an equilibrium outcome in this repeated-­game framework. Cooperation, in other words, is possible but hardly 88   For an accessible argument to this effect, see Taylor 1987. He proves the result for both two-­person and many-­person circumstances. Note here that Taylor adopts strict rational-­choice premises—­the actors in his models are strategically rational and narrowly self-­interested—­in order to assume a burden of argument. If he can show that cooperation is possible under even so stringent a set of assumptions, then it would be easy to do so under more relaxed circumstances. 89   This is a matter of time discounting, which simply is a way of taking into consideration the likelihood that a player will value future payoffs at a rate lower than current payoffs. Another way of thinking about this is in terms of how patient players are. Taylor defines what “high enough” means quite precisely by specifying for different circumstances threshold values for each players’ discount rate in terms of ratios of payoffs to the game.

81

C h ap te r 3

guaranteed. Finally, and following on this last point, there is a further reason why cooperation is not guaranteed. A static PD interaction ensures a unique, dominant-­strategy equilibrium. But in a repeated-­game framework, cooperation is a Nash equilibrium—­it is only rational for an individual to contribute to collective action if she expects others to do so and not otherwise. But game theorists have shown that, in general, any repeated game with discounting in which players sufficiently value the future will generate multiple cooperative 90 equilibria. So while cooperation is possible, that very possibility itself generates a level of indeterminacy—­embodied for players in the difficulty of coordinating on one of the available patterns of cooperative equilibrium behavior—­ that pushes us in the direction of centralized institutions. How is that the case? Some interpret the fact that cooperation can emerge in repeated games as capturing the essential features of “community,” understood as a group characterized by common beliefs, ongoing multifaceted interactions, and generalized reciprocity. In a community so conceived, social relations—­especially relations of power and influence—­remain relatively de91 centralized. And it is in that sense that community might provide some relief from the problems social choice theorists raise regarding aggregation. For a range of circumstances, at least, there apparently is no need to vote. There are several notable features of the analysis we have just described. In the first place, there is something of a mismatch between the analysis we sketch and the interpretation of community it is meant to sustain. Community demands more than specific reciprocity, in which cooperation is sustained by the threat that an aggrieved party herself will directly retaliate against others who defect on her. Instead, it requires more complex patterns of reciprocity—­ what we might call social norms—­that specify how third parties ought to respond when individuals directly involved in an interactions behave noncooperatively. In other words, community-­based cooperation is robust just insofar as any defector must anticipate generalized retaliation. The problem is that the repeated PD analysis as we sketched it sustains only specific reciprocity. So in order to capture the relevant features of community as a decentralized mechanism for providing public goods, we need to establish the possibility of generalized reciprocity. Now it may be the case that by showing how specific reciprocity can emerge under decentralized conditions, advocates of 90   The general result is called the folk theorem for repeated games because it is not the discovery of any identifiable person but simply something everyone knows. See Kreps (1990c, 72–­77; 1990b, 505–­15). Taylor (1987, 78–­81, 103), of course, appreciates the difficulties that the existence of multiple equilibria poses for his argument. 91   See Taylor (1987, 22–­23). Note the claim here is that power is relatively decentralized, not that it does not exist.

82

Th e A p p e al o f D e c e n tra liz a tion

community believe that they have taken a first step toward showing how generalized reciprocity too might emerge in a decentralized manner. And it indeed is possible to show, using much the same analytical apparatus that we described above, how social norms can be sustained as equilibrium outcomes in 92 an underlying repeated PD game. But notice that whereas specific reciprocity is simply a behavioral pattern, generalized reciprocity in the form of social norms itself represents an informal institution—­a set of rules that specifies for a relevant population who can do what to whom, when, and most important, what happens (or at least what should happen) when the rules are violated. Here two observations are important. Calvert captures both of these in the analysis of institutions that we discussed earlier. He thereby highlights why any hope for widespread community-­based cooperation will have limited applicability. These observations go to basic questions regarding the robustness of expectations about the effectiveness of community-­based cooperation. First, informal institutions can be equilibrium outcomes only within certain thresholds on population size and individual valuations of the future. Once those thresholds are exceeded, the population will need to rely on more and more centralized, formal institutions. Second, as in the analysis we sketch above, coordination problems persist here since, on this view, equilibrium institutions are sets of rules sustained as equilibria in a repeated game. As a result, the folk theorem applies. In larger populations, increasingly complex problems of co93 ordination call for the creation of increasingly centralized formal institutions.

VI.  Nonmarket Decentralized Mechanisms III: Technocracy and Incentive-­Compatible Rules Consider now one final way of resolving problems of market failure and, thus, sustaining a justification of a possible first-­order priority for decentralized institutional arrangements. We call this view technocracy. Technocrats approach collective action problems in very nearly the converse way to that adopted by advocates of community. This contrast is important if we hope to grasp the underlying logic of technocratic proposals. It also is important if one is concerned with the tensions that might emerge among the component Calvert (1995a; 1995b). In particular, we can see both how leadership emerges in these conditions and how leadership, in turn, generates social power through successful coordination (Calvert 1992; Hardin 1982). In this way, we see the inevitable pressures that size and coordination place on those who invoke community as a generalized mechanism for coordinating ongoing political and economic interactions. Compare Taylor (1987, 107). 92   93  

83

C h ap te r 3

mechanisms that comprise any institutional arrangement. Community and technocracy, insofar as they embody divergent approaches to the difficulties that strategic interdependence can pose, pull in quite divergent directions. Advocates of community ground the possibility of cooperation in the fact that strategic interdependence typically has a temporal dimension and that, within identifiable parameters, rational actors will appreciate and respond to the contingencies that emerge under such an extended time horizon. They rest their argument on the expectation that once rational actors appreciate the nature of their strategic interdependence, they will see that they no longer have a dominant strategy to defect, that they no longer have a best course of action irrespective of what they expect others to do. Technocratic proposals, by contrast, seek to mute or dampen rather than enhance actors’ appreciation of their strategic interdependence. Those who advocate them propose mechanisms that, ideally, place each individual in a situation where instead of premising her actions on what she expects others to do, she, in fact, will have a dominant strategy to cooperate. Within a decentralized environment, they seek to establish an institutional wedge that will disrupt strategic maneuvering. Technocratic strategies hold out the possibility of designing institutional arrangements that, by mimicking markets, might remedy the problem of market failure but do so in a decentralized, putatively “apolitical” fashion. This is an exacting challenge. And before examining the details of the technocratic strategy, it is important to establish the continuities and differences between it and the range of views that Robert Dahl summarizes under the label 94 “guardianship.” The latter encompasses various views that (1) rely on some criteria of competence to impose and justify substantial barriers to effective popular political participation and (2) insist on the importance of the state as a mechanism for coordinating political affairs and especially for enforcing so95 cial decisions. Thus, schemes of guardianship involve both centralization and more or less permanent political hierarchy based on claims to competence or expertise. But while it clearly risks degenerating into a “rationalization for corrupt, banal, and inept” political arrangements, guardianship in principle is no 96 more prone to do so than is democracy. The sort of technocracy we discuss below differs from guardianship broadly conceived. In the first place, it is technocratic not because it relies on prior expertise but because the institutional mechanisms it proposes are purportedly apolitical. It relies on mediators who Dahl (1989, 52–­62). Hence, guardianship stands opposed to the sort of anarchist politics that Taylor (1987) seeks to ground in the analysis of community (Dahl 1989, 55). 96   Dahl (1989, 52). 94   95  

84

Th e A p p e al o f D e c e n tra liz a tion

may or may not have special prior expertise or competence. Thus, the technocratic schemes we consider are less a response to differential competence than to the strategic problems created by inevitable information asymmetries that 97 exist among members of any population. Second, technocracy need not imply permanent hierarchy. One could imagine, for instance, picking mediators from the relevant population by lot for short fixed periods, after which they would resume their status as undifferentiated members of the population. Finally, those who advocate the technocratic schemes we discuss are staunchly committed to decentralized institutional arrangements. So, like advocates of community-­ based arrangements, albeit for different reasons, they reject the common view that the difficult task of securing voluntary collective action to supply public goods necessarily affords a rationale for the existence of government. That much said, for purposes of clarity, we start by assuming that a population has established political institutions—­a government—­effective enough to provide public goods. In such circumstances, pervasive informational asymmetries subvert the government’s capacity to generate public goods at efficient levels. After showing why this is the case, it will be easier to see why even populations without government would face the same plight. A government intent on supplying some public good can proceed in two ways. First, it can impose a tax earmarked to finance the public good in question. In this case, individual citizens, since they can consume the good regardless of whether they contribute to the cost of providing it, have an incentive to evade taxes by, for instance, misrepresenting their income, wealth, or whatever factor serves as the metric for tax liability. Second, the government might propose a general budget to cover all of its expenditures for public goods. But here individual citizens—­regardless of whether they support or oppose it—­have an incentive to misreport the value that they attach to any specific public good. Indeed, in this circumstance, all citizens, whether or not they favor the public good, have an incentive to dissemble. If a citizen values the public good at less than the per capita cost of providing it, she has an incentive to underreport the value she attaches to it. If a citizen values the public good at more than the per capita cost she has an incentive to exaggerate the value she attaches to it. These are informational problems insofar as individual citizens have private information regarding their own tax liability or their true valuation of the good. 97   One might complain that technocratic schemes rely on expertise among those who design them instead of among the population who will use them. We argue below that such schemes fail for internal reasons. However, our discussion of institutional emergence in chapter 6 will make clear that we hold out little hope that neutral or disinterested experts will actually design and implement technocratic arrangements of the sort we discuss. For the moment, we set such doubts aside.

85

C h ap te r 3

Government often would find it prohibitively expensive to collect full and accurate information on such issues. Moreover, democrats arguably would resist the extensive government scrutiny of citizens’ lives that would be required by efforts to collect complete, correct information. In any case, both sorts of informational asymmetry frustrate the ability of government to efficiently provide public goods. But notice that this predicament plagues any population—­ even one lacking a government—­confronting a decision about whether or not to generate a public good. For as a general matter, private information allows individuals to strategically misrepresent their views by exaggerating the value they attach to any public good whose provision they support or understating the value they attach to any such good whose provision they oppose. The technocratic response to problems of this sort is to ask whether it is possible to design and implement decentralized institutions that could elicit accurate information from individuals and, thereby, enable a population to provide public goods in more efficient ways. Game theorists have produced a considerable body of research on “incentive-­compatible mechanisms,” which establishes 98 just that possibility. Moreover, they demonstrate that under certain conditions, each community member not only has good reason to be truthful but has a dominant strategy to tell the truth. Incentive-­compatible institutional arrangements rely on a disinterested official or mediator who is able to elicit truthful reports from relevant individuals. The mediator determines a cost threshold for a given public good. He then establishes the portion of each individual’s tax liability that would be devoted to pay for the public good. The mediator informs the group of these amounts and proposes that they supply the public good only if the combined value they assign to providing it exceeds the announced cost threshold. The mediator then separately asks each individual to confidentially report the value that she attaches to the public good. Under this sort of arrangement, each individual has an incentive to honestly reveal the value that she attaches to the public good. Based on the information he gathers, the mediator can announce whether or not the group should provide the public good. And because they know that the mediator’s announcement reflects honestly reported information, each individual has sufficient reason to comply with his recommendation. This seems like an extremely promising result. It demonstrates the possibility that a group of rational actors might voluntarily provide efficient levels 98   Miller and Hammond (1994) survey the research on incentive compatibility in an accessible and appropriately critical manner. They offer several simple numerical examples to illustrate claims of the sort we make in the text. For a more technical treatment of mechanism design and incentive compatibility, see Kreps (1990b, 660–­714). For recent discussion, see the Nobel lectures of Maskin (2008), Hurwicz (2008), and Myerson (2008).

86

Th e A p p e al o f D e c e n tra liz a tion

of some public good. This technocratic solution is attractive because, unlike under guardianship arrangements, the mediator in an incentive-­compatible scheme has only information-­gathering capacities and not only lacks but has no need for enforcement powers. Thus, since such institutional mechanisms mimic markets, the group apparently would provide public goods in a voluntary, decentralized, and so benign, neutral, and essentially apolitical man99 ner. This would, in principle, allow any group to coordinate its activities in ways that evade the imposition of a centralized authority. Unfortunately for would-­be technocrats, there are two reasons—­one external, the other internal—­why things are not so easy as this. They force us once again to assess the robustness of the technocratic claims. First, incentive-­compatible mechanisms work best when they establish truthfulness as a strictly dominant strategy for members of the community. An agent has a strictly dominant strategy in case there is some choice that is best for her regardless of what relevant others do. The most familiar example, mentioned earlier, is that players in a single-­play prisoner’s dilemma have a dominant strategy to defect. By contrast, in Nash equilibrium, each player adopts the strategy that is best for herself given what she expects other players to do. So, in the case of incentive-­compatible mechanisms, the finding that players have a dominant strategy to tell the truth is a very strong result. And not all incentive-­compatible schemes generate outcomes that meet this standard. Some are capable “only” of inducing truth telling as a Nash equilib­ 100 rium. And as is often the case, in many such instances, these games generate multiple equilibria and attendant crucial problems of coordination. In such circumstances, communities face decisions regarding which among the available mechanisms to use and which equilibrium outcome to play and why. They therefore confront problems of strategy and justification that, in turn, place politics squarely back on the agenda. 99   Although for reasons we identify below, they rightly consider this aspiration illusory, Miller and Hammond (1994, 9–­10) capture it well: “In effect, the incentive compatible mechanisms can be seen as offering the efficiency of private-­goods markets for troublesome public-­goods decision-­ making. The failure of markets to provide efficient allocation of resources in the presence of public goods has been a powerful argument in favor of the state; but the existence of incentive compatible mechanisms offered a vision of smoothly functioning public goods markets operating without coercion. In these communities politics could be eliminated because individuals would submit voluntary and truthful ‘bids’ resulting in an efficient provision of public goods and fair allocations of tax costs under the neutral competence of the ‘auctioneer.’ The provision of public goods would be subsumed under market-­like institutions and market-­like behavior, and the result would be market-­like efficiency. . . . ‘Politics’ could finally be exorcized in favor of the neutral competence of the Invisible Hand.” 100   See Kreps (1990b, 697–­700).

87

C h ap te r 3

Although aspiring technocrats likely would find this criticism discomfiting, they nonetheless might claim that the prospect for achieving “dominant strat101 egy implementation” holds out more hope than alternative arrangements. Unfortunately, politics emerges, as it were, inside any incentive-­compatible institutional scheme. For while such a scheme can alleviate the strategic problems that information asymmetries generate within a group it exacerbates an analogous problem for the mediator. The same research that establishes the possibility of incentive-­compatible mechanisms also demonstrates that no 102 such mechanism will be budget balancing. There always will be a residual in the form of either a deficit or a surplus. In this circumstance, mediators not only have an obvious incentive to implement a mechanism that guarantees that the unavoidable residual will be a surplus, they always have the opportunity to do so. They then have an incentive to maximize that surplus by overstating the cost threshold of providing particular public goods and ap103 propriating the residual for other purposes. Here the informational asymmetry is reversed. The process of providing public goods and of implementing incentive-­compatible mechanisms to sustain that process inevitably generates specialized information that any mediator can use to mislead other members of the group. Individual members have no incentive to exercise oversight in such circumstances. Nor do they have an incentive to gather and disseminate relevant information because any such oversight or information is itself a public good. Minimally, all of this means that there will be political conflict over how any surplus will be distributed. More ominously, it provides both opportunity and incentive for manipulation. Any individual or group within the community charged with implementing an incentive-­compatible mechanism and any mediator they select to operate it have an incentive to generate and appropriate a surplus even if, in so doing, 104 they generate inefficiencies that diminish social welfare.

VII.  Conclusion The burden of justification for any social institution involves in part a process of reconciliation. Advocates of any institutional arrangement like the market begin with a set of analytical claims about the effective performance of that The quoted phrase is from Kreps (1990b, 697). Miller and Hammond (1994, 14–­15). 103   Miller and Hammond (1994, 16–­19). 104   Miller and Hammond (1994, 15–­21). 101   102  

88

Th e A p p e al o f D e c e n tra liz a tion

institution. This will include claims about the relationship between the conditions under which the institution operates effectively and the positive consequences that it will produce. In the case of the market, these are claims about the relationship between decentralized competition and social efficiency. The degree of difficulty involved in satisfying the burden of justification is a function of the extent of the discrepancy between these analytical claims and the realities on the ground. That is, our explanatory accounts of the actual conditions under which the institution operates in reality will often diverge, and often in substantial ways, from the analytical ideal type. When they do, as we argue they do in the case of the decentralized mechanisms considered in this chapter, then the normative task focuses on the question of robustness. To the extent that an institution can continue to perform effectively as the conditions under which it operates diverge from the optimal, the more persuasive will be the normative claim about the value and appropriateness of the institution. For the market, the question of robustness hinges on how well the market continues to produce socially efficient outcomes as the conditions of decentralized competition wane. Robustness emerged as a significant issue for each of the decentralized mechanisms we considered in this chapter. Whether the problem was market failure, problems (and even outright failures) of coordination, or the intrusion of politics in the form of persistent distributional conflict, each of the institutional forms we examined encounters significant challenges to its capacity to operate effectively. This is true when either the conditions under which it in fact is typically situated or the conditions likely to obtain when it emerges in the first place depart from the conditions that formal analysis reveals as necessary to its effective performance. In each of these cases, there is pressure toward greater centralization of the tasks of coordination and, therefore, of the structure of decision making. However, we should state clearly that all we have shown at this point is that the justifications of a priority for decentralization fail on their own terms. They fail in important ways to satisfy the claims that those who advocate decentralized institutional arrangements make about the effectiveness of those arrangements. As the advocates of decentralization would readily remind us, the mere fact that decentralized mechanisms are insufficiently robust to serve as general mechanisms of social and political coordination and that they often produce normatively unattractive outcomes hardly entails the conclusion that more-­centralized institutional forms afford a preferable alternative. Regardless of the criticisms we level at decentralized mechanisms, the centralized institutions that instantiate democratic decision making have their own heavy burden to bear. It is to this burden that we now turn.

89

This page intentionally left blank

Part TWO

This page intentionally left blank

Chapter 4 The Priority of Democracy and the Burden of Justification

I.  Introduction As an ideal mode of social coordination and governance, democracy may well be “close to nonnegotiable” in contemporary politics, but just what this means remains subject to vigorous and persistent debate.1 Participants in this ongoing debate articulate a wide range of positions. At one extreme are those who rather grudgingly concede that democracy simply is the best of a bad set of alternatives, emphasizing the ways in which the sorts of centralized arrangements it entails inevitably underperform relative to one or another competing ideal, typically an arrangement consisting of decentralized mechanisms. We have had something to say about such views already. At the other extreme are those who contend that democracy is the best way of organizing all aspects of our affairs and that it should therefore enjoy a first-­order priority among institutional alternatives. Such enthusiasts often emphasize the ways in which democratic decision making instantiates some combination of fairness, justice, and the “true” collective will. Most of the rest of us find ourselves somewhere between such extremes. We defend here a pragmatist justification of democracy. It is grounded in a set of claims about the fundamental importance of effective institutions for our ongoing social interactions. It rests on the vital role that democratic decision making plays in achieving and maintaining the effectiveness of those institutions. It attends to the different tasks of political theory—­the analytical, the explanatory, and the normative. In doing so, it highlights the important ways in which normative justifications of democracy and other social institutions presuppose and thus depend upon analytical and causal claims about institutional performance. But our account also recognizes that on their own, those analytical and causal claims cannot justify institutions. Indeed, a distinctive feature of the view we advance is that the analytical and causal approaches we have adopted, in fact, make our normative task especially difficult. We borrow the quoted phrase from Shapiro (2003, 1).

1  

C h ap te r 4

We do not defend our justification of democracy vis-­à-­vis alternative theories of democratic legitimacy in a comprehensive way. We should say a word about this. As pragmatists, we are skeptical of nonconsequentialist justifications that require prior commitment to some specific substantive principle or standard. And, as is common in contemporary debates, we seriously question whether any diverse society could ever achieve the necessary consensus on any such principle.2 At first blush, we are more sympathetic to procedural justifications, but we differ from such approaches to the extent that they fail to take adequate account of the consequences produced by such procedures. We believe that any attempt to justify democracy that does not accord significant consideration to the consequences that it generates misses something fundamental about the role of institutions in social life. Our pragmatism, then, leads us to endorse what we call a tempered or constrained consequentialism.3 We believe that even those who advocate democracy on other grounds might find our justification appealing. To other pragmatists, we suggest that our proposal does a better job of capturing the common features of the available characterizations of a commitment to pragmatism. To consequentialists of other sorts like utilitarians, we recommend our approach as a preferred alternative because of the weaknesses in utilitarianism that we noted in chapter 2. To nonconsequentialists, we offer two reasons why they might consider our approach attractive. First, they might find our approach an acceptable “second-­best” alternative in the face of the moral and political disagreement that inevitably arises from the circumstances of politics as we understand them. Second, they might come to see, through the explication of our argument, that their own approaches presuppose in crucial ways the argument about institutions that we offer here.4 Rawls (1993); Habermas (1996). This view captures the basic observation that social and political institutions generate normatively attractive consequences, when they do, only when they operate within the constraints imposed by specifiable initial conditions. In this sense, like Christiano (2004) and Sen (2003, 280–­81) for instance, we believe that justification requires attention to both procedural and substantive constraints on decision-­making processes as well as to outcomes or consequences it produces. Our defense of democracy, in part, stems from its usefulness as an institutional arrangement for reflexively structuring disagreement over how and in what admixture we might “combine” these factors. 4   Indeed, unless advocates of nonconsequentialist views embrace the sort of “ideal theory” that we discussed in our introduction and so remain expressly unconcerned with matters of implementation, they are under considerable pressure on this score. We will not argue this point here. But to the extent that political theory is not simply a species of applied ethics—­and we clearly do not accept that characterization—­this pressure is unavoidable and unrelenting. 2   3  

94

T h e B u rd e n of Ju stific a tion

The questions raised in these debates about democracy are fundamentally normative, but competing responses to those questions must be assessed in light of available analytical models and explanatory theories. Given our realist account of how social institutions emerge, we confront a heavy burden of justification as we seek to make the case for the priority for democracy. We must persuasively indicate, how, and in what ways, democracy does a better job of coordinating our social interactions than competing institutional forms. And we must make that case while applying the same standards to democracy that we applied to the market and the other decentralized mechanisms. In particular, we acknowledge that democratic institutions will arise, where they do, in much the same manner as other social and political institutions. Democracy, then, stands in need of justification. It is our aim to provide one. This is our task in this chapter and those that follow. We begin, in this chapter and the next, with an argument for the second-­order priority of democracy. We do so by establishing how voting or aggregation is both deeply problematic and essential for democratic arrangements. We rely on analytical theories of social choice to make the first point and argue against one common way of confronting the problems that confront aggregation—­various efforts to decenter voting and elections. In chapter 5, we challenge what is by now the standard interpretation of democratic deliberation. In brief, we argue that efforts to identify underlying mechanisms that render political argument simultaneously effective and normatively attractive are misguided. We advance an alternative account of political argument that keeps those tasks—­explanatory and normative—­distinct. Our argument coalesces around a set of claims regarding the virtues of democracy, namely, that it has important cognitive or conceptual effects, that it is adept at coordinating socially dispersed knowledge, and that it is reflective in an especially crucial way. In chapter 6, we consider some potential challenges to that argument. The most important of these challenges is the contention that other types of centralized institutional arrangements—­specifically, courts and bureaucracies—­ might perform as well as, if not better than, democratic decision making in accomplishing the kinds of coordination for which we think it is uniquely well suited. Finally, in chapters 7 and 8, we directly address the question of robustness and discuss the conditions necessary for the effective performance of democratic decision making. These chapters make clear how demanding our argument for democracy turns out to be. We assert the priority of democracy if the necessary conditions for its effective performance hold. And those conditions are demanding indeed.

95

C h ap te r 4

II.  Democracy We begin with some introductory comments about democracy. Our primary aim is to characterize the institutional mechanisms—­voting and argument—­ that are most relevant to our claims about effective institutional performance. In doing so, it is important that we return to a topic that we addressed in chapter 2, namely the political consequences of pragmatism. Doing so will help us clarify our characterization of democracy. It also will help us avoid certain misconceptions that arise from the way debates about democracy among political theorists have transpired. These relate primarily to our emphasis on formal institutions and to our insistence that any persuasive justification of democracy must take account of both voting and argument. Each of the three features that we have attributed to pragmatism—­ fallibilism, anti-­skepticism, and consequentialism—­is, in fact, philosophical in a rather narrow, traditional sense. As we discussed in chapter 2, contemporary pragmatists disagree widely about whether pragmatism, as a philosophical view, has political implications and, if so, what those implications might be. We reconsider one such claim for present purposes. Robert Westbrook once asserted that as a philosophical program, pragmatism “has no determinate moral and political implications.”5 Whether or not one accepts this claim centers on how one defines “determinate” political implications. Westbrook surely was correct that pragmatists are not committed in advance to any specific moral stance, political action, or social policy.6 He was, we believe, on considerably less solid ground in another respect: pragmatism has important implications for the assessment and justification of political institutions. In order to defend this claim, we rely largely but not exclusively on John Dewey’s The Public and Its Problems as a vehicle for our argument. We have two reasons for doing so. First, Dewey regularly and widely wrote about politics. Yet, contemporary interpreters recognize that The Public and Its Problems is “his only work of formal political philosophy.”7 Second, we hope to reorient Westbrook (1993, 1). In fairness, we acknowledge that he no longer subscribes to that view. See Westbrook (2005, 9). Nevertheless, his formulation of what we earlier called the Posner-­Rorty consensus is usefully specific. For, from a pragmatist viewpoint, much rides on what we mean by consequences. 6   Dewey, for example, insists that in the sort of social inquiry that pragmatists endorse is open ended in the sense that “no particular end is set up in advance so as to shut in the activities of observation, forming of ideas, and application” (Dewey 1920/1948, 146). 7   Westbrook (1991, 300); Ryan (1995, 202). 5  

96

T h e B u rd e n of Ju stific a tion

contemporary depictions of pragmatism. Current defenders of pragmatism often seek to identify it with “cultural criticism.”8 We think that theirs is too diffuse a focus and that pragmatism is impoverished—­in just the sense that we are left with an unduly truncated understanding of its consequences—­if it is considered solely the province of philosophers and literary critics. Dewey’s political thought allows us to make this case. It is fairly easy to trace in broad outline the argument that Dewey makes in The Public and Its Problems.9 He is concerned with the origin and workings of the state. However, he does not begin with the problem—­in his mind misconceived—­of how asocial individuals come together to create political society. He begins instead with the difficulties that men in associations encounter as they seek to regulate the indirect consequences of their ongoing interactions. Dewey first differentiates between the direct and indirect consequences of any interaction. He then suggests that the public “consists of all those who are affected by the indirect consequences of transactions to such an extent that it is deemed necessary to have those consequences systematically cared for.”10 Government consists in officials whose function it is to regulate the indirect consequences of social interaction. The state, in turn, encompasses both the public and the government.11 In this way, not only does the formation of a public give rise to government and the state, but these political agencies, in turn, help the public to articulate its interests and, in so doing, help to give it form.12 Dewey contends that this account of political association makes sense historically, but more importantly, he thinks that it provides a normative criterion with which to assess or evaluate political arrangements. We assess states in terms of how well they articulate and oversee the public interest and by how well they organize the institutions of government.13 In other words, we assess them in terms of how effectively they help us articulate and pursue our interests as we attempt to navigate the natural and social worlds. Democracy, according to Dewey, arises at the point where the public becomes instrumental in selecting the government and holding it accountable.14 The problem is that in large modern states, the public increasingly West (1989, 71). For examples see many of the contributions to Dickstein (1998). We ignore for present purposes both the details of Dewey’s argument and several of its ambiguous and unpersuasive features. On the latter, see especially Westbrook (1991, 305–­6). 10   Dewey (1927, 16). 11   Dewey (1927, 15–­16, 27–­28). 12   Dewey (1927, 33, 67). 13   Dewey (1927, 33, 72). 14   Dewey (1927, 77f). 8   9  

97

C h ap te r 4

is “eclipsed.” Under the pressure of economic complexity, demographic transformation, technological change, and so on, the indirect consequences of social interactions proliferate in both scope and quantity to such an extent that the public cannot effectively identify either itself or its enduring interests.15 The public no longer is grounded in local communities. But neither has it developed political arrangements capable of addressing the exigencies of the modern world. That, according to Dewey, is the problem that the public confronts. In lieu of a remedy, he proposes a form of social inquiry that can, by entering into democratic debate and contestation, contribute to the process by which the public can more readily identify itself, where this involves both coming to recognize common interests and concerns and devising political institutions that can serve to advance or protect them. Two themes that Dewey interweaves throughout The Public and Its Problems are of special significance here. The first is his conception of the main tasks of political theory and social inquiry. The second is his insistence on the importance, for pragmatists, of specifying how democracy can be embodied in actual institutional arrangements. We will take these themes up briefly in turn. Dewey ascribes a very clear role to political theory in a democratic society. Because, on his account, experimentation is central to a pragmatist conception of democracy, political theorists in particular and social scientists in general should undertake their research in the service of social experimentation. It is, he explains, “the business of political philosophy and science . . . to aid in the creation of methods such that experimentation may go on less blindly, less at the mercy of accident, more intelligently, so that men may learn from their errors and profit from their successes.”16 Political theory performs a critical function by assessing the merits of existing social practices. It performs a constructive function by proposing both new methods of social inquiry and new institutional arrangements. In both registers political theorists rely on pragmatist consequentialism to illuminate how individual actions combine to generate social policies and institutional arrangements. “What is needed to direct and make fruitful social inquiry is a method which proceeds on the basis of the interactions of observable acts and their results.”17 On this account, pragmatists must understand both the logic of social interaction at the

Dewey (1927, 125–­27, 137). Dewey (1927, 24). 17   Dewey (1927, 36). 15   16  

98

T h e B u rd e n of Ju stific a tion

individual level and the causal mechanisms by which individual interactions are aggregated into social policies and practices.18 Dewey’s views on the importance of democratic institutions are complex. Consider this passage: We have had the occasion to refer in passing to the distinction between democracy as a social idea and political democracy as a system of government. The two are, of course, connected. The idea remains barren and empty save as it is incarnated in human relationships. Yet in discussion they must be distinguished. The idea of democracy is a wider and fuller idea than can be exemplified in the state even at its best. To be realized it must affect all modes of human association, the family, the school, industry, religion. And even as far as political arrangements are concerned, governmental institutions are but a mechanism for securing to an idea channels of effective operation.19 Dewey here distinguishes between ideal and actual institutions in order to make two important points. First, while he insists both that the democratic ideal does not entail any particular institutional arrangement and that we must not restrict our attention to specifically “political” institutions but instead consider the wide array of social institutions that are related to and work in coordination with specifically political ones, Dewey seems clear, too, that we cannot adequately speak of democracy without considering how it might be instantiated in some set of political institutions. In other words, Dewey is adamant that in our efforts to understand and justify democracy, we must advance some plausible conception of how the “ideal” can be embodied in 20 actual political institutions. Despite their limitations—­at least as they standardly are understood—­rational-­choice approaches to social institutions are well suited to an analysis of these phenomena and especially to an examination of the causal nexus between action and outcomes. On the specific issue of the consequences of democratic institutions, this approach can be a primary tool for analyzing three factors: the incentive effects of social and political institutions, the role of beliefs and expectations in structuring individual action as well as the importance of institutions for shaping those expectations, and the mechanisms by which individual actions are aggregated into social outcomes (e.g., markets, political institutions). For our own understandings of rational choice models, see Johnson (2006; 2010) and Knight (1995; 2009). 19   Dewey (1927, 143). 20   Dewey indeed places great weight on the conception of democracy as a social ideal (Dewey 1939/1993, 240–­45). Contemporary pragmatists often emphasize this aspect of his work (e.g., Bernstein 1986; 2010). As we made clear in chapter 2, Dewey nevertheless was quite clear about the risks involved in focusing exclusively upon ideals at the expense of concern for the difficulties of practical implementation. 18  

99

C h ap te r 4

Second, Dewey also insists that no instance of what we might call “actually existing democracy” exhausts the democratic ideal. This highlights the importance he attributes to critique in the continuing experiments in democratic politics. We object to the common supposition of the foes of existing democratic government that the accusations against it touch the social and moral aspirations and ideas which underlie the political forms. The old saying that the cure for the ills of democracy is more democracy is not apt if it means that the evils may be remedied by introducing more machinery of the same kind as that which already exists, or by refining and perfecting that machinery. But the phrase may also indicate the need of returning to the idea itself, of clarifying and deepening our apprehension of it, and of employing our sense of its meaning to criticize and re-­make its political manifestations.21 Criticism of existing institutions does not undermine democracy as an ideal. But it does affect how we develop and maintain that goal. This last point leads some authors to misconstrue the importance of Dewey’s emphasis on institutions. For example, Margaret Radin conflates two different distinctions.22 While Dewey contrasts democracy as a social idea and as a potential institutional arrangement, Radin interprets this as a contrast between “ideal” democracy and democracy “as-­we-­know-­it.” The two distinctions are not equivalent however. As a result, Radin apparently sees no need to offer an account of how the ideal can be realized through feasible institu23 tions. We think that this misses an important feature of Dewey’s philosophy. Discussions of ideal democracy are constrained by some, obviously contestable, sense of what is and is not possible and this requires that we attend to institutions. As against Radin, we believe that this passage from Dewey reflects a central concern of a pragmatic approach to democracy. Pragmatists must treat political institutions as objects of analysis, and they must orient social inquiry, to a significant extent, around criticism of institutional arrangements and their consequences. This requires that we identify those institutional factors that are most relevant to such an inquiry. Throughout the twentieth century, political theorists have defended democracy by pointing out that—­regardless of its actual flaws or, conversely, what critics consider inflated depictions of democratic Dewey (1927, 143–­44). Radin (1994–­95, 540–­41). 23   This is so even if criteria of feasibility are contestable (Knight and Johnson 1996, 88). 21   22  

100

T h e B u rd e n of Ju stific a tion

ideals—­it affords a way of resolving conflicts politically, that is, without overt 24 resort to force or violence. In a well-­known passage, Dewey grants the insight but suggests that those who endorse democracy in so highly qualified a way miss its actual significance. The ballot is, as often said, a substitute for bullets. But what is more significant is that the counting of heads compels prior recourse to methods of discussion, consultation and persuasion while the essence of appeal to force is to cut short resort to such methods. Majority rule, just as majority rule, is as foolish as its critics charge it with being. But it is never merely majority rule. As a practical politician, Samuel J. Tilden, said a long time ago: “The means by which a majority comes to be a majority is the more important thing”: antecedent debates, modification of views to meet the opinions of minorities, the relative satisfaction given the latter by the fact that it has had a chance and that next time it may be successful in becoming a majority. . . . The essential need, in other words, is the improvement of the methods and conditions of debate, discussion and persuasion. That is the problem of the public.25 Here Dewey identifies the two primary mechanisms underlying institutions of democratic decision making. It is important that democratic theorists not follow Dewey in too readily conceding the foolishness of majority rule as a voting method, a process for “the counting of heads.” Indeed, in what follows, we stress the centrality of majority decision making to democratic politics. And it is equally important that democratic theorists understand why Dewey is correct when he insists that in contemporary democracies, processes of political decision making necessarily rely not just on methods of aggregation or voting but on forms of political argument involving “debate, discussion and 26 persuasion.” This theme runs from as early as Walter Lippmann’s realism in the 1920s through Adam Przeworski’s contemporary “minimalist” interpretation. See Lippmann (1927/1993); Przeworski (1999). For other recent formulations of this view, see Hardin (2009, chap. 3); Posner (2003). This should not be taken as a modest claim, since what minimalists deem the best recommendation for democratic arrangements—­that they sustain the ability regularly to remove politicians from office in a peaceful manner—­is a considerable, if precarious, achievement. 25   Dewey (1927, 207–­8). Here, it is helpful to recall some of the themes that Spitz (1984) pursues in her treatment of majority rule. First, she recognizes that the practice of majority rule emerges from something very much like what we have called the circumstances of politics. Second, like Dewey, she directs attention to the process by which majorities form and the conditions under which those processes operate. (See Spitz 1984, xii–­xiii, 3.) 26   See Dewey (1927, 143f). 24  

101

C h ap te r 4

In this respect, Dewey falls squarely in the lineage of political theorists who analyze and assess “representative government.” Indeed, he flatly insists that “all governments are representative.”27 The task for democrats, on his account, is to implement institutional arrangements that will allow publics to advance their interests by holding representatives accountable. Historically, both voting and argument have been integral to the practice and the justification of 28 representative government. And both are crucial to the ongoing operation and justification of democratic institutions. This is a point that many contemporary democratic theorists overlook or downplay. For even if we grant, with Dewey and many others, that government hardly exhausts the domain of democratic institutions and practices, it surely remains central, indeed unavoidable. And if, as Dewey rightly remarks, asserting democratic accountability over representatives is “the most serious problem of government” it remains a serious problem for putatively nongovernmental democratic practices and 29 institutions as well. We make this point largely to preempt one foreseeable challenge to our focus in this chapter on voting as a central feature of any democratic decision-­ making process. Not only do we reject the possibility that democratic theorists can evade the vicissitudes, both practical and normative, of aggregation, we find it difficult to believe that Dewey would disagree with us on that point. Political theorists of various stripes, of course, commonly depict voting as a necessary component of democratic institutional arrangements in any society, 30 and especially in any large, heterogeneous, complex society. Yet, some advocates of deliberative conceptions of democracy, by contrast, seek to evade this imperative by “decentering” voting and elections as components of demo31 cratic institutional arrangements. To this end, advocates of deliberation propose institutional arrangements that situate deliberative practices in “a multiple, anonymous, heterogeneous network of many publics” that they locate,

Dewey (1927, 76). On this point see, especially, Plotke (1997). See Manin (1994; 1997). 29   The quoted phrase comes from Dewey (1927, 76). It is crucial to recall that, as he subsequently makes clear, democrats must not resort to “magic” in their efforts to resolve this problem. 30   See, e.g., Christiano (1996, 45 n. 17); Dahl (1989, 109); Riker (1982, 1–­8); Weale (1999, 14–­15). 31   John Dryzek, for instance, pursues the theoretical strategy of “downgrading the centrality of voting” to democratic practices and institutions in hopes of deflating the critical relevance of social choice theory. He claims to identify a “logically complete deliberative alternative to the aggregative idea” of democracy. The quoted phrases are from Dryzek (2000, 47, 50). As will become clear, we think this approach is dramatically wrongheaded. 27   28  

102

T h e B u rd e n of Ju stific a tion 32

in turn, in “the associative realm of civil society.” Such proposals draw inspiration from wide-­ranging literatures on social movements, nongovernmental organizations, and emergent forms of transnational governance. And in many ways, such informal, often oppositional, sometimes outright insurgent associational forms are crucial to vibrant democratic politics. We do not deny their importance. We nevertheless think proposals that substitute such associations for formal political institutions remain incomplete in at least two ways. In the first place, efforts to decenter voting pay insufficient attention to the shape and operation of those “encompassing political institutions that organize representation along traditional territorial lines” that provide the unavoidable “backdrop” against which secondary associations operate and from which they derive not only funding (directly or indirectly) but legal standing 33 and power. In other words, secondary associations of various sorts operate within a state-­centered context of formal political institutions, and it is difficult to envision a time when this will cease to be the case. At a minimum, then, a plausible account of deliberation must come to terms with the central role that elections play in this larger political environment. In particular, any such account must specify the interface between deliberative practices and aggregation mechanisms. Ironically, by insisting that we recognize the emergence and importance of alternative associational forms, advocates of democratic deliberation highlight rather than eliminate this task. From our perspective, then, this observation more or less completely derails the sort of decentering strategies under consideration. However, for those who advocate such strategies, the situation actually is even more dire. For, in the second place, advocates of deliberation tend to be overly sanguine about the politics—­and especially the difficulties of effective participation, monitoring, accountability, representation, and so forth—­internal to the associational forms on which they pin their hopes. This perhaps is clearest in the case 34 of informal cultural groups. But it also is true of more organized political groupings. Here we might consider objections to the leadership role that un35 elected clergy play in African-­American politics. It is true, as well, of various 32   Benhabib (1996, 87, 73). Compare the proposals for “associational democracy” and “directly deliberative polyarchy” proposed by Joshua Cohen and his coauthors (Cohen and Rogers 1995; Cohen and Sabel 1997). 33   The quoted phrases come from Cohen and Rogers (1995, 64), who, to their credit, recognize this problem. 34   Johnson (2000; 2002). 35   For pointed objections on this point, see Reed (1986). We are not claiming that clergy should play no political role but rather that insofar as their leadership is a by-­product of their religious authority, it raises questions of accountability and justification.

103

C h ap te r 4

sorts of decentralized local political reforms that rely on the participation of 36 community members. It is true, finally, of nongovernmental forms of transnational or international governance just insofar as such regimes reproduce the decision-­making structures of social and political institutions more gener37 ally. The point should be obvious. Advocates of “associational democracy” (by whatever name) require a clearer argument for how, without recourse to aggregation, democratic accountability might operate within both secondary associations and more formal political institutions. Absent such an argument, their efforts to decenter voting remain unpersuasive. Once we acknowledge, with Dewey, the central importance of both voting and argument for the operation of democratic institutions, we can more persuasively approach the task of constructing a normative justification of democracy. Such a justification will focus on the causal effects of these two mechanisms on the collective outcomes of the democratic decision-­making process. But here we need to distinguish between effects that produce normatively attractive outcomes and those that will not. For it is only the normatively attractive effects that will support such a justification. From a pragmatist perspective, the normatively attractive effects of democratic institutional mechanisms are related in important ways to the effective performance of the institutions as a whole. And, as we have suggested earlier and will argue in what follows, effective performance is in large part a function of the conditions under which the institutions operate. In this way, our pragmatist justification of democracy will be grounded in the interrelationship of consequences and conditions, that is, in a tempered consequentialism.

III.  Democracy’s Burden of Justification There is, despite common practice, no compelling reason to privilege markets in the explanation or normative assessment of institutional arrangements. Yet, the same argument we advance against the priority economic and political theorists commonly accord to markets raises clear and especially vexing difficulties for the task of justifying social institutions generally and democratic institutions in particular. If actually existing social and political institutions 36   See Fung (2001) on the problems of monitoring, accountability, and representation in community-­based reforms in education and policing in Chicago. 37   For a discussion of international regimes, see Young (1999). He endorses a prominent role for nongovernmental forms of governance in international affairs, yet he defines them as institutions in very much the sense that we do and stresses the complexity of internal decision-­making and problems of monitoring and compliance within international regimes of various sorts.

104

T h e B u rd e n of Ju stific a tion

typically are by-­products of strategic interactions characterized by asymmetries of power, they are morally arbitrary. Given this fact, how can we hope to justify any given or proposed institutional arrangement? This inference clearly presents a heavy burden of justification. We see two general strategies that one might adopt in hopes of meeting it. One might argue that while our assessment of markets and other decentralized mechanisms is persuasive, it nonetheless remains possible to identify some alternative institutional form that should receive default priority. This seems to be the strategy of many who argue that democratic decision making is the best way of coordinating the full array of our social interactions. Alternatively, one might reject the idea that any institutional form properly enjoys default priority and embrace more fully and explicitly the plurality of institutional mechanisms. We opt for the latter strategy. In doing so, we acknowledge that the burden of justification, extending to the full array of available institutional forms, is an unavoidable implication of our “realist” explanation of institutional development.38 Moreover, we acknowledge that this account imposes rather tight constraints on the possible ways we might approach our normative task. In particular, it undermines confidence in strategies that seek to derive justification for any institutional arrangement from the purportedly attractive processes by which it emerges. Conversely, it prompts us to focus our justificatory strategy in broadly pragmatist fashion on the consequences we anticipate different institutional forms will produce and the conditions under which we can expect them to do so. Once we acknowledge the plurality of feasible social institutions, the task of persuasively ascribing priority to any institutional form in any particular setting appears quite daunting. This, in turn, raises the difficult task of discerning how, in such circumstances, any heterogeneous constituency might determine which array of institutional forms to use for purposes of coordinating its ongoing interactions across various domains. Such decisions unavoidably will raise difficult political questions to which we return below. Here it is important to reiterate two claims that underscore just how arduous this task actually is. First, as standard textbook accounts of markets lead us to expect, any available institutional form will generate normatively attractive outcomes only under particular (and, in principle, specifiable) conditions. Second, as 38   Stated otherwise, we believe that the explanation we advance regarding institutional emergence—­namely, that it typically is animated by bargaining among asymmetrically situated parties and hence reflects the interests of the relatively advantaged—­does not hold just of objectionable or potentially objectionable institutions. It holds, for example, of universal suffrage as it has emerged in modern democracies (Johnson 2005). Thus, even such normatively attractive institutions cannot be justified in terms of the way they were brought about.

105

C h ap te r 4

the realist implications of our bargaining account suggest, in any given circumstance, these conditions are unlikely to either govern the processes by which institutions actually emerge or characterize the arrangements that those processes generate. The central perplexity of institutional reform is whether any among the available institutional forms provides the sorts of resources necessary to address this acute, systemic discrepancy. Processes of institutional reform, then, highlight the normative task of identifying a decision-­making mechanism capable of allowing relevant constituencies (1) to assess the extent to which, in particular settings, actual conditions approximate those that various institutional forms presuppose if we expect them to generate normatively attractive outcomes and, crucially, (2) to reflexively monitor whether the conditions of their own effective operation actually obtain.39 Democratic institutional arrangements differ from other available political and economic institutional forms precisely to the extent that they are capable of performing these tasks. And it is on just this basis that we rest our case for the priority of democracy. This justification rests on the interrelated effects of voting and argument. In the remainder of the chapter, we address these mechanisms in turn. We postpone discussing how they interact until chapter 5.

IV.  Voting Voting is one—­arguably the central—­mechanism for aggregating the values, views, judgments, or preferences of individuals into a collective decision. Over the last half-­century, a well-­developed research tradition has emerged around the effort to systematically analyze the abstract features of electoral mechanisms. Although this research deploys a variety of analytical methods, it was inspired by findings in what commonly is called social choice theory. It addresses two broad domains: the features and operation of vote-­counting schemes and the comparative performance of electoral systems. And while these domains are analytically distinct, individual studies regularly address 40 issues on both sides of the boundary between them.

39   The actual problem is more complex than this. For example, in any given case, we will be interested not only in whether actual conditions approach those presupposed by various institutional ideals but, if they do not or if they do so only imperfectly, whether and how the degree of approximation might be improved. 40   See Austen-­Smith and Banks (1998; 1999); Balinski and Young (2001); Cox (1997); Dummett (1984; 1997); Johnson (1998); Myerson (1999); Riker (1982); and Sen (2003) for overviews

106

T h e B u rd e n of Ju stific a tion

We concentrate primarily on analyses of vote-­counting schemes. This research typically proceeds at a high level of mathematical abstraction. It generates a set of robust, subtle results that identify systematic difficulties with aggregation processes. More specifically, social choice theory highlights two 41 sorts of problem with voting: instability and ambiguity. Voting is unstable because under certain conditions to which we return below, all known ag42 gregation mechanisms can generate cyclical or intransitive social orderings. And because cycles are a form of electoral indeterminacy, collective decisions 43 are subject to manipulation via strategic voting or agenda control. In any particular case, it therefore is difficult if not impossible to know whether or not a decision is the product of such manipulation. The problem this poses for democrats thus is not that every outcome actually has been manipulated. Rather, it is that all outcomes are susceptible to manipulation, that some in fact result from manipulation, and most important, that we have no reliable 44 basis for differentiating manipulated from unmanipulated outcomes. Voting is ambiguous because, starting from the same initial profile of individual preferences or judgments, different methods of aggregation—­different methods of counting votes—­can yield different, sometimes dramatically dif45 ferent, social decisions. In other words, the social decisions elections generate are, at least in part, artifacts of the process by which votes are counted. Moreover, since each sort of aggregation mechanism violates one or another common criterion of fairness or consistency, no independent, external stan46 dard exists for discriminating between methods of vote counting. In this sense, if we assume that voting is essential to democratic decision making and discussion of this research tradition. For historical antecedents to this tradition, see McLean and Urkin (1997). 41   The terms “ambiguity” and “instability” are from Coleman and Ferejohn (1986). As will become clear, while we do not dispute that social choice theory demonstrates that aggregation mechanisms are systematically unstable and ambiguous, neither do we endorse prominent dire interpretations of those results (e.g., Riker 1982). 42   It is important to recall our earlier discussion of stability from chapter 1. For social choice theorists, stability or its absence is an attribute of collective decisions, of outcomes. It means only that a decision does not generate a cycle. It says nothing per se about the normative status of any such decision. 43   In now classic work, Gibbard (1973) and Satterthwaite (1975) establish that any voting mechanism meeting certain minimal normative criteria (unrestricted domain, nondictatorship, range constraint, and strategy-­proofness) is susceptible to strategic manipulation. 44   Riker (1982, 115–­95, 236–­38). 45   See Riker (1982); Dummett (1984; 1997); and various contributions to the “Symposium: Economics of Voting” (1995). 46   Riker (1982, 99–­113, 234–­36); Dummett (1997).

107

C h ap te r 4

and if we further assume that the aim of such decision making is to discern the popular or collective will, we confront a stark problem—­there is no way to determine which, if any, of the available voting methods most accurately represents the popular or collective will. How should we interpret these general results? Some observers conclude rather despairingly that voting is “meaningless” or “empty.” From this perspective, electoral instability and ambiguity subvert “populist” theories, exemplified by Rousseau, that see the outcomes produced by democratic practices as 47 corresponding to something like a coherent popular will. By contrast, on this account, the instability and ambiguity of voting bolster a “liberal” theory of democracy in which, echoing Schumpeter, elections perform the solely nega48 tive task of disciplining elected officials. From here it is a very short step to one or another realist or minimalist depiction of democracy. These alternatives are much too starkly drawn. Because voting is an essential component of any democratic institutional arrangement in any large, heterogeneous, complex society, the systematic instability and ambiguity that social choice theorists establish raises serious, unavoidable difficulties for some interpretations of democracy. Yet populism and liberalism hardly exhaust the theoretical vantage points from which we might interpret such findings. Indeed, we offer a reading of social choice theory that suggests an obvious, justifiable response to the putative dilemma fabricated by theorists who insist that the only available options are an impossible populism or an unpalatable liberalism. And our reading moves precisely in directions that Dewey suggests. 47   There are at least two noteworthy issues here. First, there is an exegetical question about whether Rousseau is a populist in Riker’s sense. For doubts, see Cohen (1986a, 28–­29). We do not pursue this issue. For if we understand populism minimally as a position that identifies democratic legitimacy by reference to a popular will, it seems that Rousseau is, in this sense at least, a populist. Thus a populist need not be committed to Rousseau’s strong vision of the “general will,” only to—­as we refer to it in the text—­some popular or collective will that serves as the basis for legitimacy. Similarly, we will not take up the question of whether or not Rousseau advocated deliberation. For conflicting views, see Manin (1987) and Cohen (1986b). Second, there is an ambiguity in Riker’s objection to populism. He at times complains that populism is politically suspect because it relies on a notion of the general will that “reinforces the normal arrogance of rulers with a built in justification for tyranny.” But his argument does not depend on such contingent claims. In fact, the results he reports sustain a more modest but still very troubling position. “Populism fails” in this latter sense, “not because it is morally wrong, but merely because it is empty.” (See Riker 1982, 249, 239.) 48   Riker (1982, 238–­46). Note that for liberals, on this interpretation, this need not be a reasoned or rational decision—­an electorate might remove officials in an arbitrary way. Indeed, since for Riker election outcomes remain “meaningless,” there is no other way to interpret this process.

108

T h e B u rd e n of Ju stific a tion

Before proceeding, we need to make three terminological clarifications. First, elections can aim at selecting among candidates or, as in referenda, between alternative policies. When it makes a difference, we refer to candidates or alternatives specifically. When, as is typically the case, it does not, we often refer to “options” and take this term to cover both candidates and policy alternatives. Second, a vote can be interpreted as expressing either an individual’s preferences or tastes or her beliefs or judgments. It will often make a difference which sort of “input” we are discussing, and in such cases we will use the relevant terminology. When it does not make a difference, we often use the term “values” to cover both preferences or tastes and judgments or beliefs. Finally, we regularly have occasion to refer to a general, common, collective, or popular will. We use these terms interchangeably not to refer to any specific philosophical formulation (e.g., Rousseau’s “general will”) but more generically to refer to the notion of a standard of morally or cognitively correct collective decision that exists independently of the political processes used to identify or discover it.

V.  Aggregation and Its Vicissitudes I: Instability Systematic analysis of aggregation mechanisms was inspired in large measure by fundamental results in social choice theory. And here the most important or, depending on one’s perspective, the most notorious findings were produced 49 by Kenneth Arrow in the early 1950s. Arrow’s main argument establishes in quite simple but apparently flawless terms that no institutional procedure for aggregating individual values into social decisions can generate a consistent collective outcome while simultaneously complying with several seemingly innocuous normative constraints. And while subsequent work has refined this impossibility theorem by, for instance, weakening its premises and clarifying 50 its interpretation, it has not altered Arrow’s essential conclusions. It is important to be clear about the thoroughly political, and so normative, nature of this undertaking. Those intent on claiming Arrow as a progenitor of “positive” political theory regularly obscure this point. In the first place, the focus of social choice theory is on the operation and justification Arrow (1951/1963). “It is now clear that the theorem is not based on a trivial combination of unusual requirements” (Johnson 1998, 30). Austen-­Smith and Banks (1999, 25–­54) synthesize the technical results as they now stand. For overviews of many subsequent developments in the field, see the essays included in Sen (2002). 49   50  

109

C h ap te r 4

of institutional mechanisms. Arrow’s results are not tied to extreme rational-­ choice premises regarding selfish motivations or maximizing orientation of individuals. They simply presume that individuals are minimally rational in the sense of being more or less consistent much of the time. Second, recall that Arrow himself explicitly frames his inquiry as an exploration of whether and how the institutional arrangements characteristic of “capitalist democracy” generate coherent social decisions, as compared with two loosely identified alternative arrangements—­premodern traditionalism and authoritarian dictatorship.51 This hardly was a naive project in the immediate post–­WWII period. Nor is it a naive project today. Finally, each of the “apparently reasonable conditions” Arrow imposes on collective decision-­making procedures is normative.52 The standard presentation of his “impossibility theorem” establishes that when the electorate faces three or more options, any aggregation mechanism that complies with the following demands: (i) that social decisions be consistent and stable (collective rationality), (ii) that individual choices not be subject to manipulation (independence of irrelevant alternatives), (iii) that all views have equal access to the decision procedure (universal or unrestricted domain), and (iv) that procedures should be responsive and unbiased with respect to feasible options (Pareto) necessarily violates the further demand 53

(v) that all participants be treated equally (nondictatorship).

This is a sobering conclusion for those who seek to defend democracy and who recognize that elections are an essential component of any democratic institutional arrangement. For if Arrow is correct—­as there is every reason to believe—­there “turns out to be a fundamental trade-­off between what one might loosely call ‘collective rationality’ and the extent to which decision-­ 54 making power is concentrated in the hands of a subset of individuals.” And his conclusion has been rendered even more challenging by subsequent analytical work that, shifting from Arrow’s axiomatic framework to spatial models and ultimately a game-­theoretic approach, establishes that his “impossibility” result can be generalized. Theorists have demonstrated that within a multidiArrow (1951/1963, 1–­2). Arrow (1951/1963, 25, 30). 53   Johnson (1998, 12–­20) provides an accessible presentation of the proof. 54   Austen-­Smith and Banks (1999, 52–­3). See also Riker (1982, 132). 51   52  

110

T h e B u rd e n of Ju stific a tion

mensional policy space, some agenda exists that will lead, via majority vote, from any given alternative to some other alternative, and from there to some further alternative, and so on, eventually returning to the initial point of de55 parture. This instability is so general that the resulting cycles, in principle, encompass the entire policy space. Moreover, since the resulting cycles are dependent on the agenda by which alternatives are considered, the entire process is susceptible to strategic manipulation. This raises clear normative questions regarding the legitimacy of outcomes reached by democratic voting procedures.

VI.  Aggregation and Its Vicissitudes II: Ambiguity Turn now to the second systematic difficulty social choice theorists identify with aggregation mechanisms. We call this ambiguity. It starts from the recognition that assuming the same initial distribution of values and even assuming that voters vote sincerely, different vote-­counting procedures, each with particular attractive normative features, can generate quite different collective 56 decisions. Note, first, that this is not inevitable. We return to this point below. Note, second, that ambiguity need not even pose a problem. One might take the fact that collective decisions are an artifact of some combination of the initial values of individual voters and the method by which those are aggre57 gated as a simple, hardly surprising reminder that institutions are important. But this seemingly innocuous claim actually raises a tangle of explanatory and normative questions. For present purposes, apprehension emerges most clearly with the realization that an electorate might be manipulated by a subset of its members who are able to propose and implement a voting procedure 58 that redounds, or is likely regularly to redound, to their particular interest. 55   We return to the implications of this point at the end of this chapter. The original results were presented by McKelvey (1976) and Schofield (1978). Enelow (1997), Johnson (1998), and Cox and Shepsle (2007) trace the contours of subsequent inquiry in accessible ways. Austen-­ Smith and Banks (1998) argue against drawing any overly strict distinction between social choice approaches and game-­theoretic inquiry. 56   “There are a number of distinct types of voting procedures . . . and it is probable that many decisions arrived at by voting would have gone differently had a different voting procedure been employed” (Dummett 1984, 1). 57   Coleman and Ferejohn (1986). 58   Arrow (1951/1963, 89–­91) ends his original monograph with a warning. He observes that political agents often will prefer some institutional arrangement over available alternatives because they anticipate that over time, it will work to their own distributional advantage. And he warns, in supremely understated tones, that it would be quite naive for us to expect most political agents to

111

C h ap te r 4

But in any case, relevant actors typically lack the information necessary to know whether, in a given instance, the outcome of voting might have turned 59 out otherwise had the counting been done in another way. Further apprehensions arise when one realizes that no unambiguous criteria exist by which one might identify any of the available voting procedures as normatively superior to all others all of the time. Nearly all aggregation mechanisms have some normatively attractive properties, and conversely, nearly all 60 violate one or more seemingly unobjectionable normative criteria. As we suggest below, there are ways to mitigate such apprehensions. But more fundamental problems arise if one interprets voting as revealing a popular or general or collective will that is unique and independent of political processes. For if one does interpret democracy in that fashion and if one also concedes that different aggregation mechanisms generate different decisions from the same initial profile of preferences or judgments, then one is left with a view that verges on incoherence. Voting is problematically ambiguous just to the extent that one lacks confidence that any given procedure actually picks out a 61 unique, independent collective will. 62 There is a great variety of voting procedures. They can be categorized in various ways. We begin, for reasons that will become clear below, by setting aside simple majority rule. There then remain a number of “majoritarian” methods (the Condorcet method and several variations on it), which aim to select a unique winner from among the set of feasible options, and a number of “positional” methods (e.g., Borda count, approval voting), which aim to generate more or less complete rankings of the options on offer. We will not 63 64 explore any of these in detail. The general point is easy enough to state. Majority rule works fairly and generates meaningful decisions so long as voters confront a binary choice. But since the social and political world typically does not present us with such convenient, relatively tractable choices, voters often—­indeed usually—­need to choose between three or more options. set such preferences aside in the politics of institutional choice and design. We agree and think that this problem is central to debates over the emergence and change of social and political institutions more generally. 59   Dummett (1984, 3); Riker (1982, 21). 60   Riker (1982, 65, 111). 61   Riker (1982). As will become clear, we neither interpret democracy in this “populist” manner nor concede that the only alternative is a thin “liberal” interpretation. 62   Levin and Nalebuff (1995) discuss more than thirteen. 63   See Riker (1982); Dummett (1984); and “Symposium: Economics of Voting” (1995) for detailed treatments of individual methods and comparisons between them. 64   In the next paragraph, we summarize Riker (1982, 65–­112).

112

T h e B u rd e n of Ju stific a tion

Majoritarian methods work well in such circumstances so long as there is a “Condorcet winner,” which is some option that can defeat each feasible competitor in a pairwise comparison. Here a further difficulty arises since, predictably enough, in many circumstances such a winner does not exist. And in cases where a winner does not exist, the various majoritarian methods can generate quite different decisions. They do not, in other words, identify a unique or unambiguous winner. Moreover, it is a virtue of majoritarian procedures that while they can identify a winner, by allowing voters to consider more than two alternatives or candidates, they also generate information about the way voters rank options that finish in second or third or whatever place. Advocates of different positional vote-­counting schemes insist that such information is especially relevant since majoritarian approaches cannot iden65 tify a unique winner. The problem then is not only that in any given case, it often will make a difference to the collective decision which category of mechanism—­majoritarian or positional—­is used to count votes but that, as is the case with the various majoritarian procedures, it also will make a difference which positional method one employs. Ambiguity, it seems, is unavoidable. This would seem significantly less troubling if one among the available voting procedures is clearly more attractive normatively than others. As we already have mentioned, however, this is not the case for any method under all circumstances. Partha Dasgupta and Eric Maskin, however, suggest that in important respects, simple majority rule—­a voting procedure that declares the option receiving 50% + 1 of the votes cast the winner—­is the most promising candidate here. They argue that in a quite specific sense, simple majority rule is more “robust” than any other voting 66 procedure. In the first place, simple majority rule is “reasonable” in that it meets several minimal, attractive criteria—­it treats voters equally (anonymity), it treats options equally (neutrality), and it is responsive to constituency preferences in the sense that the more votes an alternative or candidate receives, the more likely it is to prevail (Pareto). And although Arrow shows that no rule meeting those criteria will generate transitive social decisions over a universal domain of preferences, simple majority rule performs better than compet67 ing, reasonable procedures in a precise sense. If, over any domain of initial This, of course, raises the obvious additional difficulty of deciding when and for what purposes such additional information is relevant. 66   See Dasgupta and Maskin (1999, 72–­74; 2008). 67   Anonymity implies nondictatorship, and any neutral voting rule satisfies independence of irrelevant alternatives (Dasgupta and Maskin 2008, 2–­3). It follows that since simple majority rule complies with the Pareto criterion, it cannot generate transitive collective decisions over an unrestricted domain. 65  

113

C h ap te r 4

preferences, some voting rule both is “reasonable” in the above sense and generates a transitive collective decision, then simple majority rule does so as well. Moreover, for some domains of initial preferences, simple majority rule alone, among available “reasonable” procedures, generates a transitive collective decision. In other words, simple majority rule is impervious, and uniquely so, to intransitive collective decisions over a significantly more expansive domain of initial preferences than any other reasonable vote-­counting scheme. This is a quite remarkable finding. And it seemingly should give some confidence to democratic theorists who wish to avoid the problems posed by the plurality of voting rules. Indeed, we think that it offers considerable relief in that regard. For this argument persuasively establishes that simple majority rule is the best of the available voting mechanisms in the sense that it not only is reasonable but demonstrably avoids instability more effectively than any of the going alternative procedures. It is crucial to note, however, that the argument that simple majority rule is more “robust” than other reasonable voting procedures is importantly incomplete. In the first place, robustness is a largely procedural characteristic and does not directly address the sense in which voting procedures might be ambiguous with respect to some independent, unique collective will. There is, in other words, no guarantee that the outcome generated by simple majority rule corresponds to the “general will.” Although we do not consider this objection fatal—­since it presupposes that democratic arrangements must in some sense be justified by reference to a general or collective or popular will—­we set it aside for now. Second, those who advocate simple majority rule because it is robust relative to other voting mechanisms must recall that their preferred rule operates most attractively in situations where voters confront a binary choice. As we also noted earlier, the social and political world is seldom so amenable as to present us with neat binary choices. Voters must therefore be concerned with how the choices they face are fashioned in processes of political interaction. And this simply 68 pushes the burden of justification back one step. Finally, the whole argument from “robustness” quite transparently operates across various restricted domains of initial preferences that already have somehow been shaped. And “somehow” is crucial here, too, for explanatory questions lurk at one remove.69 Riker (1982, 58–­64). This point seems lost, for instance, on Mackie (2003), who repeatedly asserts that such restrictions reflect “natural similarity” (stress added) among the preferences of a given population. Riker (1982, 122, 128) rightly recognizes that any such “fundamental similarities of judgment” probably emerge via complicated causal processes, including “discussion, debate, civic education and political socialization.” In other words, homogeneous preferences are a product of social and political processes. They are not “natural.” We would add that nothing in any of the causal 68   69  

114

T h e B u rd e n of Ju stific a tion

We rightly will ask how preferences have come to be constrained in the ways that they have. Ultimately we seek to short-­circuit this potential regress by advancing a pragmatist interpretation of democracy—­one that stresses, as Dewey does in the passage we invoked at the outset, the role of political argument and ultimately of deliberative processes in democratic politics but that, like Dewey too, rejects “causal authorship” views that justify democratic decision making 70 by reference to a general or collective will. Before turning specifically to the way in which political argument might alleviate the problems identified by social choice theorists, we need to do two things. First, we need to address a potential concern regarding our defense of majority rule. We take up that task in the next section. Second, we must offer a more general sketch of how argument enters into democratic decision making. We turn to that matter in the subsequent section. Only once we have addressed these two matters can we return, in chapter 5, to our exploration of the central role of argument in democratic politics.

VII.  Arguing for Majority Rule: Can That Be All There Is? It may be helpful at this juncture to pause and entertain a possible objection. Some readers may find our case for majority rule, insofar as it rests so heavily on the Dasgupta and Maskin analysis, to be uninspiring. The question such readers might press is whether there is a more inspiring case—­by which we imagine such critics would mean a more “principled” one—­to be had. In a series of recent papers, Risse explores just that question.71 Ultimately, his argument for majority rule converges with the position Dasgupta and Maskin advocate. Two features of Risse’s exploratory argument are especially helpful. In the first place, he is attuned to the crucial comparative question—­majority rule as compared to what? Second, he grasps the importance of keeping an eye on the conditions under which particular decision-­making rules operate in normatively attractive ways. With those factors in mind, he concludes that, mechanisms that Riker mentions (or any others, for that matter) suffices, absent additional argument, to ensure that the resulting pattern is normatively attractive. We elaborate this in what follows. 70   Dewey (1927, 53–­55). 71   Risse (2004; 2009a; 2009b).

115

C h ap te r 4

in significant ways, “the case for majoritarian decision-­making is plainly pragmatic.”72 He is not, we should be clear, offering a “pragmatist” defense of majority rule; he simply deflates the sorts of aspiration that our imagined reader articulates. After surveying various objections to majority decision—­especially criticisms leveled by advocates of the sorts of positional voting schemes we mentioned earlier—­Risse suggests that notwithstanding those objections, there are strong arguments for majoritarian decision-­making whenever it is reasonable: (a) to use merely information about the relative standing of any two options in individual rankings; (b) to restrict the voting to pairs of options; and (c) to use aggregation methods at the exclusion of other decision procedures, in particular fair division. Once (a), (b), and (c) are granted, the standard arguments for majority rule can do their work. So, justifying the use of majority rule means identifying the reason for or circumstances under which they ought to be imposed.73 The “standard arguments” in favor of majority rule, on Risse’s account, include the following: it maximizes the number of individuals who exercise self-­determination; it respects individual judgment under circumstances of disagreement; it comports with the conditions of the Condorcet jury theorem in instances where one can presume there is a ‘correct’ outcome; and it comports with what commonly is known as May’s theorem.74 Risse points out that each of these arguments is unpersuasive insofar as it also applies to other decision processes, and so is indeterminate, or it dismisses competing possibilities (say, making decisions via some process of “fair division”) without justification, and so is question-­begging. Risse (2009a, 794). Risse (2009a, 797). 74   May (1952). Dasgupta and Maskin’s (2008) argument differs from May’s better-­known analysis. As they point out: “For the case of two alternatives, May shows that majority rule is the unique voting rule satisfying a weak version of decisiveness, anonymity, neutrality, and a fourth property, positive responsiveness. Our Theorem 3 strengthens decisiveness, omits positive responsiveness, and imposes Pareto and IIA to obtain a different characterization” (pp. 953–­54). This difference is important because Goodin and List (2006) have generalized May’s theorem to show that in circumstances where a decision-­making system collects a single ballot from each voter, “plurality rule, which always chooses the option with the most votes, is the uniquely compelling aggregation procedure; it is so, in the sense that it uniquely satisfies May’s well-­known minimal conditions on a democratic procedure generalized to decisions over any number of options” (p. 940). They contend that in other informational environments (e.g., where voters cast multiple ballots), other voting mechanisms will satisfy May’s conditions. In other words, nothing in Goodin and List undermines the argument Dasgupta and Maskin advance. 72   73  

116

T h e B u rd e n of Ju stific a tion

Having reached that conclusion, Risse remains unpersuaded by the standard criticisms of majority rule. We will not reiterate his response to critics of majority rule other than to indicate that in his analysis of competing voting methods, Risse rightly notes that critics who advocate using this or that mechanism because it generates greater information about the views of voters are under some burden to justify their confidence that the increased information is reliable.75 And he rightly suggests that problems surrounding interpersonal comparisons and concerns about strategic voting should, at a minimum, give us pause on that score. The “principled” arguments for and against majority rule, then, seem interminable because no party is able to persuasively defend her own preferred mechanism or definitively dismiss all alternative voting methods. Risse proceeds to argue that perhaps the best reason to restrict ourselves to consideration of pairwise comparisons via simple majority rule is solely pragmatic. It reduces complexity. In addressing the question of why we ought to rely on aggregation (voting) methods generally instead of some principle of fair division, Risse is similarly persuasive.76 Given the sorts of persistent disagreement that characterize what we call the circumstances of politics, the notion of “fair” is itself highly contestable. There is scant hope, in other words, that the relevant population will reach sufficient agreement on the matter of what constitutes a fair procedure. In that case, we require what Risse terms a “default group decision method.” Given the argument for majority rule he advances, Risse plausibly suggests that it has a claim to play that role. His conclusion, like the one he reached with respect to alternative aggregation mechanisms above, is that “the best case for majority rule is much more pragmatic and less principled than one might have thought.”77 In short, although he takes a different route, in his quest to justify majority rule Risse ends up in much the same vicinity as Dasgupta and Maskin. In this sense, we find his conclusions congenial. They allow us to look more confidently beyond the preoccupation with voting toward the crucial role of argument in democratic politics.

Clearly, there are numerous alternatives to majority rule as a decision-­making mechanism. Sometimes, for certain purposes, we might want to rely on one or the other of those alternative mechanisms (Riker 1982). Indeed, for certain purposes we might—­and indeed already do—­ sometimes opt for minority-­rule mechanisms (Vermeule 2007). But in any given instance, we would want to justify any decision to opt for such alternatives (Holmes 1988). And for such purposes, majority rule itself arguably serves as the default decision rule. 76   Young (1994) provides a useful discussion of fair division procedures. 77   Risse (2009a, 799). 75  

117

C h ap te r 4

VIII.  Argument Political theorists now conventionally distinguish aggregative mechanisms and democratic deliberation. Indeed, they often inflate that basic analytical distinction into an unbridgeable dualism or dichotomy between competing 78 “aggregative” and “deliberative” conceptions of democracy. Insofar as any plausible conception of democracy will incorporate both voting and arguing, we find such broad contrasts misleading. Ultimately, we identify a systematic connection between these two essential features of democratic decision making. We risk confusion, however, if we embark on such analysis too hastily. Not just any sort of political talk counts as political argument, and not just any sort of political argument or debate counts as democratic deliberation. In the remainder of this chapter, we take up the first claim; in the next chapter, we take up the second. Before proceeding, however, it will be helpful to articulate our general understanding of the distinction between argument or debate and deliberation. The bulk of the following discussion is animated by the need to vindicate our views on how the two differ. In our discussion of markets, we defended the view that not just any pattern of exchange of goods or services constitutes a market, with the attractive normative features commonly attributed to the latter. Only when exchange occurs under something approximating “textbook” conditions can we infer from the persistence of exchange to the presence of markets. Likewise, the literature on social choice that we have just discussed largely reflects the fact that not just any casting and counting of votes counts as an election. Indeed, Arrow’s famous results emerge precisely because he recognizes that any aggregation of votes must take place under some set of normative conditions in order to qualify as an election. Similarly, not just any process by which participants argue or debate by advancing and entertaining utterances constitutes democratic deliberation. We can infer from arguing to deliberating only when the former takes places under specific conditions of freedom and equality. Unsurprisingly, the “initial conditions” to which we refer in each instance are very nearly the same. The nature of our argument hopefully is clear. In each case, we place normative weight less on the underlying mechanism (exchange, voting, or arguing, respectively) and more on the conditions under which those mechanisms operate. Just as economic exchange approximates market interaction as standard microeconomic See, e.g., Benhabib (1996); Cohen (1996, 98–­99); Young (2000, 18–­26); Dryzek (2000, 50–­ 51); and Shapiro (2002a). Recently, Mansbridge et al. (2010) have backed away from nearly all of the specific positions that sustain the aggregative-­deliberative dichotomy. In the process, they embrace many of the positions we have defended in numerous earlier papers. 78  

118

T h e B u rd e n of Ju stific a tion

models depict it only under stringent conditions, so only when they transpire under particular conditions of freedom and equality do voting and arguing respectively resemble elections and deliberation as these appear in debates among democratic theorists.79 We start with the concept of political argument, by which we mean concerted efforts to coordinate ongoing social and political interactions via the exchange of reasons under conditions of potential, and often of actual, disagreement and conflict. We differentiate political argument so understood from conversation, which consists of informal talk that is not aimed at reaching decisions or, for that matter, at any sort of end. We thus would place the sort of “everyday talk” that Mansbridge promotes into the latter category just insofar as it “is not aimed at any action other than talk itself.”80 Likewise, we would categorize the sort of “political discussion” that Conover and others explore as conversation insofar as it remains largely “spontaneous, unstructured and without clear goals.”81 Such conversation often is crucially important in practical terms to the extent that individuals rely on it as a source of information or of exposure to more or less divergent views. But on our conception, conversation of this sort is best understood as preliminary or supplemental to political 82 argument and deliberation. Political theorists commonly draw a further distinction between political argument and bargaining, in which relevant parties seek their own, typically material, advantage through the exchange of more or less credible threats and 83 promises. We, too, make this distinction but do so in an unfamiliar way. Rather than see arguing and bargaining as distinct categories, we see bargaining as one mode parties might adopt when they engage in political argument. In this sense, we consider the arguing-­bargaining distinction to be significantly less important than do other democratic theorists.

79   Given our approach, much of the social scientific literature that has emerged in recent years in the effort to test the claims made for democratic deliberation is quite problematic. In part, that is the thrust of our argument in the remainder of this chapter. For critical reviews of this empirical literature, see Delli Carpini, Cook, and Jacobs (2004); Ryfe (2005); Thompson (2008); and Mutz (2008). 80   Mansbridge (1999a). 81   Conover, Searing, and Crewe (2002). 82   Indeed, we suspect that the reticence toward public discussion of politics expressed by the respondents in the Conover, Searing, and Crewe (2002) study reflects shortcomings in extant institutional arrangements and especially social norms, namely, the absence of conditions of freedom and equality under which political argument would approximate democratic deliberation. 83   Elster (1995) lays out the basic distinction between arguing and bargaining. For a recent treatise on the latter, see Doron and Sened (2001).

119

C h ap te r 4

We adopt this view largely because we find the two criteria on which political theorists commonly rely when differentiating arguing and bargaining especially misleading. In the first place, the distinction does not ride, as is often supposed, on whether or not the parties involved act strategically. For an utterance that in a bargaining setting might be construed as a threat could, in other types of argumentative setting, constitute a warning (and so a reason for doing something or not) even if a speaker might issue the warning for the 84 strategic purpose of advancing her argument. Conversely, in analyzing other forms of political argument, we typically are concerned with how and why 85 actors “do things with principles.” Since actors use principles to establish similarity or dissimilarity among cases, and since such determinations rarely will operate impartially, there is every reason to suspect that in politics, actors will avail themselves of opportunities to do strategic things with principles. It will, in other words, often be to my advantage if I can induce you to see some new event or phenomenon as “like this”or as “unlike that.” Second, one cannot reliably distinguish bargaining from other sorts of argument in terms of the sorts of claims that parties make. Most importantly, we think it misguided to categorically exclude particularistic, even self-­interested, claims—­or the conflicts that such claims might generate—­from the range of admissible topics that participants in political argument generally or demo86 cratic deliberation more specifically might address. Members of previously excluded groups, for example, typically demand entry into relevant decision-­ making arenas precisely because, so long as they remain excluded, their in87 terests are not adequately considered. In this sense, relatively particularistic interests provide the scaffolding for principled claims to representation, equal88 ity, and rights.

84   On this topic, see Elster (1995). One might consider as well the “rhetoric of reaction” that Albert Hirschman (1991) explores, which consists largely of warnings of various sorts regarding the baleful consequences that will attend the adoption of this or that proposed reform. 85   We draw this phrase and the insight in the next sentence from Nozick (1993). His argument converges in interesting ways with the discussion of principles, tradition, and precedent in Kreps (1990a), Schelling (1960), and Dworkin (1988), respectively. In each instance, the particular sort of speech act is used to, in Nozick’s words, sustain “subjective inferences,” typically in the face of what economists now refer to as “unforeseen contingencies.” On the latter, see Dekel et al. (1998). 86   More generally, we resist the notion that self-­interested and moral claims are necessarily mutually exclusive and opposed. On this point, see Nagel (1989, 909, 914). 87   See, e.g., Phillips (1995). 88   On the conceptual relation of interests to rights and representation, see Waldron (1993, 11–­ 12) and Pitkin (1967, 113, 155), respectively.

120

T h e B u rd e n of Ju stific a tion

Having now dismissed what arguably are the most common ways of differentiating between bargaining and other forms of argument, it remains for us to identify the grounds on which we do so and to sketch the reasons for proceeding as we do. The source of confusion here lies in the propensity of political theorists to collapse their distinct explanatory and normative tasks. In particular, they tend to overburden explanatory mechanisms with normative 89 expectations. Our central analytical task is to identify the sorts of mechanisms that coordinate social and political interaction. We differentiate bargaining from other forms of political argument quite simply in terms of the sorts of mechanisms that each invokes. When parties to political argument issue threats the credibility of which is grounded wholly or predominantly on exogenous social, economic, or political resources, they are engaged in bargaining. But on our view, the linguistic and other resources available to participants in political argument always are both deeply entangled and asymmetrically distributed. Bargaining, then, simply represents the limiting case of argument in which parties rely most heavily and directly on nonlinguistic resources to coordinate their ongoing interactions. This analytical distinction carries no direct normative significance. In particular, we do not think that it sustains any plausible judgment about what might properly count as a “reason” in political argument. When in the course of argument I issue a threat, I am offering a reason of a specific sort. I am announcing that should others act contrary to my stated view, I will, in response, act in ways that run counter to their interests. To the extent that my announcement is credible, others then have reason to comply with my demand because it is in their interest to do so. We see no basis for excluding this sort of reason from the domain of political argument in a blanket way. For while certain kinds of threat may be suspect in normative terms, not only will this be untrue of many others, but it often will be the case that normatively unobjectionable In this sense, our view differs from the one Cohen and Rogers (1995, 247) advance. They argue that interest-­based claims merely afford reliable “information” about relevant interests but do not in themselves constitute a reason for adopting or rejecting a policy. On their view, any such reason must be more general—­e.g., that the policy “represents a fair accommodation of the interests of all, or advances the good of those who are in greatest need.” On our view, it is rarely possible for parties to political argument to even articulate claims without direct, if tacit, reference to their own interests, and for parties making demands (e.g., for inclusion) such claims in fact constitute reasons even if those reasons might not be decisive. 89   This tendency is symptomatic of the tacit, and we think unhelpful, assumption that political theory is best understood as a subsidiary of ethics or moral theory. As a consequence, normative attention is focused unduly on individual actions and dispositions rather than on institutional factors.

121

C h ap te r 4

threats will effectively induce cooperation of precisely the sort required to coordinate ongoing social and political interactions. Here, finally, we return to the distinction between political argument and the more restrictive category of democratic deliberation. On our account, political argument consists in the concerted effort by participants to coordinate their ongoing social and political interactions through the exchange of reasons under conditions of both interdependence and potential, indeed often of real, disagreement and conflict. Actual political argument approximates deliberation—­and so acquires normative force in addition to practical effect—­to the extent that it occurs under specifiable substantive and procedural 90 conditions that embody constraints of freedom and equality. It is just these initial conditions that afford the leverage needed to differentiate the kinds of threats (or other utterances) that are normatively admissible to deliberative forums from those that are not. They do so by providing general criteria for differentiating the sorts of resources that parties to political argument can properly bring to bear on their interlocutors in deliberative settings. Under what circumstances, by our lights, would threats be admissible in democratic deliberation? Clearly not in instances where one party bases the credibility of his threat on access to asymmetrically distributed social, economic, or political resources. Hence, threats made by a boss to his employees—­“if you do not support my preferred policy, I will fire you from your job”—­or by the politically well-­situated to those less advantaged—­“if you back a candidate or policy that I oppose, in the future I will use my political position (e.g., seniority, or committee chairmanship) to prevent issues of concern to you from even reaching the agenda”—­would be proscribed. By contrast, threats of the sort “if you do not support my preferred candidate or policy in the decision-­making process this time, I will oppose your preferred polices and candidates in the future even if I myself prefer them to the other alternatives on offer” would be perfectly unobjectionable because their credibility need not derive from economic or political asymmetries.91 Both sorts of threat 90   This claim is not controversial in itself. For instance, Cohen (1996, 99) writes that “democracy, on the deliberative view, is a framework of social and institutional conditions that facilitates free discussion among equal citizens.” Likewise, Dryzek (2000, 38) insists that “close attention needs to be paid to the conditions under which authentic deliberation can proceed.” That said, the particular way we spell out, in chapters 7 and 8, the conditions of freedom and equality that constitute democratic deliberation and the way we see them entering into the task of political justification might prove somewhat more controversial. For in that discussion, we justify these constraints primarily in terms of their consequences for the effectiveness of democratic institutions. 91   Note that both sorts of threat are costly to carry out ex post and so might not be credible ex ante. That is a different matter.

122

T h e B u rd e n of Ju stific a tion

might well have the practical effect of inducing the intended audience—­by presenting them with reasons—­to comply with the speaker’s wishes and so provide a basis for coordinating social and political interaction. However, threats of the first sort—­since they import economic asymmetries into politics or trade upon asymmetries within the political process itself—­would violate the conditions that democratic deliberation, as distinct from simple political argument, demands. Threats of the second sort are admissible in deliberative arenas just insofar as they do not violate those initial conditions of freedom 92 and equality. As a general matter then, while the processes of “debate, discussion and persuasion” characteristic of political argument can have significant practical effects, those effects need not, simply by virtue of their “communicative” character, be normatively justified or even justifiable. The sheer fact that a decision emerges from political argument no more confers normative standing on that decision than if the same decision had been reached simply via voting. Since democratic deliberation, on our account, consists of political argument under a set of specifiable conditions, collective decisions reached wholly or partly through political argument derive normative force largely to the extent that “discussion, debate and persuasion” transpire under circumstances that approximate those conditions. In this respect, our approach is analogous to that of social choice theorists. Concern about the normative standing of various voting methods lead them to focus on the outcomes that aggregation devices generate and on the conditions under which they operate. Likewise, we suggest that any plausible normative defense of deliberative procedures must focus less on mechanisms that animate parties to political argument than on the conditions under which argument transpires.93 The difference between the second of our inadmissible threats and the threat we deem unobjectionable may be unclear. Think of the former as an instance of agenda manipulation and the latter as one of strategic voting. Only some participants have effective agenda control while nearly everyone can vote strategically. So the ability to credibly issue the former threat trades on political resources that are not generally available, while the ability to credibly make the latter threat is, in principle, available to anyone willing to bear the cost of carrying it out. It seems to us that some of the procedural reforms (e.g., rotating committee chairs) that Guinier (1995), for instance, advocates are aimed precisely at leveling asymmetries of the agenda-­setting variety and so set the stage for more deliberative legislative interactions. Similarly, we see the preconditions (both procedural and substantive) for democratic deliberation as working to mitigate the political relevance of resources that are distributed in asymmetrical and typically arbitrary ways. 93   We take up the themes of this paragraph at considerable length in chapter 5. Failure to grasp the points we make here has consequences for current debates among political theorists. It animates, for instance, the disagreement between Steiner (2008) on the one hand and Austen-­Smith and Fedderson (2008), Goodin (2008), and Schneider (2008) on the other. 92  

123

C h ap te r 4

Therefore, we hope to reorient analysis of political argument in several ways. In general, we challenge the tendency among political theorists to overburden the mechanisms that animate political argument by placing normative as well as explanatory demands upon them. To that end, we aim to deflate explanatory and normative expectations regarding the consequences of political argument, especially expectations regarding the type and extent of agreement that it is likely to generate. For neither political argument nor deliberation ensures or, for that matter, necessarily aims at agreement in any unambiguous sense. This leads us to argue for a pronounced shift in theoretical attention away from claims about how argument induces substantive convergence among participants and toward an analytical focus on how it shapes the broadly cognitive structure of their views. And it leads us to underscore the fact that argumentative processes and the outcomes they do induce are, just on their own, normatively arbitrary. The practical effects that they induce will be normatively attractive only when argument proceeds under conditions that approximate those required for democratic deliberation. Before proceeding, we consider two examples of the sorts of confusion that arise when theorists fail to differentiate adequately between political argument and democratic deliberation. Consider first a hypothetical example, namely the concern about “enclave deliberation,” understood as consisting of political argument among some group whose members subscribe to homogenous, relatively extreme views and who, deliberately or otherwise, radically limit 94 the “argument pool” to which their views are subject. Predictably, argument under such conditions can generate “group polarization.” On this view, such polarization, which involves the propensity of members to embrace ever more extreme versions of their “predeliberation” views, risks degenerating into “in95 creasing extremism, fragmentation, hostility, and even violence.” It indeed is possible that political argument might have just such effects. And insofar as this phenomenon forces us to question whether the mechanisms that animate political argument always or usually work in normatively attractive ways, they 96 are salutary. But the exclusionary conditions under which group polarization would likely occur are so extreme and so distant, not only from those that would constitute democratic deliberation but arguably from those we might regularly encounter in politics, that it is hard to ascertain with any confidence 94   Sunstein (2000; 2001; 2002a; 2002b) has advanced this scenario widely. For an earlier, less well developed version of the same case, see Femia (1996). 95   The quoted phrases in this and the preceding two sentences are from Sunstein (2002a, 176–­77, 187). 96   See Sunstein (2002a, 187).

124

T h e B u rd e n of Ju stific a tion 97

just how they constitute “a danger of deliberative democracy.” They may perhaps constitute a danger of political argument, but they are a danger of democratic deliberation only under a thoroughly confused understanding of what the latter entails. A skeptic might insist that we are being overly dismissive and that worries about enclave deliberation are warranted by actual empirical cases. One such example involved a recent dispute over educational reform aimed at remedying the widely acknowledged “apartheid in American public schools.” The cast of characters consists largely of “white middle class voters” who benefit from that apartheid yet who purportedly “live in towns that are paragons of deliberative democracy.” Given these circumstances, [i]n 1995 . . . a statewide Connecticut plan to reduce school desegregation was duly deliberated upon at great length in repeated New England 97   This last phrase is the title to Sunstein (2002b). A charitable reader might claim that Sunstein simply is raising a warning against various theorists who propose that deliberation be located in a loose array of diverse publics or associations (e.g., Fraser 1992; Cohen and Rogers 1995; Benhabib 1996, 87). We think this interpretation goes quite wide of the mark. Consider the extended version of his argument (Sunstein 2001). There, Sunstein observes that “it is impossible to say, in the abstract, that those who sort themselves into enclaves will generally move in a direction that is desirable for society as a whole or even for its own members.” This is a plausible claim. But so, too, is the converse claim. What we need, for either analytical or normative purposes, is to examine the conditions under which political argument proceeds. And it is just here that Sunstein presents his case in a fantastic package. He devotes the first half of the book to advancing a warning of how individuals might use Internet technologies in ways that, in the aggregate and unintentionally, could subvert the deliberative politics so essential to his own republican vision. But his tale is quite literally incredible. He depicts a dystopian world populated by large numbers of wholly isolated consumers who—­because their ability to filter out discordant views is “total,” “complete,” “perfect,” or “unlimited”—­not only can but will successfully deploy information and communication technologies to avoid entirely any and all unexpected, unplanned, and often unwanted exposure to divergent points of view. The risk is a population dangerously bereft of common experiences, impervious to divergent political views, and susceptible to rampant, rapid cascades of fallacious information. A dire, fearful, perilous world indeed! In painting this portrait, however, Sunstein trades on a mix of equivocation, exaggeration, and implausible inference. The irony in all this is not simply that none of the theorists who might constitute plausible targets for Sunstein’s concerns themselves envision publics or associations anything remotely so isolated as the groups he depicts. Nor is the irony that Sunstein seems himself to doubt that new communication technologies in fact systematically pose the dangers against which he warns. The real irony is that his rhetorical strategy, relying as it does on inducing apprehension, fear, and suspicion, may well leave duly alarmed readers even less inclined to engage in reasoned, deliberative politics than they might be otherwise. It is quite difficult to see how this might contribute to Sunstein’s avowed aim of fortifying “the preconditions for maintaining a republic.” The passages quoted in this note are from Sunstein (2001, 77, 10, 14–­15, 48, 101–­2, 68, 201).

125

C h ap te r 4

town meetings in which the inner city residents of Hartford and New Haven had virtually no effective voice at all. As a result, their interests were simply ignored and [the] plan easily was defeated.98 This might stand—­on a very charitable interpretation—­as an example of exclusionary political bargaining or of insulated political debate. As such, we might take it, as perhaps it was offered, as an example of why we require more deliberative arrangements. But our objection remains. In the first place, this hardly is an instance of an “enclave” population completely isolated from external views. Rather, it is an example of a privileged population who, confronted by claims and arguments that they find discomfiting, are able and willing to ignore those discordant demands by excluding those who press them. Second, a situation such as the one sketched here, in which the most disadvantaged of those concerned effectively are deprived of meaningful voice in discussions about how extant arrangements might be reformed, hardly counts 99 as democratic deliberation under the most charitable definitions. As a warning against too readily idealizing which, if any, of our extant political practices count as “deliberation,” this surely is valuable. In most instances, as in the case here, what this will demonstrate is just how distant actually existing arrangements are from deliberative ideals. The risk is that we come to see the former as a sham, a venue for rationalization, ignorance, prejudice, and selfishness, and so the latter as impotent. Now consider what may well be the converse claim. Some theorists insist that there are “mechanisms endogenous to deliberation itself to change views 100 and beliefs in a benign direction.” This view is very nearly as credulous as the preceding one is confused. We hardly dispute that political argument can have crucially important practical effects. Yet little work has been done to persuasively identify the mechanisms by which this occurs. And—­even setting the group polarization thesis to one side—­there is little reason to suppose that the mechanisms that animate political argument necessarily generate outcomes that are benign or justifiable in any general sense. Just as is the case 98   Shapiro (2002b, 207–­8). For background on this example and especially the dire educational prospects for the children of “inner city residents” excluded from the deliberative process here, see Hochschild and Scovronick (2003, 46–­48). Note that they, too, characterize the decision-­making in this case as “deliberation.” 99   Indeed it does not meet Shapiro’s own definition of deliberation, which involves “solicitous search for right solutions in circumstances of conflict.” The emphasis in this definition “is less on giving reasons for one’s own views and more on soliciting reasons from others for theirs” (Shapiro 2002b, 196, 198). It is impossible to see the example Shapiro sketches as a solicitous search for anything but continued educational segregation. 100   Dryzek 2000, 169, 42–­7.

126

T h e B u rd e n of Ju stific a tion

with voting, where we have little warrant to suppose that the sheer act of casting and counting votes will generate a determinate outcome, let alone one that might be normatively justified, so we cannot presume that “discussion, debate and persuasion” per se will do so. In fact, it is entirely possible that political argument, even under conditions that approximate democratic deliberation, might, by clarifying or exposing previously unappreciated differences, 101 exacerbate rather than ameliorate disagreement. More generally, there is no reason to claim, without further argument, that the mechanisms by which argument coordinates social and political interaction will have “benign” effects or that the effects it does induce will emerge in “benign” ways. We return to this question in the next chapter. At present, we simply wish to suggest that advocates of deliberation risk creating confusion and incredulity if they prejudge empirical or analytical matters in ways they find normatively congenial. 101   We made this point some years ago. See Knight and Johnson 1994. Shapiro (2002b, 198–­ 99) has subsequently reiterated it.

127

Chapter 5 Reconsidering the Role of Political Argument in Democratic Politics

I.  Introduction We have argued that when considering democracy, problems of aggregation are unavoidable. In so doing, we have sought both to recognize the problems to which aggregation is susceptible and to suggest that despite those problems, and the strenuous efforts of theorists to minimize or decenter the role of voting in democratic politics, it remains central. We also have raised the possibility that political argument might afford a remedy for the persistent problems of aggregation. To that end, we offered a broad sketch of how we understand political argument and how it operates, along with a promissory note regarding the ways we plan to fill in that initial sketch. We devote the bulk of the present chapter to cashing in that promissory note. In this chapter, we first take up, and disagree with, what we call the standard case for democratic deliberation. That case, as we see it, seeks to identify some aspect of reasoned argument that will lead participants to alter their substantive preferences over outcomes. It holds out the possibility, on that basis, that parties to deliberation will, through the back-­and-­forth of offering and responding to reasons, minimally mitigate conflict and ideally at least approach if not actually attain agreement or consensus. In this way, the standard account fails to differentiate between analytical, explanatory, and normative undertakings. In response, we advance a different view of political argument and how it might work to structure disagreement among relevant actors. And we suggest how the persistence of disagreement allows us to grasp the crucial consequences of democratic decision making—­its cognitive effects, the way it coordinates socially dispersed knowledge and information, and its particular reflexive character. Our distinctive case for the priority of democracy rests on that final quality—­the reflexive character of democratic institutions.

Th e R o le o f Po litic a l A rg u me nt

II.  Identifying Mechanisms Dewey insists that “language . . . is . . . the instrument of social cooperation 1 and mutual participation.” Even the most die-­hard skeptic would have a difficult time disagreeing with this view. But the skeptic might well quote Dewey back at us—­and at defenders of democratic deliberation more generally. For Dewey also suggests “that the fruit of communication should be participation, 2 sharing, is a wonder by the side of which transubstantiation pales.” From the perspective of democratic theory, the analytical task of identifying how language “works”—­more precisely, of specifying the mechanisms that animate communication in social and political interactions and the circumstances under which they do so—­whether in conversation, in political argument or 3 in deliberation, is crucial. This is especially so, to the extent that what we see as the standard case for deliberation locates both the efficacy and normative attractiveness of deliberative processes in the force of communication. There can be little doubt that language does work in the ways that Dewey, among others, intimates. Recall that the circumstances of politics, on our account, are characterized by irreducible diversity and the disagreement that accompanies it under conditions of interdependence. Suppose, for the sake of argument, that diversity consists solely of differences of material interests and that members of the relevant population are motivated solely by such concerns. And suppose, as well, that the interdependence they confront is unavoidable. In such circumstances, social and political actors face the task of coordinating their interactions despite their inevitable disagreements over how best to do so. This is precisely the sort of situation best captured in austere game-­theoretic models. Even in this unlikely environment, communication has causal force. In some sense, this is no surprise. After all, the entire program of Nash refinements that has preoccupied game theorists for several decades concentrates on establishing the credibility or otherwise of communication in various strategic settings. But in the refinement literature, credibility is tied tightly to payoffs in the sense that no player has any reason to believe and Dewey (1929/1958, xiii). As he later elaborates: “The heart of language is not ‘expression’ of something antecedent, much less expression of antecedent thought. It is communication; the establishment of cooperation in an activity in which there are partners, and in which the activity of each is regulated by partnership. To fail to understand is to fail to come into agreement in action; to misunderstand is to set up action at cross-­purposes” (p. 179). Compare Honneth (1998). 2   Dewey (1929/1958, 166). 3   Although we cannot pursue the point here, this is true even of “post-­modern” political theorists (Johnson 1997). 1  

129

C h ap te r 5

act on any statement (threat or offer) another might make unless it is in the second player’s interest—­as defined by the payoffs to the game—­to follow through on the statement.4 Game theorists, however, also demonstrate that “cheap talk”—­understood as consisting of “costless, nonbinding, nonverifiable messages that may affect the listener’s beliefs”—­can, under some conditions, help social and political actors coordinate their interactions.5 In these models, talk is cheap in the sense that it is not tied directly to players’ payoffs. From a strictly rational-­choice perspective, the notion that cheap talk can have important effects is anomalous. Rational actors—­at least as the actors who populate game-­theoretic models typically are portrayed—­have no business responding to such “payoff-­irrelevant” communication. We, however, are not concerned with exploring this anomaly. Rather, we want to point out two limitations of this research. First, while suggestive, it does not sustain strong normative conclusions. Cheap talk does not necessarily generate determinate practical effects. And the effects it does produce, when it does, need not be normatively attractive.6 Second, the results of cheap-­talk models do not sustain strong analytical conclusions. Not only does such communication not necessarily have determinate practical effects, but game theorists have little to say about why the players who populate their models actually respond to payoff-­irrelevant messages when they do so. They do nonetheless place just that analytical issue on the theoretical agenda.7 Before we turn to that problem—­the problem, that is, of identifying plausible mechanisms—­it will help to consider another literature. Here we have in 4   Kreps (1990c, 108). See also Rubinstein (2000) on the way language enters formal models more generally. 5   The quoted phrase is from Farrell and Rabin (1996, 116). See also Austen-­Smith (1992) and Matthews and Postlewaite (1995). 6   Farrell and Rabin conclude their review of this area of research by stressing that language is not quite as miraculous as Dewey might have us believe. Misunderstanding and conflict in particular can significantly dampen the effects of cheap talk. The broad picture seems to be as follows: “We see no general lessons. Indeed we reject several enticingly simple (and popular) ideas: that cheap talk ensures players will play a Nash equilibrium; that an efficient outcome will emerge (even if this means efficient among Nash equilibria); or that all information will be shared (or even as much information as is consistent with incentives in the sense of equilibrium)” (Farrell and Rabin 1996, 116). 7   More generally, Rubinstein (1991, 921) is correct when he observes that “although language plays a crucial role in resolving conflicts, game theory has so far been unable to capture this role.” This remains true despite his more recent efforts on that score (Rubinstein 2000). Among political scientists, recent approaches to this general problem founder, in our view, on the unargued assumption that political argument and debate are best thought of in terms of the difficulties of disclosing or withholding “information” (e.g., Landa and Meirowitz 2009).

130

Th e R o le o f Po litic a l A rg u me nt

mind the large and growing experimental literature on cooperation in various sorts of “social dilemma.”8 Much of this work has demonstrated that experimental subjects, like the players in cheap-­talk models, regularly act in ways that confound the expectations rational-­choice theorists have inculcated in us. That is to say that experimental subjects regularly cooperate in the resolution of social dilemmas at rates far higher than rational-­choice theorists would have us expect. The details of this work are unimportant for present purposes. What is relevant is that communication plays a crucial and distinctive role in inducing cooperation among subjects in social dilemma experiments.9 Two things are important here. First, analysis of this literature suggests that communication effects are independent of such other plausible sources of cooperation as altruistic motivation, social norms, group or subgroup identification, and so forth. Second, in both one-­shot and repeated treatments, the effects of communication are both independent of such competing explanatory factors and typically more powerful. These results sharpen the conclusion of the preceding paragraph. For while they show that communication works, they in fact eliminate several plausible underlying mechanisms that might account for how it works. Having done that, however, these experimental studies, too, leave open the underlying question of just how communication coordinates social and political interaction.

III.  The Problem of Mechanisms I: Communicative Action The most systematic attempt to analyze the mechanisms animating the efficacy of language in social and political interaction is perhaps the theory of communicative action that Jürgen Habermas elaborates. As is well known, Habermas seeks to draw a central, categorical distinction between strategic and communicative action. On his account, the former is oriented to “success” at influencing the actions of other rational actors, while the latter is oriented toward the “understanding” of actors involved in the task of cooperatively defining the context of their ongoing interactions in ways that facilitate the pursuit of their individual plans. In this sense, Habermas accords communicative action, understood in a quite precise way, a privileged analytical 8   We rely on the survey and reanalysis of this work conducted by Sally (1995). Many, but not all, of the “social dilemmas” in this literature have the strategic structure of a multiperson prisoner’s dilemma that we discussed in chapter 3. 9   It is important to note that communication effects induce cooperation at rates even higher than the anomalous rates found in the absence of communication.

131

C h ap te r 5 10

role in any account of social and political order. We obviously cannot discuss 11 this work in any detail here. It nevertheless is important to recognize several crucial aspects of Habermas’s project. In the first place, the position Habermas elaborates is ambitious in a sense that is particularly important for our purposes. The distinction between strategic and communicative action is, for him, categorical rather than simply analytical. Actual interactions might incorporate a complex mix of types of action. But in any given instance, competent agents can, in principle, grasp whether their speech acts are oriented toward success or understanding. Second, in their efforts to reach and sustain understanding, parties to communicative action necessarily advance and respond to what Habermas terms “validity claims.” These claims—­claims to truth, normative rightness, and sincerity that competent speakers unavoidably raise in their speech acts—­constitute the distinctive mechanism in his social and political thought.12 But they also afford the basic component of Habermas’s project of providing a normative fulcrum for critical social and political theory. Finally, communicative action involves participants in the effort to cooperatively negotiate shared understandings of their ongoing social and political interactions. To do this, it operates on two levels. In everyday communication—­what we call conversation—­this process of mutual negotiation remains implicit. In more reflective moments, usually those that emerge in response to situations that are “problematic” in the sense that shared understandings are contested, it is made explicit in processes of argument. The difficulty with this enterprise is that Habermas has not successfully defended the basic categorical distinction between strategic and communicative action.13 His own account presupposes familiarity with the intricacies of philosophical analysis of speech acts. For our present purposes, it will suffice 10   As he states succinctly: “With the concept of communicative action, the important function of social integration devolves on the illocutionary binding energies of a use of language oriented to reaching understanding” (Habermas 1996, 8). 11   For a précis of his views, see Habermas (1996, 1–­27). For a more detailed discussion of the points we make in this and the next paragraph, see Johnson (1991; 1993). 12   Thus, Habermas insists that communicative action is distinct from simple conversation, is not reducible to interest-­governed bargaining, is not merely conventional, and is not equivalent to norm-­governed action. For relevant citations, see Johnson (1991, 188–­99). In this respect, his work converges with the game-­theoretic and experimental research we report above. 13   As one sympathetic critics remarks: “Habermas’s position may in the end turn out to be correct, but the arguments he advances are in general insufficient to settle the difficult questions that must be answered in order to establish this” (Heath 2001, 45).We thus can argue that Habermas has yet to advance a convincing case for his elementary distinction between strategic and communicative action while remaining agnostic about the prospect that he might do so in the future.

132

Th e R o le o f Po litic a l A rg u me nt

to state the problem in schematic form.14 Habermas wishes to differentiate not only between the sort of effects that strategic action and communicative action induce but between the ways that they do so. To this end, he argues that strategic action consists of perlocutionary speech acts that, for simplicity’s sake, we label SA1. Utterances of this sort have causal impact such that in making them, we do something by saying something. So if, in the course of political argument, the parties engage in invective, make jokes, or rely on other rhetorical strategies in hopes of inducing interlocutors to change their minds, they would be engaging in this sort of action oriented to success.15 Habermas likewise identifies communicative action with illocutionary speech acts for which we use the label SA2. Utterances of this sort have what might be called constitutive force such that in making them we do something in saying something.16 So far, so good. However, Habermas also insists that when an agent engages in communicative action, she relies solely on “the force of the better argument” in the expectation that, by presenting her best argument, she might convince or persuade those who disagree with her. The speaker, in other words is oriented toward understanding rather than success. Yet, within the speech act framework on which he relies, any endeavor to convince or persuade another, unfortunately, is a form of SA1 and so cannot, on Habermas’s own account, be an instance of communicative action which, by his own definition, consists solely of utterances in category SA2. Stated more forcefully, action oriented to success emerges at the center of his account of communicative action, thus depriving his theory of its most provocative implications. This may seem like an overly pedantic criticism. Yet the distinction between types of action is absolutely fundamental to Habermas’s larger project and, in particular, to his case for “discursive democracy.” Thus, Habermas claims that “the legitimating force of a discursive process of opinion-­and will-­formation” 14   For those familiar with speech act theory, Habermas identifies strategic and communicative action with perlocutionary and illocutionary acts, respectively. For a fuller analysis of this claim, see Johnson (1991; 1993). Having said this, however, it is important to note that, as conceived in the speech act theory upon which Habermas draws, the illocutionary force of an utterance derives in large measure from its place in ritualized or conventional practices. It therefore is difficult to see how he can maintain the further distinction he draws between communicative and norm-­ governed action. We set that matter aside. 15   Shiffman (2002) and Basu (1999) advance the examples of invective and humor respectively as part of more or less friendly criticisms of overly rationalist conceptions of deliberation 16   By contrast, it is important to note that some advocates of democratic deliberation who claim “roots in critical theory” present their work as distinctive just insofar as it recognizes “discourses as real entities with causal force when it comes to the content and transmission of opinions” Dryzek (2000, 56). Others like Young (2000, 66) back away from the categorical nature of Habermas’s distinction.

133

C h ap te r 5

in law and politics operates through processes of argumentation that trade upon SA2—­“the illocutionary binding force of a use of language oriented to mutual understanding”—­which, in turn, operates “to bring reason and will together and lead to convincing positions to which all individuals can agree without coercion.”17 Here we see how his analytic and normative projects converge: the noncoercive agreement of individuals in legal and political interactions reflects the fact that communicative action, since it trades solely on SA2 rather than SA1, does not operate causally to induce agreement of compliance. Agreement instead is voluntarily undertaken. Clearly, claims of that sort simply make no sense unless one can distinguish between strategic and communicative action in the way Habermas proposes. To the extent, then, that Habermas fails to sustain the categorical distinction between strategic and communicative action, the larger edifice of his social and political thought is unsound. And while we are not concerned narrowly with Habermas’s project, we wish to underscore the extent to which theorists who ground their case for democratic deliberation in some notion of “deliberative” or “communicative” or “discursive” reason are under a considerable burden to explain just what they mean.18 The difficulties with broadly Habermasian approaches do not end there. Recall that we are concerned with identifying the mechanisms that generate the sorts of independent effect that communication has in game-­theoretic models and in experimental settings. Suppose that Habermas or his followers or those involved in some analogous undertaking successfully establish that efforts to convince or persuade people actually involve utterances of type SA2. Such utterances typically derive their force from the often informal institutions, the ritualized or conventional practices, in which they are embedded. Suppose that in response to a court officer’s query “Do you promise to tell the truth, the whole truth and nothing but the truth, so help you God?” you reply Habermas (1996, 103). Some, in fact, seek to do just this (e.g., Bohman 1996b; Heath 2001; we return to their work below). Others, however, do not. Hence we are told “that a defensible account of deliberative democracy has to be underwritten by a conception of communicative rationality” (Dryzek 2000, 147). Still other theorists more loosely claim that “the discourse model of ethics formulates the most general principles and moral intuitions behind the validity claims of a deliberative model of democracy” (Benhabib 1996, 70). See also Young (2000, 24) who invokes the notion of “deliberative reason.” This is true not just of the theorists we explicitly mention here but, by extension, of anyone who follows Habermas in grounding their accounts of democratic deliberation in his “ideal reconstruction of the presuppositions of communication” (Chambers 1996, 8, 90–­105). Given the difficulties we have identified with Habermas’s account, it would be helpful to know whether the language of “communicative rationality,” “discourse ethics,” “validity claims,” and so forth, serves as more than window dressing in such statements. 17   18  

134

Th e R o le o f Po litic a l A rg u me nt

“I do.” In saying “I do,” you do something, namely assume a legal obligation. But notice that your utterance is part of an institutionalized legal practice of swearing an oath and that your utterance draws its force from this setting. The same is true of “testimony” regarding personal experience in the context of African-­American social and religious practices, of “greetings” or “story-­ telling” as modes of social acknowledgment and inclusion, and of oratorical 19 styles more generally. Here it is difficult, if not impossible, to attribute independent force to communicative action or its analogues. Moreover, one must adopt an exceedingly sanguine view of informal institutions and practices to suppose that they both sustain the independent effects these sorts of speech act generate and afford them some sort of presumptive normative privilege 20 as well. After all, informal institutions typically will emerge, on our view, in just the way we suggest in chapter 3, and they therefore will stand in need of justification before they can supply any themselves. Despite its flaws, Habermas’s work raises two issues that are especially germane. The first revolves around the task of identifying underlying mechanisms. Habermas sees this as a normative as well as explanatory task. He not only wants to flesh out a particular explanatory account. He also wants the elements of that account—­and especially his conception of communicative rationality—­to bear normative weight. Hence, because parties to communicative action rely on utterances of type SA2, any understanding they establish will not result from causal influences and, in that sense, will be “free.” We think this ambition is misplaced. Habermas, for reasons we have provided, has not delivered on it. And, as we show in the next section, once democratic theorists retreat from Habermas’s ambitious project, they are no more successful at the task of identifying plausible analytical mechanisms—­let alone ones that simultaneously can satisfy their normative expectations. Second, insofar as parties to communicative action, on Habermas’s account, aim at shared understanding, they aim at consensus or agreement. We, too, think that political argument succeeds, when it does, by helping to establish shared understandings among relevant actors. However, we also suggest that such shared understandings operate in a broadly cognitive manner and in no way entail agreement or consensus on substantive political matters. Indeed, we suggest why such agreement is neither necessary nor desirable in democratic polities. 19   These examples are advanced as more or less friendly criticisms of overly rationalist conceptions of deliberation by Sanders (1997), Young (2000), and Remer (1999; 2000), respectively. 20   Minimally, then, it seems safe to say that those who seek to revitalize some notion of communicative action by reference to informal practices of testimony, greeting, or oratory face a considerable burden of argument. To be clear, this is not an approach Habermas would endorse. Nor are those who adopt it necessarily inspired by his work.

135

C h ap te r 5

All of this raises obvious questions concerning the nature of agreement in politics and, in particular, about whether and, if so, how political argument might induce it. We turn to this question in section 5 below.

IV.  The Problems of Mechanisms II: Alternative Possibilities Advocates of democratic deliberation devote relatively little attention to other mechanisms that might animate political argument. Some revert to a revised theory of social norms. But this strategy retreats from the task of identifying 21 the independent efficacy of “discussion, debate and persuasion.” Others 22 present admittedly open-­ended lists of mechanisms of varying plausibility. Consider some of the possibilities. Most obviously the exchange of reasons in political argument can generate information both about those who are party to it (their preferences, beliefs, experiences, and identities) and about the circumstances in which they find themselves (about the “facts” of those circumstances as well as about the consequences of either sustaining or, in one or another way, altering them).23 But political argument also might work in other ways. It might help to mitigate the bounded nature of human rationality by, for example, (1) commanding participants’ attention or capturing their imagination, thereby opening or making available unexplored, unrecognized alternatives or possibilities, (2) highlighting unappreciated interdependencies in which they find themselves, or (3) framing or reframing issues in more or less productive ways.24 Political argument also can help clarify (or obscure) the tacit, often incompletely shared premises and understandings that inform 21   Heath (2001). It also is fair to say that the particular, provocative theory of social norms that Heath develops is contestable. The reasons why are beyond our current concerns, however. 22   In this paragraph, we draw on the overlapping lists compiled by Bohman (1996b, 55–­66), Fearon (1998) and Dryzek and List (2003, 9). Bohman (1996b, 59) offers his list as “open-­ended” and denies any suggestion that it might be “exhaustive.” He suggests that offering and responding to reasons can, via such mechanisms, generate “deliberative uptake” for views in the face of problematic social interactions. Fearon (1998, 44), too, concedes that there may well be mechanisms other than the ones he mentions, but he suggests that his list probably includes the “most important” ones. Likewise, we make no claim to completeness. We seek only to illustrate some of the possibilities. 23   Landa and Meirowitz (2009). 24   The first and third of these possibilities depart from an information framework precisely because the speech acts involved will operate—­as in our earlier discussion of principles, traditions, and precedent—­in the subjunctive rather than the indicative mode. That is, they take up not what is or is not the case but what might be possible.

136

Th e R o le o f Po litic a l A rg u me nt

extant practices, including argumentative practices. And finally, it can help constrain indeterminacy in situations where the parties must apply general norms or rules to particular instances or cases. Whether and how any of these mechanisms might operate in particular settings is largely an empirical matter. For our purposes, it is more important to see how they fit into broader defenses of deliberation. In highly schematic form, the argument proceeds as follows: Singly or in combination, mechanisms like those we just mentioned induce parties to political argument to reflect on their preferences, beliefs, presuppositions, and so forth.25 Such reflection, in turn, leads relevant actors to transform or change their preferences, beliefs, and presuppositions in one of two ways. Either they come to hold their preferences, beliefs, and presuppositions more self-­consciously, and so more 26 confidently, or they actively change their mind on some substantive matter. And as a result of such transformations, the standard argument for deliberation presumes that parties to deliberation will, through the back-­and-­forth of offering and responding to reasons, minimally mitigate conflict and, at least ideally, approach if not actually attain agreement or consensus. One might challenge this argument at several places. One might question whether, as an empirical matter, political argument induces the necessary transformations of preference, belief, or presupposition at the individual level. Granting that such transformations might occur, one might question 25   Thus, Benhabib (1996, 71) suggests that deliberation can move individuals to a greater “level of conceptual clarity about their choices and preferences” and suggests that this is endogenous to deliberation insofar as the “very procedure of articulating a view in public imposes a certain reflexivity on individual preferences and opinions.” Similarly Dryzek (2000, 162) defines “authentic” deliberation in terms of the requirement that arguments “induce reflection upon preferences.” We set aside the issue of whether such reflection actually presupposes public deliberation (Goodin 2000). 26   The focus typically is on preference transformation. This might take the form of exposing and revising objectionable preferences or of inducing reflection and consideration of the grounds for holding otherwise unobjectionable preferences. In either case, we repeatedly are told that a core feature of deliberative mechanisms is that they change preferences. Thus, Dryzek announces on the first page of his recent book that “preferences can be transformed in the process of deliberation” and subsequently presents this as the “defining feature” of deliberative approaches (Dryzek 2000, 1, 32). So, too, one critic asserts in his very first sentence that deliberation aims “to change the preferences on the basis of which people decide how to act” (Przeworski 1998, 140). For similar views, see Cooke (2000, 948); Elster (1986, 112); Phillips (1995, 148); and Young (2000, 26), It is therefore unsurprising that one recent review of the literature on democratic deliberation defines it as “discussion aimed at producing reasonable, well-­informed opinions in which participants are willing to revise preferences in light of discussion, new information, and claims made by fellow participants” (Chambers 2003, 309).

137

C h ap te r 5 27

whether they occur in the “right way.” By this, we mean that one might ask whether the mechanisms that animate political argument reliably operate in normatively “benign” ways. One might also question—­again granting that the relevant transformations occur at the individual level—­whether in the aggregate such transformations will have discernible practical effects. Finally, one might question whether, assuming both that reflection transforms preferences, beliefs, or presuppositions at the individual level and that the resultant views have discernible aggregate-­level effects, the resulting outcomes will be normatively attractive. For democratic theorists, these last two questions are especially pressing insofar as we are concerned with the genesis and justification of binding, properly collective decisions. It may be difficult to appreciate the force of such potential objections. In order to establish it more clearly, we discuss two potential mechanisms that defenders of democratic deliberation have advanced as parts of their arguments. The first example is quite speculative and so potentially objectionable. The second, however, is more innocuous and so perhaps more telling. Our aim is not to establish that each and every one of the mechanisms that we identified above are susceptible to just the objections we raise. We suspect that many of them are. Instead, we simply wish to show how, to the extent that they base normative claims on the analytical task of identifying the mechanisms underlying political argument, common defenses of deliberation assume a very heavy burden of argument. It is a burden that, to be charitable, they very rarely meet. And since it is a burden of their own making, it is one we think they need not even bear. Consider the first example. Some defenders of democratic deliberation speculate about the possible “civilizing force of hypocrisy” as part of an account for how, over time, those who initially invoke some impartial princi28 ple for strategic reasons might over time come to sincerely embrace it. On this view, public debate might generate psychological pressures toward self-­ censorship. These pressures deter speakers from advancing claims, expressing preferences, or endorsing beliefs that the speaker anticipates others will consider objectionable. Such strategic self-­censorship expresses not so much genuine conviction as anticipation of disapproval or reproach or even stronger social sanctions. Yet, over time, something like dissonance reduction might induce such parties to actually adopt the “unobjectionable” positions to which 27   Here we draw on Davidson’s (1980) discussion of the requirement that reasons must work “in the right way” when they cause our actions for us to attribute rationality to those actions. 28   See Elster (1995, 250; 1986, 113) who advances this possibility as “speculative.” Several advocates of deliberation nevertheless endorse it as one of several possible mechanisms at work in democratic deliberation. See, e.g., Sunstein (1993, 244); Williams (2000, 144); Dryzek (2000, 46); and Dryzek and List (2003, 16).

138

Th e R o le o f Po litic a l A rg u me nt

they earlier paid only lip service. This is one way that political argument might transform individual preferences, beliefs, or presuppositions. There are several difficulties with this view. First, at the individual level there is little reason to assume that hypocrisy indeed “civilizes” in anything like the manner that those who propose it suggest. Indeed, hypocrisy might simply induce parties to political debate to embrace modes of argument that allow them to attack policy proposals or candidates indirectly. In this way, they might privately retain their opposition to, say, widely embraced principles of democratic equality, publicly express that opposition indirectly through warnings about the grave unintended consequences that policies aimed at promoting those principles would generate, and then bask in an exaggerated sense of their own political superiority that they infer from the clever nature of their 29 argument. In this scenario, hypocrisy is not simply reinforcing itself but is breeding detachment, reinforcing contempt, and inflating self-­importance. Whether hypocrisy tends to “civilize” as some propose, works in the manner we sketch, or perhaps operates in some other way is, of course, an empirical question. Our point is simply that the plausibility of the civilizing force of hypocrisy as an explanatory mechanism in the current context hardly is immune to the first level of objection we mentioned earlier. It may not, in other words, actually transform individuals at all. And the problems hardly end there. Notice, in the second place, that this mechanism hardly works in the “right way.” The entire process captured by the idea of the civilizing force of hypocrisy is parasitic on practices of dissimulation and exploitation. It presupposes that speakers engage in the “strategic use of impartiality” in the sense that, by invoking some principle for strategic purposes, they not only misrepresent their own actual views but exploit the expectations of others regarding the existence, if not prevalence, of some among the relevant population who sin30 cerely endorse the principles in question. Moreover, the alleged civilizing effect of hypocrisy trades on the existence and functioning of psychological mechanisms that first induce speakers to engage in self-­censorship and eventually cause them to embrace the resultant views. Over time, such mechanisms “might even bring about a transformation of preferences and values, simply by making venal or self-­regarding justifica31 tions seem off-­limits.” But these sorts of causal process operate behind the 29   Hirschman (1991, 10–­42) details the “perversity thesis” common to conservative political rhetoric in just these terms. 30   The quoted passage is from Elster (1995, 249). 31   Sunstein (1993, 244). We set aside the issue of how much time is required for hypocrisy to civilize.

139

C h ap te r 5

backs of agents to induce a conformity that is at once rather shallow, since it does not reflect considered commitments, and normatively suspect insofar as it emerges not from the consideration of reasons but from “blind” psychologi32 cal processes. Finally, when we turn to the aggregate level, this process might well have quite pernicious effects. For even if we set aside the matter of how speakers, having first engaged in self-­censorship, might then come to embrace the censored or filtered view, the prior question remains about whether they will do so. Some might. But others might not. It surely is plausible to suspect that at least some members of the population, perhaps even a majority, might be pressured into endorsing conformist positions despite the fact that those positions do not reflect their actual beliefs, preferences, or presuppositions. Here the aggregate consequences can take at least two different unattractive forms. First, under pressure of what they deem a hostile intellectual climate, some actors might express their opposition through various “reactionary” forms of argument. At the individual level, this can breed the sort of arrogant conservative nonconformity we already have mentioned. At the collective level, it can undermine political engagement insofar as such argumentative strategies represent “contraptions specifically designed to make dialogue and deliberation impossible.”33 Second, conformist pressures might induce simple hypocrisy absent any purportedly redeeming civilizing effects as significant members of the population engage in “preference falsification” and similar psychological defense mechanisms.34 In such instances, political argument is potentially impoverished insofar as consideration of certain topics is preempted and use of some modes of argument proscribed not because they have been shown faulty in reasoned debate but because potential speakers are dissuaded from raising them for fear of political or social sanctions. These objections, of course, raise analytical and empirical questions, and difficult ones at that. We do not pursue them here. Our point is simply that political theorists cannot presume that just because they can identify some mechanism that might animate political argument, such a mechanism 32   For example: “Dissonance reduction does not tend to generate autonomous preferences.” Elster (1986, 113). Consequently, claims about the “civilizing force of hypocrisy” hardly comport well with the effort to ground democratic arrangements on autonomy. For such an effort, see Sunstein (1993). 33   Hirschman (1991, 170). Note well that this is not a failing to which conservatives are uniquely susceptible. Hirschman makes this point emphatically. And, of course, Hirschman (1993) subsequently conceded that, like conservatives, liberals too engage in such gambits. 34   For recent arguments to this effect, see Loury (1994) and Kuran (1990; 1993). But the theme here is easily traced back at least as far as J. S. Mill’s arguments on liberty.

140

Th e R o le o f Po litic a l A rg u me nt

necessarily will lead to the outcomes they posit or, in instances where it does, necessarily do so in normatively attractive ways. Those who endorse the standard case for deliberation, however, might respond that our argument rests on a deliberately provocative example. In response, we consider a second example that should be considerably less troublesome in that respect. It turns out that it is susceptible to similar sorts of objection. Many advocates of deliberation envision it as a process of political argument that generates and disseminates information in light of which participants might reflect upon their preferences and opinions.35 Again, information effects might enable parties to political argument to develop more complete preferences in instances where they previously may not have had them. They might simply illuminate the preferences or beliefs that relevant others hold. Or they might allow the parties to revise their original views when they more clearly understand the consequences of implementing those views.36 At the individual level, it is easy enough to grant that, as an empirical matter, political argument might have such effects. But it is less easy to attribute this to the operation of “endogenous mechanisms.” There is some tentative evidence that under highly stylized conditions, debate and argument may, in fact, have such effects.37 Yet there also is reason to expect that in less rarified strategic settings, political argument will do little to increase available information.38 Here we see that while political argument might well generate the sorts of information that enable participants to develop more complete preferences, envision new solutions to political problems, and recognize, if not appreciate, Benhabib (1996, 71). Compare Dryzek (2000, 2, 162). On the possible effects of greater information, see Manin (1987, 349–­51). Compare Przeworski (1998) on the difference between “technical” beliefs about matters of fact or the consequences of adopting alternative policies and “equilibrium” beliefs regarding the beliefs and preferences of relevant others. 37   Luskin, Fishkin, and Jowell (2002). We do not wish to dismiss this evidence. Nor, however do we wish to overstate its significance. As the authors concede in reflecting on the particular changes in issue preference (crime and punishment) that their study reveals: “Why these particular changes? We can offer a few impressions, based on first-­hand observation and statistical bits and pieces” (Luskin, Fishkin, and Jowell 2002, 470). That said, three caveats are in order. First, we place to one side the question of whether deliberative opinion polls of the sort that Luskin, Fishkin, and Jowell report actually reproduce conditions of democratic deliberation. Second, Luskin, Fishkin, and Jowell are quite clear that while they can measure information acquisition relatively easily, they have no reliable measure for claims that participants hold their views more reflectively. Finally, since what the authors assess are levels of statistically significant (or otherwise) attitude change, they actually have little to say about underlying mechanisms. See Luskin, Fishkin, and Jowell (2002, 480, 484). 38   See Austen-­Smith (1992; 1995) and Farrell and Rabin (1996). 35   36  

141

C h ap te r 5

competing views, its ability to do so will depend on the extent to which it takes place under conditions that approximate democratic deliberation. Suppose that the additional information generated through political argument not only induces parties to reflect on their preferences, beliefs, and presuppositions but does so in normatively defensible ways. When we turn to the aggregate level, we must wonder if such changes in fact amount to anything and, if so, what. There are several possibilities. It might be the case that, as standard defenses of deliberation presume, parties to political argument not only will come to express more reflectively held preferences but that their preferences will converge and so generate agreement. But it also may turn out that argument-­induced changes in individual preferences, beliefs, and presuppositions generate little change in the aggregate. Indeed, the same study we mention above suggests that even if individual political actors change their policy preferences in the face of discussion, the net aggregate change might be quite low or even nonexistent since changes in individual preferences may be offsetting.39 Finally, it need not follow from the supposition that political argument causes actors to hold preferences in a more considered, reflective, self-­conscious manner that they will, as a group, reach agreement or even a decision. It may, to the contrary, simply ensure that relevant parties more fully appreciate the nature and depth of their disagreement. And given the circumstances of politics as we depict them, it is not only possible but entirely likely that relevant parties will continue to disagree in good faith. If under these circumstances individuals continue to disagree, however reflectively, at some point either the matter will have to be set aside (which itself constitutes a decision) or it will have to be put to a vote.40 In sum, then, there is no reason to presume that reflection on preferences, even if it induces their transformation, will generate practical effects, let alone “benign” ones at the aggregate level.

V.  The Normatively Attractive Effects of Political Argument on Democratic Decision Making Democratic arrangements operate through the two basic mechanisms—­voting and argument. Voting generates normatively attractive outcomes just insofar as it is constrained by the commonly recognized conditions of freedom and Luskin, Fishkin, and Jowell (2002, 471, 485). Holmes (1988) argues forcefully that removing matters from the political agenda can have salutary effects but rightly insists that adopting such an avoidance strategy in any given instance stands in need of justification. 39   40  

142

Th e R o le o f Po litic a l A rg u me nt

equality. This may seem a contentious claim. Yet even those who defend putatively “minimalist” conceptions of democracy concede that, insofar as they themselves insist that elections must be competitive, democracy must incorporate a rather substantial scaffolding of rights and liberties.41 Indeed, in spelling out the view that foreshadows such recent contributions, Schumpeter rightly notes that his “theory is no more definite than the concept of competition for leadership,” that this demands some approximation to “free competition for a free vote” and where, in turn, this would involve “a considerable amount of freedom of discussion for all.”42 As we have stated—­not just any casting and counting of votes amounts to an election. Unfortunately, the results of social choice theory make us painfully aware of the difficulties—­in the form of unstable or cyclical aggregate outcomes—­that emerge when we impose a set of minimal and apparently unobjectionable conditions of equality and freedom on aggregation procedures. This is where argument or debate plays a crucial role in democratic institutions. But political argument, too, generates normatively attractive outcomes only insofar as it is underwritten by similar conditions. When it operates in circumstances approximating these conditions, argument or debate constitutes what now commonly is called democratic deliberation. Quite felicitously, political debate can operate to enhance the quality of voting decisions in three important ways. First, it can mitigate qualms about democracy expressed by theorists who focus too exclusively on mechanisms of voting. It can do so by diminishing some of the problems identified by social choice theory. More specifically, political argument can shape the cognitive structure of the debate in such a way as to reduce the possibilities that the democratic process will be subject to strategic manipulation. In doing so, it can increase the meaningfulness and stability of voting outcomes. Second, political argument can facilitate collective outcomes that are “better” than those See, e.g., Riker (1982) and Przeworski (1999). Schumpeter (1947, 271–­72 [emphasis in original]). O’Donnell (2001, 9–­13) stresses the importance of this broad normative dimension for putatively “minimalist” views of democracy. This observation, in turn, underscores Isaac’s (1998, 32) insight that Schumpeter, far from offering a “realist” view of democracy, actually advanced a normative argument masquerading as an “empirical” theory. Isaac’s claim makes it plausible to suspect that it hardly is a coincidence that Schumpeter consigned the importance of freedom and equality to a footnote. Such observations, while quite persuasive, seemingly have had scant impact on contemporary debates. See, e.g., Mackie (2009) who excoriates Schumpeter for being both minimalist and realist when arguably he is neither. Likewise, this crucial normative dimension of Schumpeter’s thought should prove discomfiting to those who hope to recruit him to the cause of rendering pragmatism less “utopian,” thereby sustaining the case for “pragmatic liberalism” (Posner 2003). 41   42  

143

C h ap te r 5

that might emerge from alternative institutional forms, better in the sense that they make more effective use of the distributed knowledge that exists in society. Third, and perhaps most important, political argument in a democratic decision-­making environment will allow for a reflexivity that is lacking in alternative arrangements. This reflexivity makes democratic institutions especially appropriate for the tasks of coordinating institutional experimentalism and choice. We believe that these three features of democratic decision making are tightly interdependent. With that caveat in mind, for ease of presentation, we address these effects serially in the next three sections.

VI.  Cognitive Effects on Dimensionality Advocates and critics of democratic deliberation alike commonly presume that 43 it aims at achieving agreement among relevant parties. Recall the standard case for deliberation as we sketched it earlier. It holds that political argument works through some endogenous mechanism or mechanisms to induce parties to reflect on their preferences, beliefs, presuppositions, and so forth. Such reflection, in turn, leads relevant actors to transform or change their values and presuppositions. As we noted, this typically is taken to mean that they change their substantive preferences. Such change minimally requires that speakers embrace their views more reflectively (and so perhaps more confidently, perhaps more tentatively) but more strongly implies that in the face of convincing arguments, speakers actually change their mind on some substantive matter. The consequences of such transformations, according to the standard argument, is that parties to political argument will minimally mitigate conflict as they advance, challenge, and defend their arguments, and if they do not actually attain agreement or consensus, they will at least more closely approach it. This view is misguided for at least two related reasons. In the first place, we think that it is a mistake to focus on the ways that political argument 43   “Reasonable people enter discussion to solve collective problems with the aim of reaching agreement. Often they will not reach agreement, of course, and they need to have procedures for reaching decisions and registering dissent in the absence of agreement. . . . While actually reaching consensus is thus not a requirement of deliberative reason, participants in discussion must be aiming to reach agreement to enter discussion at all” (Young 2000, 24). Other advocates explain that when “citizens deliberate, they seek agreement on substantive moral principles that can be justified on the basis of mutually acceptable reasons” (Gutmann and Thompson 1996, 55). Compare Christiano (1996, 117). Finally, from a critic of deliberation we learn that under relevant conditions, democratic deliberation “will guide the discussion towards generalizable interests to the agreement of all participants” (Mouffe 1999, 747–­48).

144

Th e R o le o f Po litic a l A rg u me nt

transforms or changes substantive preferences. In the second place, we think that the standard view dramatically overstates both the need for and the desirability of agreement. Our aim in this section is to defend these views in turn. Consider first the common claim that political argument induces participants to change their substantive preferences. We are skeptical. But it is important to understand why. We do not question this because we think of prefer44 ences as somehow normatively fundamental and hence immune to revision. Nor do we think it impossible for political actors to change their preferences. Indeed, they might alter their preferences in the course of political argument. That, as we have already suggested, is an empirical question, as is the overall effect of argument-­induced shifts in preferences among a given population.45 Our objection, instead, is that common claims that political argument transforms preferences are either beside the point or too demanding. It is easy enough to see, from what we already have argued, why the claim that political argument induces individuals to hold preferences more reflexively is largely beside the point. Suppose we grant the empirical claim that due to political argument, individuals come to hold their substantive preferences in a more reflective manner. Notice, in the first place, that this empirical claim constrains the analytical options available to us. For it surely is implausible to expect mechanisms that necessarily operate behind the backs of individuals (like the civilizing force of hypocrisy) or ones that arguably often will do so (like framing effects) to induce those individuals to be more reflective.46 That said, suppose further that we postulate that political argument works through a relatively innocuous mechanism, such as the dissemination of information. Then, even if as a result of the increased information that political argument makes available, individuals come to hold their preferences more reflectively, it in no way follows that this will lead to greater substantive agreement at the aggregate level. And this holds for the very reasons we gave above—­any preference changes among a population might well be offsetting in the aggregate or See Goodin (1986), Sunstein (1993), and Ferejohn (1993) for a range of positions on this issue. We simply insist that the normative status of any preference change will depend on the extent to which the argument in question takes place under conditions that approximate those that democratic deliberation requires. 45   See Delli Carpini, Cook, and Jacobs (2004, 337); Ryfe (2005, 49–­50); Thompson (2008, 499); and Mutz (2008, 536). Insofar as the empirical studies surveyed in these essays are meant to gauge the effects of political argument—­instead of testing democratic deliberation—­the current state of knowledge on this score is perhaps best captured in two words that appears in virtually all of the reviews—­“mixed” and “inconclusive.” 46   This objection holds, too, of the factors that Mackie (2006) invokes to account for the lack of empirical evidence that political argument might “change minds.” 44  

145

C h ap te r 5

it may simply be the case that following argument, members of the population disagree more confidently. For there is no reason to think that information is susceptible to only one interpretation, especially if one allows that individual preferences might be shaped by a combination of diverse material interests, ethnical commitments, and cultural attachments. It might be more difficult to see the sense in which claims that substantive agreement based on argument induced preference transformations are overly strong. It will help if we focus on the interface of political argument and voting. For it turns out that what democracy requires is less agreement over either substantive preferences or underlying understandings than structured disagreement about those matters. Recall the various results from social choice theory. Arrow insists that his mathematical analysis illuminates not just the problems that beset aggregation but other important, broader political questions as well. His aim was to defend capitalist democracy against its rivals. In the analysis, he demonstrates that there exists no aggregation mechanism that simultaneously conforms to a set of several relatively unobjectionable normative criteria and generates coherent collective decisions. So, one way to circumvent the impossibility Arrow identifies is to violate one of the axioms he proposes. The standard presentation of his proof gains rhetorical force by showing how aggregation procedures generate collectively rational outcomes while also complying with unrestricted domain, Pareto, and independence of irrelevant alternatives must violate nondictatorship. And no one likes dictatorship. But notice that departing from any of Arrow’s axioms deflates his impossibility result. And so political issues arise once we ask how a given constituency might approach this task. We mentioned in the previous chapter that Arrow’s impossibility result had been generalized in the context of spatial and game-­theoretic models. The generalizations establish that (1) in a multidimensional political space, instability is so general that the resulting cycles in principle encompass the entire policy space, and (2) since the resulting cycles are dependent on the agenda by which alternatives are considered, the entire process is susceptible to strategic manipulation.47 One significant response to these analytical results has been to explore the mechanisms by which institutions constrain indeterminacy in Mackie (2003) goes to great and, we think, excessive and misguided, lengths to establish that models of cycling are empirically insupportable. He thus assumes that they are, in fact, empirically testable. To that extent he subscribes to the same positivism that animates practitioners of “positive political theory.” But there is good reason to think that that animating assumption is mistaken. See Austen-­Smith and Banks (1998, 266–­71). For an extended assessment of the naive philosophical premises that drives both the urge to construct a “positive political theory” and its critics, see Johnson (2010). 47  

146

Th e R o le o f Po litic a l A rg u me nt

ways that “induce” equilibrium outcomes.48 We do not want to discuss it here in any detail. We instead wish to highlight that this literature demonstrates how institutional rules can generate stable outcomes precisely because those rules violate one or more of Arrow’s axioms. Much of this research has focused on legislative institutions. So it seems important to stress that, for example, seniority and committee systems violate “nondictatorship” by enabling occupants of particular roles to have differential and sometimes decisive impact on outcomes. Similarly, procedural constraints on amendment procedures (e.g., germaneness rules) violate the demand that aggregation procedures be unbiased over options that the Pareto condition captures.49 Insofar as Arrow demonstrates that there is an enduring, systematic tension between the efficacy of aggregation mechanisms and the normative aspirations democrats must invest in voting, this literature on “structure-­induced equilibria” seems to punctuate his point. This has led those committed to “positive” political theory—­when they notice the matter at all—­to urge democrats to scale back their aspirations.50 It might rather have suggested to them that they explore the sorts of possibility that Dewey suggests in the passage we cited at the outset. It indeed is possible to work our way backward over the social choice results we just have reviewed, first to majority rule and from there to Arrow himself, and show how each pushes us more or less uniformly toward a particular interpretation of aggregation and of its relation to political argument. Our claim is not that the results we have surveyed in fact add up to a “theory” linking aggregation and deliberation. Instead, we think that these results call for such a theory. Start with the intuition that animates the notion of a structure-­induced equilibrium. This intuition suggests that institutional rules can induce equilibrium in what counterfactually would be a chaotic or highly unstable and, so, indeterminate situation. Social choice theorists have shown that various supermajority (δ-­majority) voting rules have just this effect. In fact, they have demonstrated that under specifiable conditions, equilibria are sure to exist in

48   On “structure-­induced equilibrium,” see Shepsle (1979) for a crucial early contribution. Ordeshook (1997) surveys subsequent work. For agenda control or “setter” models, see Rosenthal (1990). For an overview of these two enterprises and how they relate, see Persson and Tabellini (2000, 19–­45). 49   For a frank recognition of this, see Shepsle (1989, 137). 50   Riker (1982) is the most explicit, extended such exercise. But it is common for positive theorists to note “the obvious” point that “Arrow’s theorem has tremendous implications for modern democratic theory” (Johnson 1998, 21). True, but what is required is some hard thinking about just what those implications are. Riker’s view is hardly definitive on this matter.

147

C h ap te r 5

multidimensional electoral spaces where the δ = 64%.51 Of course, much of the work here is done by the “specifiable conditions.” These amount to restrictions on the structure of individual preferences (that they be “Euclidean”) and on the density of the preference distribution among the relevant population (that the distribution of values not be too polarized). This, of course, violates the condition of unrestricted domain just insofar as it places prior restraint on the sorts of individual values that the voting mechanism can accommodate and still generate a stable collective outcome. This is a theme that recurs through the other results we have surveyed. With this, we return to majoritarian procedures and in particular to simple majority rule. Here we will be especially brief. In the face of more than two options, simple majority rule falters and majoritarian procedures ensure a stable collective outcome only under extremely restrictive assumptions regarding the distribution of voter preferences.52 Simple majority rule is robust for binary choices, but again its robustness is determined relative to various partitions of the domain of possible preferences. So while, as we argued in chapter 4, majority rule remains singularly attractive, it does so only by violating universal domain. Moreover, simple majority rule operates under the threat of ambiguity. And while social choice theorists have suggested that blanket claims of ambiguity are overstated, not only have they not systematically developed such claims but they assume the same sorts of restriction both on the structure of individual preferences and on the density of the preference distribution among the population as did the results on δ-­majority voting rules.53 The underlying theme thus remains the same. Aggregation is significantly less problematic if the domain under which it is expected to operate is restricted.54 Caplan and Nalebuff (1988). The original work here is attributed to Plott 1967. For concise, accessible accounts, see Rae and Schickler (1997) and Johnson (1998). 53   See Levin and Nalebuff (1995), who systematically analyze nine voting methods using simulations. They report that with two exceptions, these methods, while they “differed in the rankings of lower candidates,” do not generate “a different top choice.” Hence they conclude that the outcomes these methods generate do not differ “dramatically.” The major problem is that plurality rule, which is among the most widely used voting methods, was one of the two exceptions (Levin and Nalebuff 1995, 4, 18, 23). 54   This conclusion would not have surprised Arrow. Indeed, having proved his famous result, Arrow almost automatically turned to consider ways to restrict in some manner the domain of individual values that are admissible to our social decision procedures. But Arrow is not concerned simply with abstract axiomatic methods. He explicitly wonders whether such “mathematical restrictions have any social significance” (Arrow 1951/1963, 80, 81). And he intimates that their significance resides in reiterating the crucial importance of “free discussion and expression of opinion” to democratic politics (Arrow 1951/1963, 85). 51  

52  

148

Th e R o le o f Po litic a l A rg u me nt

Despite claims that results highlighting pervasive instability pose a fundamental challenge to democracy, the actual state of affairs turns out to be significantly less dire. For it emerges pretty much systematically that fairly modest constraints on either the structure of individual preferences or the scope of their distribution work to mitigate apprehensions regarding instability. Stated somewhat more technically, once one violates the axiom of universal or unrestricted domain, the threat of unstable collective outcomes diminishes significantly. And it is by now a commonplace among interpreters of social choice theory that 55 political argument might well work precisely to induce just such constraints. There have been numerous more or less plausible proposals for how individual preferences might be structured so as to circumvent cyclical collective 56 choices. The best known of such restrictions is the notion of “single-­peaked” preferences.57 This condition simply states that the members of some electorate share a common understanding or description of what is at issue in some situation even if they disagree in substantive terms regarding how best 58 to respond to it. Single-­peakedness thus is a broadly cognitive or conceptual constraint on the structure but not the substance of individual orderings. It and related restrictions, of course, violate the axiom of universal domain. Yet, if relevant voters can agree about the dimension over which they disagree—­if, that is, they can agree about what is at stake in a particular political conflict—­ aggregation mechanisms need not generate cyclical social orderings. More precisely, if relevant voters have single-­peaked preferences there will be no cycling when choice ranges over only one dimension. This, in turn, would diminish the likelihood that the outcomes of collective decision processes had been manipulated. Political argument might help accomplish this by allowing relevant constituencies to sort out, and hopefully reduce, the dimensions over which they 59 disagree. Indeed, there is some limited empirical evidence that it operates 55   We pointed this out in prior work (Knight and Johnson 1994), as did D. Miller (1992). Not only Arrow (1951/1963, 74–­91) but also Riker (1982, 122, 128, 189) did so even earlier. More recently, List (2002b) and Dryzek and List (2003) have picked up on this theme. 56   See Sen (1966) and List (2002b). The issue of plausibility revolves around whether it is possible to provide an interpretation for one or another mathematical constraint of preference orderings in terms of a social or political theory. Also, as we noted earlier, it is remarkably naive to suggest, as Mackie (2003) does repeatedly, that this structuring is “natural”; indeed, it is a product of social and political processes and so stands in need of justification. 57   Arrow (1951/1963, 75–­80). 58   Riker (1982, 126). 59   This might involve either eliminating possible sources of disagreement—­we might discover that we agree—­or it might simply mean separating dimensions. Holmes (1988) argues in a pointed way for the need to justify any such process for altering the political agenda.

149

C h ap te r 5 60

in just that fashion. That said, we interpret this phenomenon in terms of structured disagreement precisely because it does not imply social consensus. This is true in two important senses. First, the fact that some population indeed has single-­peaked preferences does not mean that its members agree on substantive matters. It means that they share an understanding about or appreciation for what is at stake in some particular conflict. They may still disagree 61 quite vigorously about how to best resolve that conflict. Second, substantial disagreement can persist within a given population even over the question of what is at stake in a conflict. Depending on population size, voting can generate stable collective outcomes if as few as 70–­75 percent of relevant individu62 als converge on a common understanding of the dimension of conflict. In other words, stable collective decisions do not presume excessive homogeneity of either substantive preferences or underlying political understandings. To summarize, democratic decision making can be robust in the face of ongoing disagreement and conflict. Under the proper conditions, political argument might induce just the sort of common or shared understanding of what is at issue in particular circumstances that makes coherent (i.e., non­cyclical) aggregate outcomes possible. Political argument operates to mitigate the instability that plagues voting by allowing participants to try to persuade one another, where persuasion contributes to the formation of beliefs and preferences in a quite specific sense. Persuasion operates, on the view we advance, less to directly influence the substantive content of individual preferences than to impose conceptual restrictions on the structure of individual preferences. It tells us, in other words, not what to think, but what to think about. It need not always succeed in so doing, but when successful, this amounts to establishing shared understanding within the relevant population regarding the dimensions of political conflict. Indeed, democratic politics is perhaps best seen less as a way of reaching consensus or agreement than as an effective List et al. (2009); Farrar et al. (2010). We remind readers of the qualms regarding the enterprise of deliberative polling that we expressed earlier. So while we think this empirical finding is promising, it hardly is definitive. Given the persistence with which this claim has been made, it is surprising that none of those contributing to the burgeoning enterprise of empirically “testing” democratic deliberation seems to so much as mention it. For overviews of this literature, see Delli Carpini, Cook, and Jacobs (2004, 337); Ryfe (2005, 49–­50); Thompson (2008, 499); and Mutz (2008, 536). 61   Riker (1982, 128) makes the point somewhat more colorfully. The fact that the relevant constituency shares an understanding of what is at stake in a political conflict, he suggests, “will not prevent civil war, but it will at least ensure that the civil war makes sense.” 62   Niemi (1969). Note that this study, too, speculates that such convergence can be induced via political argument. 60  

150

Th e R o le o f Po litic a l A rg u me nt 63

way of structuring the terms of such persistent disagreement. We return to this last point below because it is crucial to our assessment of the importance of diversity and experimentation in democratic politics.

VII.  “Better” Decisions What do we mean when we say that one institutional form will resolve such and such a political problem “better” than another? Our position is that under the proper conditions, democratic decision processes make better use of the distributed knowledge that exists in society than do their rivals. However, we need at the outset to clarify the status of this claim. We are not arguing that democratic decision making is the only way to effectively use distributed knowledge. We are arguing rather that, given the structure of democratic decision making, it will outperform other available institutional arrangements in using distributed knowledge for certain purposes. In this section, we address the way in which democratic institutions might effectively employ distributed knowledge. In the next section, where we discuss the effects of reflexivity, we identify some of the types of interactions for which democracy can do a better job than the other institutional alternatives. From our pragmatist perspective, democratic arrangements effectively use distributed knowledge in two senses. As a way of eliciting information regarding the provision of public goods, democratic institutions afford a comparative advantage over either markets or technocratic, market-­mimicking institutional forms of the sort depicted in the literature on mechanism design. Viewed on their own terms, democratic arrangements also generate outcomes that are better in a broadly cognitive or epistemological sense.64 In a polity whose members unavoidably are characterized by an irreducible diversity of interests, commitments, and attachments, processes of political argument—­especially This conclusion deprives theorists like Mouffe (1999), who promote “agonistic” alternatives to deliberative conceptions of democracy, of their primary complaint—­namely, that democratic deliberation aims at consensus. Since our argument displaces the role of agreement in democratic politics, their flamboyant insistence on “agonism” has scant traction. 64   Our position resembles the view set out by Putnam (1991) and elaborated in various direct and indirect ways by Knight and Johnson (1996; 1999), Knight (2001), Misak (2000), Westbrook (2005), Kitcher (2006), Anderson (2006) and Talisse (2011). Insofar as we stress the importance of diversity, our view converges with the argument set forth by Page (2007). Put simply, he argues that “diversity trumps ability” in collective problem-­solving. Anderson (2006) sketches the affinities between Page’s work and pragmatist experimentalism. We find her views congenial just insofar as she, too, stresses the importance of dissent and contestation. 63  

151

C h ap te r 5

under initial conditions that approximate those required to characterize such debate as democratic deliberation—­provide the sort of competitive environment in which, minimally, participants can convey information about proposals and counterproposals that might otherwise remain unavailable. Let us unpack this second claim. What might be the real cognitive or epistemic benefits of democracy? In answering this question, it is helpful to begin by considering one potential answer to this question, the approach referred 65 to commonly as the “epistemic” interpretation of voting. This position has what for many is the virtue of reasserting solid moorings for democratic legitimacy in a notion of the popular or collective or general will. It proceeds from three basic premises: (i) A standard of correct decision exists independently of the political procedures used to identify it. (ii) Individuals vote not in order to express their preferences or tastes over outcomes but to register their beliefs or judgments regarding what, in the case at hand, actually constitutes the correct decision. (iii) When they participate in political decision-­making, individuals rely on evidence they acquire from the judgments rendered by others in order to adjust or revise their own beliefs about what counts as the 66 correct decision. Seen in these terms, epistemic views have an obvious Rousseauian resonance. And it hopefully is clear how this view trades on the fundamental premise that decisions emanating from democratic processes are justified insofar as they 67 correspond to some independent common good or popular will. This remains the case even if we concede that an “epistemic” interpretation of voting need not postulate a perfect relationship between electoral outcomes and the independently ascertainable popular will. It need only establish that electoral For a recent, seemingly comprehensive list of citations, see the bibliography in List and Goodin (2001). 66   Cohen (1986a, 34). See also Christiano (1996, 29–­34). Although we are here concerned with the epistemic interpretation of voting, Freeman (2000, 384–­89) sketches the “epistemic” interpretation of deliberative democracy. 67   “The epistemic account sees democratic participation by all as a method of discovering the common good. What is discovered by democratic participation is independent of that participation. The common good, in the epistemic account, is not merely whatever the people decide after discussion and voting; it is something independent that those processes attempt to uncover” (Christiano 1996, 35). Likewise, the “hallmark of the epistemic approach, in all its forms, is its fundamental premise that there exists some procedure-­independent fact of the matter as to what the best or right outcome is” (List and Goodin 2001, 280). 65  

152

Th e R o le o f Po litic a l A rg u me nt

outcomes provide more or less reliable, but in any case imperfect, evidence regarding the popular will and that they do so only over a specifiable, perhaps 68 fairly narrow range of issues. Traditionally the Condorcet jury theorem has seemed to be the analysis 69 of voting most congenial to an epistemic view of democracy. As its name implies, this work interprets elections on the model of jury deliberations. It starts from the premise that an electorate confronts some matter of common concern that is susceptible to correct or incorrect judgment. And it finds that, so long as each voter has a better than .50 chance of making a correct judgment, as the size of the electorate increases, so too does the probability that the collective judgment will be correct. In the limit, the probability of a correct collective decision approaches one. This result clearly will appeal to theorists like us, who advance claims that democracy can have cognitive or epistemic benefits for collective decision making. It nevertheless is important not to embrace the jury theorem results too hastily. At the very least, we should be clear about the circumstances under which the Condorcet results might lend support to such claims. In the first place, in standard formulations, they trade upon a set of fairly restrictive assumptions. For example, they presume that voters choose independently 70 of one another and that they vote sincerely. Moreover, further, more basic assumptions—­namely that voters confront a binary choice and that they meet a certain level of competence—­have prompted even more serious doubts regarding the broader relevance of the jury theorem to democratic theory. Champions of epistemic interpretations of democracy might well object that some relatively recent research goes a considerable distance toward miti71 gating such deep-­seated skepticism. This research extends the Condorcet jury theorem to cases where voters must choose from among more than two options. And it demonstrates that a variety of aggregation mechanisms—­both majoritarian schemes like plurality rule and positional methods such as the Borda count—­mimic the epistemic qualities Condorcet discovered in the jury setting. Which of the various available ways of counting votes is more reliable in epistemic terms depends on how competent the voters are, that is, on See, e.g., Coleman and Ferejohn (1986, 15–­19); Cohen (1986a, 29–­31). See, e.g., Cohen (1986a). For accessible introductions to the Condorcet jury theorem that relate it to broader themes in democratic theory, see Estlund et al. (1989); Young (1988); and List and Goodin (2001). 70   Recent research has explored how such assumptions might be relaxed. This technical literature is extensive. We will not explore it here. The bibliographies provided by List and Goodin (2001) and Anderson (2006) offer an auspicious starting point for the curious. 71   See List and Goodin (2001). 68   69  

153

C h ap te r 5

how much more likely they are to render a correct rather than an incorrect judgment. And under some scenarios, the required level of competence declines to significantly less than the .50 demanded in classic formulations of the Condorcet results. With a reasonably large electorate, it turns out in any case not only that various aggregation mechanisms are roughly equivalent in epistemic terms but that the chances for intransitive collective outcomes di72 minish rapidly as well. We thus seemingly arrive back in a position that those who interpret democracy in epistemic terms should find especially heartening. Yet, once more, it is important to temper our enthusiasm. There are at least three matters on which substantial room for skepticism remains. First, although the importance of voter competence is diminished in the research we just sketched, it remains relevant. And this has potentially troubling implications. Second, when in politics we make judgments or register beliefs, we tend not to do so issue by issue, or rather we tend render judgments over related sets of issues. And here we once again run into problems with aggregation. Finally, there is still the extremely contentious matter of whether, in most political disputes or disagreements, there is a fact of the matter waiting to be discovered. Put otherwise, while it may be possible to demonstrate that a variety of voting methods have similar epistemic qualities, all of this presupposes that political questions are susceptible to “correct” answers. It will help if we first offer an example that can help focus discussion of these matters. Consider what we take to be a relatively tractable domain of political decision making, namely macroeconomic policy. We say “relatively tractable” here just insofar as issues in this domain are more or less unconstrained by a penumbra of metaphysical or religious commitments. One might, by contrast, consider matters in what we might call, following T. S. Eliot, the domain of “birth and copulation and death.” In the American context, one need think only of debates surrounding the politics of abortion or of AIDS prevention or of stem-­cell research to see how disagreements over questions of inflation, unemployment, income growth, national debt, and income distribution seem relatively more tractable. Empirical and analytical research on the political economy of macroeconomic policy in the United States has generated a large and rather technical 73 literature. And this literature, despite common misperceptions, is one that List and Goodin (2001, 291–­95, 297–­304). Keech distills and dissects this literature in ways that make explicit connection to democratic theory from an avowedly “conservative” perspective (Keech 1995, 220–­21). As will become clear, we are indebted to his study. 72   73  

154

Th e R o le o f Po litic a l A rg u me nt

democratic theorists should find congenial. For it shows that although democratic politics has “costs,” these consist largely of inefficiencies that are the predictable, unavoidable, reasonable consequence of any effort to engage in 74 sustained political action in pursuit of common goals. So while democratic politics is susceptible to identifiable “risks and pitfalls,” political economists demonstrate that it “does not systematically lead to inferior macroeconomic policy.” This is not to say that democratic arrangements cannot contribute to “pathological” macroeconomic outcomes in the form of, for instance, hyperinflation or wildly unsustainable national debt. Rather, it is to say that in causal terms, democracy affords neither a necessary nor a sufficient condition 75 of such pathologies. This conclusion constitutes a powerful, if partial, response to critics of democracy who believe that modern electorates are so venal, so ignorant, or both that they not only tolerate but actively endorse government officials who purvey myopic or self-­serving economic policies. For present purposes, however, an additional theoretical point emerges from this research that is perhaps more important than this substantive observation. Any claim that democratic politics produces pathological macroeconomic policy presumes that those who advance it can specify a social welfare function by which to differentiate, in uncontroversial ways, between the unavoidable costs of democracy and obvious avoidable macroeconomic pathologies. Yet, the political econo76 mists highlight precisely how implausible that supposition must be. Precisely because any proposed macroeconomic goals or objectives (including those proposed in academic treatises) necessarily consist of revisable, contestable collective judgments that emerge from political processes, there exists no “Ar77 chimedean point” from which to assess them. Return now to the difficulties that even an extended Condorcetian analysis raises for epistemic interpretations of voting—­standards of voter competence, 74   This, Keech argues, is because macroeconomic policy-­making is a special instance of the sorts of principal-­agent problems that are endemic to politics. 75   The quoted phrases are from Keech (1995, 8, 16–­17, 50, 213f, 215). 76   Keech concludes that “democracy cannot be adequately understood within an optimizing framework, which depends on externally defined goals for objective functions.” This is because, on his account, “there is no authoritative or uncontestable welfare function that is not defined as part of a political process, and subject to revision in such a process” (Keech 1995, 212). There thus simply is no body of research in political economy that can be applied to society by would-­be social engineers. 77   See Keech (1995, 7, 17, 103–­4, 182). The recent high-­profile inquiry carried out by the French Commission on the Measurement of Economic Performance and Social Progress raises just this challenge to the practice of assessing economic performance in terms of gross domestic product. See Stiglitz, Sen, and Fitoussi (2010).

155

C h ap te r 5

problems of aggregation across interconnected judgments, and what, for want of a better term, we can call truth. We consider these briefly in turn. First, an epistemic interpretation of voting presumes that voters are competent, that they are “more likely, however slightly, to choose the ‘correct’ op78 tion, than any other” of the options on offer. Unfortunately, even supposing one endorses a view of average voter competence located somewhere between an overly naive reading of the Deweyan “faith” in the intelligence of the common man and an overly cynical embrace of Rousseau’s distrust of the “blind multitude,” this may well seem a heroic assumption. We will not explore it fully here. Rather, we wish simply to point out that while there are various ways to enhance the average competence of an electorate by partitioning issue domains or imposing qualifications on suffrage, these risk being quite 79 antidemocratic. For instance, one might seek to insulate entire domains of decision making from electoral pressures. A clear example of this partitioning strategy is the tendency to delegate authority over monetary policy to an “independent” central bank. This is a political decision. And it is one that minimally rests on a set of contestable political judgments concerning the threat democratic arrangements pose to sound monetary policy and the likelihood that the “independent” decision makers somehow will be less politically 80 motivated than the electorate as a whole. One may not find partitioning strategies of this sort objectionable per se to see that they, in any case, stand in 81 need of justification. This is even more transparent when we consider other, more direct strategies for enhancing political competence. Instead of insulating issue domains List and Goodin (2001, 300–­301). Jeremy Waldron makes this point in his contribution to the symposium on Condorcet and democratic theory (Estlund et al. 1989, 1324–­26). We borrow the distinction between “partitioning” and “qualification” strategies from Weale (1999, 148f). For a statement of the putative dangers that prompt such strategies, see Mill (1861/1991, 313): “The natural tendency of representative government, as of modern civilization, is towards collective mediocrity and this tendency is increased by all reductions and extensions of the franchise, their effect being to place the principal power in the hands of classes more and more below the highest level of instruction in the community.” This tendency notoriously prompted Mill to endorse the Hare system as an antidote. 80   On the politics of central bank decision-­making, see Blinder (1999) and Stiglitz (1998). In particular, Blinder offers insightful remarks on the internal decision-­making processes at the Federal Reserve, on how the bank’s mandate is interpreted, and the fact that the bank hardly is faulted when it is responsive to financial markets. Stiglitz is insightful regarding the neglect in central bank deliberations of input from unions and worker’s organizations. 81   Holmes (1988), for instance, both recommends the use of “gag rules” as a means of allowing democratic government to work effectively and insists that no such rule is self-­justifying. 78   79  

156

Th e R o le o f Po litic a l A rg u me nt

from putative popular threats, one might seek to differentiate competent from incompetent voters on the basis of one or another test or assessment. This is clearest when we consider suffrage qualifications. There now exists virtual consensus across countries that only mental deficiency (however arbitrarily defined) and age afford grounds for restricting the franchise. Indeed, there is no agreement across democracies about whether prison inmates, nonresidents, or 82 even noncitizens can rightly be excluded from the vote. Qualifications based on, for instance, age and mental deficiency reflect conventional and thereby morally arbitrary standards of competence. And anyone even vaguely familiar with the political disenfranchisement of African-­Americans, for instance, will grasp how judgments of competence grounded in social conventions can be used for extremely objectionable political purposes. The point is not that any judgment of political competence is inherently problematic but that any such 83 judgment readily lends itself to “invidious” uses. So any way of seeking to enhance average political competence of a general electorate by withholding the franchise from some constituency stands in need of justification. Since epistemic interpretations of democracy rest squarely on matters of voter competence, this problem is especially pressing for those who advance such an interpretation. Second, while the epistemic approach to voting seemingly garners support from the extended Condorcet analysis we just sketched, that analysis presumes voters render judgments on single issues rather than on sets of interrelated issues. The domain of macroeconomic policy suggests just how contestable an assumption that must be. For our judgments on one issue, say, inflation, will necessarily be intertwined with and constrain the judgments we coherently can make on other issues, such as unemployment levels or aggregate growth rates. And here, not surprisingly perhaps, an Arrow-­like impossibility result suggests that there exists no aggregation mechanism that will generate a collectively rational judgment on some set of issues while remaining mini84 mally responsive to the views the individuals who comprise the electorate. Given this result and the complexity of real-­world political decision making in democracies, Condorcet extensions confront the significant challenge of identifying a justifiable way of determining in any given circumstances how Massicotte, Blais, and Yoshinaka (2004, 15–­39). Smiley (1999) rightly stresses this point. 84   List and Pettit (2002). This result imposes rationality constraints on initial individual views and on outcomes as well as minimal normative conditions on the aggregation mechanism. The latter are “universal domain,” which requires that any individually rational view should have access to the voting arena; “anonymity,” which requires that all individuals are treated equally; and “systematicity,” which demands that no issue is accorded privileged treatment. 82   83  

157

C h ap te r 5

issues might be disentangled or, failing that, for determining which of the intertwined issues should be accorded privilege. Finally, extensions of Condorcet’s jury theorem pointedly broach what is perhaps the least plausible feature of epistemic interpretations of voting in particular and of democracy more generally. This is the basic assumption that in most political disputes there is some fact of the matter, independent of political processes themselves, the discovery of which can serve as a criterion of “correct” decisions. The various extensions of Condorcet that we report above assume that there is some independent fact of the matter even as they 85 remain agnostic about whether that assumption is warranted. Regardless of whether we ascribe the honorific label “truth” to such a fact of the matter, the assumption that any such procedure-­independent standard exists seems deeply implausible, in general; and research in even the particular, relatively tractable domain of macroeconomic policy simply accentuates that skepti86 cism. And so the burden remains on defenders of an epistemic interpretation of democracy to suggest what it might mean to claim that there is a general or collective or public will, independent of political procedures, that those procedures nevertheless help relevant political actors to identify. Although the Condorcet jury theorem and the related epistemic interpretation of democracy usually are invoked as support for claims about the epistemological benefits of democracy, they are not the most fruitful line of argument in this regard. A more promising approach would emphasize the ways in which democratic decision making draws on the benefits of social diversity. We can best see this by highlighting the role diversity plays in processes of competition.87 We considered two important forms of competition in chapter 3— ­competition within a market framework and competition as a mechanism of That research thus concludes that various aggregation mechanisms “are good truth trackers—­insofar, of course, as there are any ‘truths’ for politics to track at all” (List and Goodin 2001, 295). 86   “Growth, unemployment, and inflation are all legitimate and appropriate areas for macroeconomic policymaking. There are good reasons for the public and for politicians to want to increase growth and reduce unemployment and inflation. However, even if there were no trade-­offs between these and other goals, there are no obvious, authoritative numerical targets that public policy ought to aim for, with the possible exception of inflation. But the fact is that these three goals may conflict with one another as well as with other goals not discussed here. With no external authority for individual goals, or for objective functions that could define optimal combinations of goals, we must consider that public goals are to be formulated and reformulated in a public political process. Each formulation is likely to be subject to subsequent reformulation. This fluid dynamic is likely to be an inevitable feature of macroeconomic policymaking in a democracy” (Keech 1995, 117–­18). 87   Again, our views converge here with Page (2007) and Anderson (2006). 85  

158

Th e R o le o f Po litic a l A rg u me nt

institutional selection and change. Similar forms of competition—­like electoral competition or other forms of competition for political advantage—­ commonly are incorporated into virtually every conception of democracy.88 There also is the type of competition facilitated by the adversarial process in most legal systems. We rely on notions of competition when we invoke metaphors like the “marketplace of ideas” to describe forms of social inquiry. And we rely on the concept of competition when we make claims about the value of experimentation in science as well as other aspects of social life. What is striking about each of these examples is that they trade on much the same underlying logic. When effective, competition basically involves a process of experimentation and testing. The standard justifications for many of our basic legal, political, and economic institutions are grounded in the idea that better decisions follow from processes by which individuals contribute their individual knowledge to collective activity. In the process of contributing their knowledge, individuals test the merits of their own ideas and beliefs as well as the ideas and beliefs of others. In other words, competition—­whether economic, political, intellectual, or whatever—­affords precisely the challenge to our basic beliefs, preferences, and understandings to which we referred in chapter 2, when we were spelling out the pragmatist commitment to fallibilism. Such institutionalized testing produces a collective understanding that is superior to that previously held by any individual member of the group. With this superior knowledge base, better solutions to collective problems become possible. The outcome may be better goods and services, better political representation, better policies, better ideas, and so forth, depending on the particular case, but the process of achieving it is very much the same. And this is where diversity plays a fundamental role. For the various accounts of competition and its virtues invoke some process by which the distributed knowledge in a society is employed to achieve socially beneficially goals. Competition depends for its effectiveness fundamentally on two features: (1) diversity of inputs, whether these entrants to the competitive process be ideas, individual or collective agents, or institutions and practices, and (2) an institutionalized process of experimentation, inquiry, and testing. On the one hand, the diversity of initial entrants broadens and enhances the base on which experimentation and testing occurs. Here the diversity-­of-­inputs requirement recommends diversity of participation: the greater the diversity of participation by people of different backgrounds and experiences, the greater the range Lest we be accused of subscribing overly much to a narrowly Schumpeterian position, recall that Mill (1861/1991, 315) identifies the “function of antagonism” as central to the way political argument can operate to hold representatives to account. 88  

159

C h ap te r 5

of ideas and information contributed to the institutional process. Successful experimentation and testing requires variation, and this variation is a product of diversity of participation. On the other hand, processes of competitive testing must be systematized so that ideas and information are assessed in terms of their quality and not in terms of arbitrary factors or conditions exogenous to the aggregation process.89 The condition of effective experimentation and testing thus entails procedural constraints on extraneous factors that would distort the test of information and ideas. For example, testing procedures should not consider characteristics of an individual unrelated to the quality of his or her ideas and beliefs. Therefore, the overall effectiveness of the competitive mechanism is a function of our ability to identify and satisfy these conditions—­the requirements of diversity and effective experimentation and testing. Thus, to put it simply, one important argument for the value for diversity follows from the justification of the value of competition: given the appropriate procedural conditions, competition is most effective when participation is most diverse. We could support this claim about diversity and competition by reference to the justifications for a number of basic institutional forms.90 Although these different institutional forms vary markedly in terms of the ways in which they organize and structure competition, they share the basic logic of diversity and effective testing in their underlying institutional processes. Here it is important for us to note one additional feature of this logic that is especially relevant to our ongoing emphasis on conditions for effective institutional performance. Obviously, diversity alone will not guarantee effective competition. These interactions among diverse constituencies must take place in the appropriate institutional setting, and as we pointed out earlier, even allegedly “minimal” defenses of democracy trade on surprisingly robust conditions of freedom and equality. Those taking part in experimental processes generally and democratic ones more specifically must be free to (1) contribute their ideas and beliefs on the input side of the process and (2) make uncoerced choices as participants in determining the outcomes of the process. Similarly, the argument for competition as an effective process of experimentation and testing entails constraints on the use of power to perpetuate inequality within the sphere of competition. 89   This discussion clearly simplifies what is a complex process. On some of the complexities that we neglect, see Kitcher (2006) and especially his discussion of what he terms the Inquiry-­and-­ Information System upon which a democracy relies to generate, authenticate, and disseminate information to potential users. 90   Knight and Johnson (1996); Knight (2001); Epstein, Knight, and Martin (2003).

160

Th e R o le o f Po litic a l A rg u me nt

We return to this issue in later chapters where we elaborate the importance of free and equal participation for democratic decision making. For now, we do want to reiterate and emphasize the basic point that this condition of participation extends beyond democracy to all forms of institutional arrangements grounded in the logic of competition. The overall effectiveness of each of these institutional forms depends on our ability to satisfy conditions such as the free-­and-­equal-­participation requirement. Democratic decision making structures competition in such a way as to take special advantage of the benefits of diversity. It does so primarily in terms of the particular way in which it employs distributed knowledge in a society. Unlike decentralized approaches that test a particular idea or good through an ongoing set of small-­n interactions, democratic decision making constructs, in principle, a more comprehensive test for any idea or proposal. The collective nature of the decision-­making process forces any idea to face the scrutiny of an expansive array of perspectives while simultaneously giving each perspective the opportunity to challenge any other feasible alternative. In doing so, democratic institutions attempt to guarantee that resulting collective decisions will be the product of a thorough and comprehensive testing process. In this way, they seek to give social diversity the maximal effect on the decision-­making process among the feasible set of institutional alternatives.91 To the extent that diversity can enhance the epistemological basis of social coordination, then this is an asset in the case for democracy. It is a direct product of the salutary effect of political argument on democratic decision making.

VIII.  Reflexivity What do we mean when we say that democratic institutions operate in a reflexive manner? The reflexivity of democratic arrangements derives from the fact that political argument—­again, increasingly so to the extent that it transpires under the appropriate conditions—­requires relevant parties to assert, defend, and revise their own views and to entertain, challenge, or accept those of others. It derives, in other words, from ongoing disagreement and conflict. In this respect, democratic institutions clearly differ from market interactions, 91   However, we reemphasize that this claim alone does not support the view that democratic decision-­making is always the most appropriate mechanism for institutionalizing competition as a coordinating device. This view, which would accord democracy something close to a first-­order priority among social institutions, is a strategy that we have already rejected. The epistemological effect of any institution is just one factor, albeit an important one, that goes into the consideration of the appropriateness for particular coordination problems.

161

C h ap te r 5

where parties simply “exit” to seek more favorable terms of exchange elsewhere in the case of disagreement and simply trade without concern for the efficiency or distributive features of aggregate outcomes in cases of agreement. In this sense, too, the reflexivity of democratic arrangements resembles judicial decision making where (in the ideal) arguments are exchanged under established, fair conditions. But in legal settings, too, the parties—­judges, juries, litigants—­typically are concerned primarily with settling the case at hand rather than either with the law as an entity or structure writ large or with the way established rules will impact the outcome of future cases. The key here is that market interaction and legal decision making are decentralized in the sense that those party to any particular interaction are charged neither with reviewing or monitoring in an ongoing way the conditions under which institutions of exchange and legal disputation themselves operate nor with assessing the consequences of how broader institutional arrangements operate. This hopefully is clear from our discussion of markets. We argue in the next chapter that it is true as well of judicial decision making. It is just here that democratic institutions enable a level of reflexivity unavailable in other institutional forms. For political argument not only allows members of a democratic polity to collectively revisit past substantive decisions, it also allows them to collectively reconsider and revise the terms of their ongoing interactions. Democratic politics, in other words, structures disagreement not just over policy decisions but also over the operation and reform of social and political institutions. Moreover, because democratic arrangements structure political argument so as to allow the emergence of new ideas, perspectives, and interests—­and, thereby, new constituencies and oppositions—­this reflexivity has an endogenous, dynamic dimension. This propensity to inclusiveness is an attractive normative feature of democratic arrangements. But it raises a potential problem when we recall the practical effect we previously ascribed to political argument—­namely, that by establishing shared understanding of the dimensions of political conflict, it can mitigate the sort of instability that social choice theorists show threatens outcomes generated by voting. Given this, an apparent tension emerges within democratic institutions between their normatively attractive inclusiveness and their hoped-­for practical effect of inducing stable outcomes.92 The rise and activation of new constituencies, and of the interests, attachments, and commitments to which they would give voice, would surely threaten already settled political understandings. This tension, however, is much less acute than it might at first appear. In practical terms it is, as we argued above, Knight and Johnson (1994).

92  

162

Th e R o le o f Po litic a l A rg u me nt

less useful to characterize political debate as inducing agreement than to see it as structuring disagreement. It is a mistake to focus on the agreement here. We must instead keep our eye on the conflict. That is because even when political argument succeeds in establishing a shared assessment of the relevant dimensions of conflict, it leaves considerable space for at least two sorts of persistent disagreement. First, the fact that members of some population agree on the dimensionality of conflict in no way entails that they agree substantively on any particular issue. So, for instance, the fact that the parties to some controversy might agree that what is at stake is a matter of national defense rather than, say, facilitating interstate commerce (e.g., as in the argument over the U.S. interstate highway system) in no way means they agree regarding how much to spend on that issue. Indeed, it is entirely possible that clarifying the dimensions of conflict might sharpen disagreement along that dimension. Second, the persistence of substantive disagreement does not exhaust the room for conflict in democratic arrangements. That is because we have for some time known that aggregate stability presupposes far less than consensus over even the dimensions of conflict. As many as 25 to 30 percent of a population can dissent from a commonly held view regarding what is at stake in any political conflict without subverting the possibility that voting on the matter will generate a noncyclical collective outcome.93 The grounds for our claim that democracy is reflexive, then, emerge when we see how, when it is successful, it structures disagreement and, thereby, potentially fosters still further disagreement. The irreducible diversity of interests, commitments, and attachments that gives rise to politics in the first place means that members of a population have an interest in monitoring the institutions that coordinate their ongoing interactions. Because democratic institutions afford space for the two kinds of disagreement we identify, they also provide the agents who populate them with ample opportunity to challenge the terms of their ongoing interactions. There is little reason to suspect that such space will remain vacant for long. Those individuals and groups who lose in political competition, or who are newly emergent, or who have remained in opposition become the agents of democratic reflexivity.

IX.  A Second-­Order Priority Several important things follow from the conception of democratic decision making that we offer here. First, nothing that we have said about the benefits Niemi (1969).

93  

163

C h ap te r 5

of democracy supports the view that it has a presumptive privilege as a first-­ order mechanism for coordinating social interactions. In many domains of social life, the reflexivity inherent in democracy will make it inappropriate—­in the sense of being cumbersome and ineffective—­as a first-­order mechanism for coordinating ongoing social interaction. Conversely, it is a virtue of competing institutional forms, such as markets and common law adjudication, that they allow us to more effectively coordinate social interaction precisely because they are not reflexive in the sense we indicate. In circumstances in which the requirements of coordination recommend nonreflexive choice, institutions like the market will be more appropriate as a first-­order institutional mechanism. Second, democracy nevertheless enjoys a second-­order priority precisely because it allows constituencies to monitor and maintain the conditions presupposed by effective operation of prevailing institutional arrangements—­ including, crucially, the conditions that govern its own effective operation. Democratic decision making provides a comprehensive framework for coordinating these fundamental tasks. Of course, at this point, our claim about the priority of democracy rests on a claim about whether and, if so, how the conditions necessary for the effective operation of democratic institutions can be established and maintained. And this obviously raises the same issues of robustness that undermined claims of priority for decentralized mechanisms. We set those issues aside for the moment and return to them in chapters 7 and 8, where we discuss these conditions in detail. Third, it also follows that the process by which constituencies must decide which institutional forms they will rely on and for what purposes is unavoidably political. Here there is room, at least, for disagreement over both outcomes and initial conditions. With markets, for example, efficiency competes with equity, freedom, and other considerations as a defensible criterion for assessing aggregate outcomes. Likewise, since actual conditions on the ground will only ever approximate the ideal conditions presupposed by the effective operation of markets, legitimate disagreement will persist over whether particular goods or services—­say, sex or reproductive capacity—­ought to be available on markets in the first place.94 Markets hardly are unique in these respects. The same issues will arise with alternative institutional forms, including democratic ones, with the consequence that social actors cannot reasonably expect to discover or construct a neutral, apolitical metric that might be applied when making such decisions. Even under ideal conditions, there is no reason to expect consensus on such matters as how to weight competing Satz (1992; 1995).

94  

164

Th e R o le o f Po litic a l A rg u me nt

normative criteria or how to judge whether empirical conditions sufficiently approach those necessary to ensure normatively attractive outcomes. The fact of pluralism we recognized in the first chapter suggests that there will always be space for legitimate disagreement about such matters. Fourth, skeptical readers may find our emphasis on politics and disagreement an unsatisfactory feature of a pragmatist justification of democracy. They may question why pragmatists would focus on possibility results for ideal-­typical institutions rather than on the actual effects of actually existing institutions operating within a context that departs from optimal conditions.95 They therefore might recommend that we instead examine actual conditions in hopes of opening avenues for further theoretical analysis, something akin to comparative-­static analysis of markets for other institutional alternatives. They might reason that by doing so, we might find, for example, that markets are more robust than democratic decision making to in the face of deviations from optimal operating conditions.96 And this could give greater support to claims about the appropriateness of using markets to coordinate various social interactions. Such a skeptical response misunderstands two important aspects of our argument. First, even if we could conduct such robustness tests at the theoretical level, the results would only bear on the choices for first-­order tasks, on questions such as whether we should rely on markets or bureaucracy to coordinate a specific domain of activity. It would not affect the argument about the second-­order priority of democratic decision making, an argument that rests crucially on an assessment of what tasks institutions can and cannot accomplish when they are operating effectively. Second, and more important, such skeptics err in thinking that questions about robustness can be resolved at the theoretical level. Theoretical analysis can make important contributions to our understanding of issues of institutional performance, but alone it cannot provide definitive answers to questions of actual effect.97 Such answers can only come from the cumulative experience of using the various institutions at our disposal. And this further highlights the importance of institutional

Dixit (2009, 8) nicely articulates something close to the view we endorse. Like him, we are comparing and exploring “ideal” or “first best” models because it helps sharpen our argument. We also are well aware that in fact, “Everything is ‘second-­best’ at best, subject to numerous constraints of information, incentives, commitment, and rules of the political game.” 96   As we already have noted, there is good reason to think that even slight divergence from the initial conditions required for textbook analysis of efficient markets raises considerable doubt about the robustness of that account (Stiglitz 2002). 97   Posner (2003). 95  

165

C h ap te r 5

experimentalism in generating knowledge about the relationship between institutional performance and changing social conditions. This, in turn, supports, finally, our specific claim about the priority of democracy. Reflexivity is a crucial feature of experimental procedures. We trust or value experimental outcomes insofar as they emerge under proper conditions, and we expend much effort in monitoring those conditions. The priority of democratic institutional arrangements emerges because they embody the sorts of reflexivity needed to monitor the conditions under which alternative institutions and practices, including democratic procedures themselves, generate normatively attractive outcomes. But in keeping with our commitment to a plurality of institutional forms, it is important to highlight one, hopefully obvious implication of our discussion to this point: the reflexivity of democratic decision making will often allow constituencies to decide to coordinate their ongoing social interactions across expansive domains through various nondemocratic institutional forms.

166

Chapter 6 Refining Reflexivity

I.  Reflexivity and the Organization of Competition Of the three kinds of consequences political argument has for democratic decision making, reflexivity is the most important for the justification of a second-­order priority of democracy. All three factors go some distance toward satisfying the burden of justification for democratic institutions. The effects of political argument on the cognitive structure of the issues under consideration, the constraints on dimensionality, are a necessary response to general challenges to the normative significance of democratic decisions. Social choice results illuminate inherent problems of instability and ambiguity in voting that, in turn, highlight significant opportunities for strategic manipulation and, thus, undermine the common claims about the normative legitimacy of democracy. This challenge is exacerbated when advocates of democracy rest their justifications on overly optimistic and unrealizable claims about the capacity of democratic institutions to facilitate the general will or other forms of principled consensus. By reconceptualizing the role of political argument as one of structuring disagreement, as opposed to establishing agreement and consensus, we provide the basis for a practical counter to the social choice challenge and, in doing so, maintain the possibility of mounting a normative theory of democratic legitimacy. The other two effects of political argument enhance the quality of voting outcomes and contribute to the effectiveness of democratic institutions. The epistemological effects of diversity on democratic decision making rely on an interrelated structure of debate and voting that facilitates maximal effect to the distributed knowledge in society. Although they are not the only effective means of coordinating distributed knowledge, democratic institutions structure competition and choice in such a centralized way that it offers the best way of aggregating information and knowledge for certain categories of interaction. And this is where reflexivity becomes central to our argument. For it is in the reflexive structure of democratic decision making that we see the distinctive benefits of this centralized form of coordination.

C h ap te r 6

As a way of clarifying the role of reflexivity within democratic institutions, it is helpful to compare the ways in which competition and choice interact in markets and in democratic decision making. In doing so, we can also begin to highlight the primary reason why we think democratic institutions perform better than markets in satisfying the second-­order requirements of institutional choice. In their early demonstration that equilibria exist in competitive economies, Arrow and Debreu characterized the nature of market competition as follows: “It was assumed that each consumer acts so as to maximize his utility, each producer acts so as to maximize his profit, and perfect competition prevails, in the sense that each producer and consumer regards the prices paid and received as independent of his own choices.”1 In other words, “individuals and firms take prices as given.”2 In this sense, choice in the market is parametric in just the way we stressed in chapter 3. In conditions of market competition individual participants act independently in that they make choices based exclusively on their preferences and the information provided by the price system that market institutions embody. They reasonably expect that nothing about their choices will influence or be influenced by the choices of any other actor. Competition is structured as a network of small-­n exchanges. Effective choice, in the sense of maximizing the net benefits of such an exchange, is nonreflexive. Effectiveness is a function of an actor’s ability to focus exclusively on the information necessary to make the transaction at hand. It does not require, and in fact can be undermined by, reflection on past choices and on mistaken notions that one’s choices can have broader effects on the economy as a whole. In this last sense, effective market choice depends upon confident—­unreflective—­reliance that the market is working effectively and thus eschews broader considerations about the overall effectiveness of the market as an institution. One of the major virtues of the market is that it simplifies the requirements of effective economic choice. Competition within democratic institutions is organized quite differently. There, competition is structured as a centralized process. Individual participants interact with all others in the group and must reach a decision as a collectivity. Choice within democratic institutions is strategic in the sense that the choices participants make are interdependent. That is, we make our decisions with the expectation that our choices can affect the choices of other actors and vice versa. This interdependence opens up possibilities for significant debate and persuasion prior to the ultimate act of voting. Effective democratic

Arrow and Debreu (1954, 265). Arrow (1994, 4).

1   2  

168

R e finin g R e fle xivity

choice is reflexive in large measure because it is strategic in this sense.3 It commonly requires consideration of a broad range of implications of any individual choice and frequently recommends reconsideration of past decisions. While reflexivity can undermine success in the market, it often becomes essential to effectiveness in a democratic environment. It is easy to see why reflexivity is important for the tasks of institutional choice. As we set out in the introduction, the most encompassing normative task related to institutional choice is to identify some institutional arrangement (or set of arrangements) capable of allowing relevant constituencies (1) to coordinate effective institutional experimentation, (2) to monitor and assess effective institutional performance for the range of institutions available in any society, and most importantly, (3) to monitor and assess its own ongoing performance. Each of the specific tasks requires a framework in which populations can collectively assess the interrelationship of conditions and consequences over time. Such assessments present the possibilities for conflict at various stages of the process. For example, we might anticipate conflict over the ways in which we interpret the data about the relationship between conditions and consequences. We might also anticipate conflict over the appropriate criteria of assessment and conflict over the extent to which the criteria are satisfied in any particular case. Examples such as these highlight one important fact about institutional choice: there is an unavoidable political dimension to these decisions. Due to its emphasis on reflexivity, democracy is uniquely structured to provide, under the proper conditions, an effective framework for these complex and often contentious tasks. Our argument for the second-­order priority of democracy rests on the comparative claim that democratic decision making has properties other institutional forms either lack or possess to a lesser degree.4 One potential challenge 3   See Elster (1979) for a now standard explication of the difference between parametric and strategic action. 4   Pragmatists will make initial institutional comparisons in terms of effective performance. This requires us to assess the alternatives as characterized in their most ideal form, which leads us to make the following distinction between two types of criticisms. One type, empirical criticism, challenges the proffered justification on the grounds that the facts of actual practice differ from the institutional practice as characterized by the theoretical justification. A second type, institutional criticism, challenges the appropriateness of an institution for a particular task on the grounds that the institution, when performing optimally, is not the best mechanism for undertaking the task. While both types of criticism are relevant to pragmatist considerations, it is the second type that we emphasize here because it is more decisive for institutional choice. Indeed, if Rodrik (2007, 162) is correct when he remarks that the “institutional arrangements that we observe in operation today, varied as they are, themselves constitute a subset of the full range of potential institutional possibilities,” then empirical criticism, while important, will remain

169

C h ap te r 6

to our argument is that some other institutional form can reflexively monitor and assess the conditions of effective institutional performance in ways comparable or superior to those we ascribe to democratic mechanisms. Any such comparison would primarily focus on the different forms of communication facilitated by different types of institutions. In this way, we can make precise our rejection of the market for these purposes. Democratic institutions are most effective when decision making is properly strategic and reflective; markets are most effective when decision making is parametric and unreflective. These differences affect the types of communication that are fostered by the two institutional arrangements. The crux of our argument is that markets do not facilitate the types of communication necessary for the tasks of institutional assessment and monitoring. In perfect markets, prices convey all of the information necessary for effective choice. Communication between actors is about particular exchanges and not about the general functioning of the market itself. As communication extends to other types of issues, markets tend toward imperfection and inefficiency. Now, an advocate of markets might object at this point on the grounds that we have overstated the differences in the way that competition is organized in the two environments. This objection would be based on the idea that there is, in practice, no sharp line between the centralized and decentralized environments. Given the lack of a sharp distinction, so the argument would go, market advocates might press for markets on the grounds that decentralized mechanisms provide other benefits that offset the benefits of reflexivity.5 Although we disagree with the premise of the objection, we would insist nonetheless that this debate also involves a political question that is best taken up democratically. For it is just the sort of question about whether a certain set of conditions obtain for one institutional form or another that lends itself to the reflective nature of the democratic mechanism. In the next sections of this chapter, we consider the objection as it applies to other centralized mechanisms that might embody the necessary reflexivity importantly incomplete. It will require the sort of theoretical examination we pursue here insofar as we need to explore possibilities. 5   One might infer such a complaint from correct observations as, for example, “Institutions are rarely either private or public—­‘the market’ or ‘the state’” (Ostrom 1990, 14). Likewise, “a market economy relies on a wide array of nonmarket institutions. . . . Once these institutions are accepted as part of a market based economy, traditional dichotomies between market and state or laissez faire and intervention begin to make less sense” (Rodrik 2007, 160–­61). As pragmatists, we agree that it is important to resist the temptation to inflate the distinction between state and market (and related distinctions) into a conversation-­stopping dichotomy. It nevertheless is important to inquiry that we draw and use distinctions of the sort we presuppose here.

170

R e finin g R e fle xivity

to handle the second-­order institutional tasks. We think this raises a more serious challenge to our basic thesis. We consider three such alternatives: courts and judicial decision making, bureaucracy, and a hybrid form that combines informal norms within formal institutional arrangements. We take these up in turn.

II.  Courts To what extent can courts adequately undertake the tasks of monitoring conditions and assessing consequences in the ongoing process of effective institutional performance? To fairly answer this question, we must assess the courts, as we do with the market and democracy, as characterized in their most favorable, yet plausible, justification.6 Hart offers such an account, a positive conception of the rule of law that highlights the distinctive institutional features of judicial decision making.7 On Hart’s account, both the role of the courts and the standards of their effective operation are established by what he calls the rules of recognition—­ the set of secondary rules that dictate how the primary rules, the laws, are to be implemented, administered, and changed.8 The rules of recognition distribute authority among various legal and political institutions over the different tasks involved in maintaining the legal system. Within this distribution of labor, courts are primarily responsible for adjudicating controversies over how law is to be applied in particular cases. Rules of recognition are effective just so long as members of the relevant community voluntarily accept them.9 Hart thus links effective institutional performance to shared views about the legitimacy of the various institutions. For Hart, courts—­typically in the form of unelected judges—­derive legitimacy from the particular nature of their most common form of decision making, the application of primary rules to new fact situations. He acknowledges that there is always room for discretion in the application of rules, but he insists that rules remain crucial even in such cases. He argues that discretionary judgments “must not disguise the fact that both the framework within which Like other institutional forms, the actual practice of judicial decision-­making has often deviated from such justifications (Epstein and Knight 1998). Such issues are of secondary importance here because we are making a comparison of the relative merits of different institutional forms when they are operating effectively. 7   Hart (1994). 8   Hart (1994, 94–­96). 9   Hart (1994, 114–­115). 6  

171

C h ap te r 6

they take place and their chief end-­product is one of general rules. These are rules the application of which individuals can see for themselves in case after case, without further recourse to official direction or discretion.”10 Hart thus clearly emphasizes the “framework” in which judges act, highlighting the importance of institutional procedures as a precondition for effective performance. Although courts rely on numerous unique procedures vis-­à-­vis other institutional forms, most important here are those that define (1) the scope of the judicial process and (2) the nature of acceptable judicial argument. Both are intended to maintain the rule-­application focus of judicial decision making. The scope of judicial decision making is limited in at least two important ways. First, the agenda-­setting power in judicial institutions is largely exogenous. While judges often are authorized to refuse to hear various kinds of cases, they typically cannot initiate causes of action. They must instead rely on interested parties to bring controversies before the court. Second, participation in the process is limited in most jurisdictions to those actors who satisfy restrictive definitions of who constitutes an interested party. Rules of standing limit participation to those who can demonstrate that they are directly affected by the controversy at hand. Although standing may be defined differently across jurisdictions, the accessibility of participation under even the most expansive definitions restricts the scope of relevant interests. In addition to constraining the scope of decision making, judicial procedures also restrict the types of claims that legitimately can influence court decisions. Hart’s justification of the court as an institution rests on the idea that judges will render impartial decisions.11 This criterion restricts both the types of arguments that parties can employ when pressing claims and the ways judges can justify their decisions. According to Hart, an impartial judge ideally will justify her decision in terms of predetermined rules. In response to the criticism that citing precedent merely rationalizes decisions that are made on other grounds, he notes that “some judicial decisions may be like this, but it is surely evident that for the most part decisions, like the chess-­player’s moves, are reached either by genuine effort to conform to rules consciously taken as guiding standards of decision or, if intuitively reached, are justified by rules which the judge was antecedently disposed to observe and whose relevance to the case in hand would generally be acknowledged.”12 Hart (1994, 136). Hart (1994, 207). 12   Hart (1994, 141). For some of the complications involved in the ways political actors use principle and precedent, see Calvert and Johnson (1999) and Murphy et al. (2005). 10   11  

172

R e finin g R e fle xivity

This characterization of judicial institutions clearly is incomplete. It is sufficient, however, for the purpose of assessing whether courts might effectively monitor institutional performance. It is useful to distinguish three features of this task. The first two involve judicial oversight of other institutions, specifically monitoring conditions of effective performance and assessing the ongoing consequences of institutional choice. The third involves self-­assessment, that is, monitoring conditions and assessing outcomes related to their own performance. Courts arguably might afford the appropriate venue for monitoring conditions presupposed by other institutions. This, for example, is what constitutional courts normally do when they address questions of constitutional rights. Rights can be conceptualized as institutional protections for individual activities and prerogatives in various institutional contexts. They enable individuals or groups to claim that their participation is being infringed in some important way. In such instances, courts must ascertain whether or not effective participation, as constitutionally defined, is being adversely affected in the particular case. Where the rules defining effective participation are clear, courts can normally do an adequate job of monitoring such conditions. As Hart suggests, courts are best at applying predetermined rules to particular sets of facts. However, courts may not be the best or only mechanism for monitoring conditions. Waldron, for instance, argues that the protection of basic rights is too important to be left to unelected judges.13 He insists that when issues of rights are at stake in a democratic society, the right to participate in the resolution of the controversy is of prime importance. There is room for disagreement here. Waldron’s claim presumes a prior definition of how we define who is affected when some basic rights are infringed. If we define the impacted group narrowly to include only those who are directly affected by the infringement under question, judicial decision making may, given our characterization of such processes, allow for adequate resolution of particular cases. If, on the other hand, we define rights controversies broadly, arguing that such claims are always challenges to the basic definition of effective participation, then the scope of the challenge is much more expansive and the courts may prove inadequate to the task.14 Waldron (1999). Whether courts can adequately handle these more expansive rights claims is a question that must be answered in an institutional forum appropriate to the task of assessing effective institutional performance. As we argue below, that forum is not the courts but rather the democratic process. 13   14  

173

C h ap te r 6

When we move on to consider how courts might handle the task of assessing institutional outcomes, we encounter the real weakness with claims that courts can satisfy the requirements of institutional effectiveness. While courts might be able to adequately monitor conditions, they are not well suited to assess the consequences of institutional choice. The relevant questions in assessing effective performance are: “to what extent is the institution satisfying its intended goals?” and “would another institutional form do a more effective job?” These are general questions the answers to which involve basic considerations of institutional goals as well as criteria of effectiveness. The institutional restrictions on judicial decision making render courts inadequate for the effective consideration of these questions. Institutional assessment consists in ongoing processes of experimentation. These processes must take account of the full array of interests, attachments, and commitments in a society by affording them equal opportunity of access to political influence.15 If the range of interests is restricted arbitrarily, the outcome of assessment processes correspondingly will be suspect. The constraints operative in judicial decision making—­constraints that restrict arguments to invoking and applying predetermined rules—­preclude consideration of many issues relevant to the assessment of institutional consequences.16 Similarly, restrictions on accessibility to judicial procedures circumscribe the participation of potentially relevant parties. Unlike controversies over rights violations that may be conceptualized in terms of the specifics of a particular case, assessment questions generally affect broad segments, if not all members, of the community. Therefore, constraints on participation will tend to overly restrict the range of relevant factors taken into account in the assessment of institutional consequences. Most importantly, courts are restricted in the types of remedies they typically seek to implement. These normally consist of sanctions aimed at forcing individuals to comply with the dictates of predetermined rules. Because judicial decision making consists primarily in applying rules, courts are ill-­equipped to pose broadly normative questions about a particular policy or institution. Simply put, when effectively undertaking their legitimate role, courts are not reflexive in the sense that questions regarding the general validity of either institutional forms or procedural constraints remain beyond their purview. Their basic prerogatives are set by either constitutional or legislative pronouncements. Knight and Johnson (1996; 1997); Knight (2001). If we were to insist that processes of institutional experimentation operate under such constraints, we would simply contribute to the juridicalization or constitutionalization of political argument that has gained force in recent years. For a discussion of some of the practical effects of this process, see Stone-­Sweet (2000). 15   16  

174

R e finin g R e fle xivity

This reflexivity deficit is even more pronounced when we consider whether courts might satisfy the requirements of self-­monitoring and self-­assessment. The factors that prevent courts from effectively assessing the consequences produced by other institutions are even more relevant to self-­assessment considerations. Restrictions on their agenda and on standing to participate diminish the likelihood that courts will be challenged to examine their own effectiveness. The same barriers restrict the range of interests that might influence such an assessment. Finally, the emphasis on rule application constrains the methods by which courts might address problems within the judicial system. Together these institutional restrictions undermine the ability of the courts to be a self-­regulating institution. Put simply, when the court is effectively performing its first-­order functions, it must rely on other institutions for an adequate assessment of its own performance. In summary, courts are appropriate for resolving particular controversies (including particular violations of institutional conditions) but are inadequate to the task of ultimately assessing the requirements of general institutional performance (the reconsideration of conditions as well as the assessment of institutional outcomes).17 To the extent that courts might be remodeled to better undertake these more general tasks, they would have to be refashioned to look more like institutions of democratic decision making.

III.  Bureaucracy Bureaucracies have become an increasingly significant form of social coordination in most societies. In the face of ever growing complexity, bureaucrats Advocates of judicial review may object that our conclusion about the superiority of democratic decision-­making is a function of a mis-­specified comparison. They might argue that the appropriate focus is not on the operation of an individual court but rather on the system of courts that make up the judicial branch of government. We lack the space here to fully address this challenge. However, there are at least two reasons why we think that a comparison with a system of courts would lead to the same conclusion. First, the constraints on agenda setting, participation, and types of legitimate argument that affect the ability of an individual court to adequately assess institutional performance apply across the board to all of the courts in a judicial system. Second, even in an interlocking system of courts, no one is given overall responsibility, as part of its judicial function, for assessing and monitoring the system. For example, the chief justice of the U. S. Supreme Court is considered the chief operating officer of the federal system and, in that capacity, provides Congress with an annual report about the workings of the judicial branch. However, this responsibility is an administrative task over and above his primary responsibilities as a member of the court. In addition, the task extends only to the reporting function, with the responsibility to act on changes and reforms to the judicial system resting with Congress. On this last point, see Office of Judges Programs (2003). 17  

175

C h ap te r 6

are called upon to play important roles in policy development and implementation as well as in the regulation of various forms of public and private activity. They are generally organized hierarchically, with varying schemes for the distribution of authority among the various actors in the system. Envisioned as a centralized authority, bureaucracy is the primary critical target of advocates of decentralization like Hayek and Coase. For such critics, bureaucracy is the central organizing principle of state activity. But this account offers too simplistic a conception of how bureaucracies might coordinate ongoing social interactions. The Hayekian account limits our understanding of bureaucracies to what we might imagine them to be in an authoritarian state. Frankly, this would be too easy a target for our assessment because it fails to take adequate account of how bureaucracies can and often do work within a society characterized by markets and democratic decision making. This is the most interesting and, in many ways, most challenging case, and so it is the conception that we address here. A complete analysis would require a consideration of the circumstances and types of social interaction that would recommend bureaucracy as the best way to coordinate our affairs. This would inevitably introduce important questions about the role of specialization and expertise in public and private decision making. These are difficult and, at times, vexing questions for democratic theorists. From a pragmatic perspective, they do not lend themselves to easy, a priori answers. As we have argued above, these are questions that must be answered by the relevant constituencies themselves based on their best understanding of their own particular needs and interests. Furthermore, too detailed a discussion would lead us far away from what is the primary issue at hand: can bureaucracy, when operating effectively, facilitate the reflexivity in decision making necessary to perform the second-­order tasks we have assigned to democracy? Therefore, we only consider the substantive functions of bureaucracy to the extent necessary to address this question. To develop an answer, we focus on the two primary ways that bureaucratic authority is organized in a democratic society.18 Both conceptions share common features of hierarchical internal organization with various levels of decision making and review. Ultimate decision-­making authority within the organization rests with either a single head of the organization or a small executive committee. But the models differ in terms of the attribution of ultimate authority vis-­à-­vis democratic decision makers. The first and most common contemporary formulation of this relationship is the principal-­agent model.19 On this Here we draw mainly on G. Miller (1992) and Knight and Miller (2007). Miller (2005); Przeworski (2003, sec. 2).

18   19  

176

R e finin g R e fle xivity

conception, bureaucrats serve mainly as the agents of democratic representatives. This is a conception that gained currency in the study of U.S. government bureaucracies in the 1980s. Proponents of this view argued that results borrowed from principal-­agency theory in economics are applicable to the relationship between congressional oversight committees and bureaucracies. Weingast pointed out that the information asymmetry between congressional committees and bureaucratic agencies is analogous in all analytical respects to economic relationships like that between the insurance companies and their potential customers.20 The problem here is one of moral hazard. The interests of the company and the potential customers diverge in important ways. To the extent that customers can act in a way harmful to the interests of the company, the insurance company has a reason to be concerned about their actions. Yet the company does not have access to the information necessary to perfectly control its customers. Weingast argued that congressional oversight committees have the same relationship of moral hazard with government bureaucracies. The solution to the oversight committee’s problem is the use of incentives that align the interests of bureaucrats with the (reelection) interests of the members of the committee.21 The principal-­agent logic offers the following answer to the question of the proper role of bureaucrats in a democratic society. A central claim of principal-­agency theory is that, appearances to the contrary, bureaucrats are rational agents who make their decisions in response to incentives that are set by elected officials. On this account, public concern about out-­of-­control bureaucracy is overstated; bureaucrats are, in fact, largely responsive to elected officials. Here the principal-­agent theory has both explanatory and normative implications. So, the argument goes, shifts in bureaucratic decisions can be traced back to shifts in the preferences of elected officials, demonstrating a responsiveness to democratic processes that helped make elections meaningful.22 Thus, principal-­agency theory explains how bureaucrats can be controlled by elected officials. Furthermore, advocates of this approach claim that it is the normative responsibility of elected officials to control bureaucrats. They argue that the U.S. Constitution has guaranteed effective control of elected officials by the general citizenry, so that elected officials can be counted on to represent those interests authoritatively, and empowered to represent those interests effectively vis-­à-­vis the bureaucracy. From this, it then

Weingast (1984). Weingast and Moran (1983). 22   McCubbins, Noll, and Weingast (1987). 20   21  

177

C h ap te r 6

follows that the normative solution to the problem of democratic legitimacy is for the bureaucrats to carry out the policy intentions of elected officials. Contrast the principal-­agent model with an alternative conception of bureaucracies based on the notion of trusteeship, an idea commonly associated with Schelling.23 This alternative conception grants bureaucrats an autonomous role vis-­à-­vis democratic representatives. One justification of this autonomy can be the possible value of experts in the policy process. For example, the scientists who work at various government agencies are not supposed to be representative of either the general public or elected representatives with regard to the subject of their expertise. There would be little reason for the agency to exist if they were representative. Because of this expertise, we may want bureaucrats to serve, not just as neutral implementers of policies and goals as set by elected officials, but as contributors to the policy-­making process itself. For example, if the legislative mandate is to decrease illegitimate births and outlaw many forms of birth control, it might be worthwhile to hear from a bureaucrat who points out the inconsistency of those goals. In addition to the issue of expertise, there may be other circumstances in which we would prefer to insulate bureaucrats from the interests of elected representatives. Such an institutional strategy would violate the view that principals will normally do better if they delegate to an agent with preferences closer to their own. As Bendor and Meirowitz put it, “this ancient political principle—­the boss picks the (ideologically) closest agent—­is valid given very general assumptions about preferences, shocks, and policy technology.” 24 But disrupting this close relationship is exactly the underlying notion of the trustee model. The difference is precisely that in principal-­agency theory, incentives are used to align the agent’s preferences with those of the principal; the trustee, on the other hand, is useful only to the extent that its preferences are different from those of the principal. This trustee model may be especially valuable in circumstances in which it is in the interests of the representatives to precommit to a certain kind of policy. These are circumstances in which there is a problem of time inconsistency faced by elected officials: the anticipation of self-­interested actions further down the game tree will invoke harmful actions by other players. For example, investors who anticipate election-­oriented inflationary boosts will be deterred from optimal investments. Or firms that anticipate politicized tariffs will be led to distorted decisions as well. The creation of autonomous bureaucratic institutions, the actions of which are predictable Schelling (1960). Bendor and Meirowitz (2004, 299).

23   24  

178

R e finin g R e fle xivity

and routinized by professional and legal constraints, offer an alternative institutional form that might benefit the public and even the elected officials. The value of the trustee bureaucrat, when acting properly, is that she will make decisions based on what she thinks are the long-­term interests of the relevant population and not in terms of the short-­term interests of elected representatives. In principal-­agent theory, one of the punishments available to create the right kind of incentives for the agent is termination. Ideally, the principal’s control is improved by the possibility of replacing an agent who does not act in the principal’s interests. However, in the trustee model, the delegation of authority to the trustee must be inviolable. The representative may not be able to come to the trustee and force her to act in a way consistent with the demands of the representative. Nor can the representative fire the trustee in order to replace him with a more “reasonable” person who can see the wisdom of accommodation. While principal-­agent theory affirms that the structure of the relationship must be elastic to the principal’s wishes, Schelling’s trustee theory stipulates quite the opposite; any suspicion that the principal can redefine the relationship in response to changing conditions is counterproductive. For our purposes, there is an important normative point to make at this point. The primary arguments in favor of the trustee model tend to reflect a skepticism about the effectiveness of democratic decision making. A trustee approach to bureaucratic authority will seem attractive when the interests of democratic representatives tend to deviate from what is taken to be the long-­ term interest of the relevant population. For many, this is a problem inherent in elections in representative democracies. There is a sense in which moral hazard between citizens and elected representatives is built into the system. On this account, politicians will always find it an attractive use of political authority to dispense government contracts to favorites, to build entry barriers around political supporters, and to dispense justice with a bias toward political allies. In a democracy, elected officials are always going to be the locus of this temptation; and that is why, on this account, controlling this moral hazard is not accomplished simply by making bureaucrats responsive to democratically elected officials. So, one response to the choice between the principal-­agent model and the trustee model of bureaucracy might be that the trustee model is the wrong response to the problem of democratic ineffectiveness. To the contrary, the better response ought to be greater attention to the conditions that are subverting the effectiveness of democratic decision making. This is fair enough, but we think that such a response unnecessarily usurps the institutional choices of relevant constituencies. Here we need to reiterate a basic point. We are not arguing that democracy is entitled to any first-­order priority vis-­à-­vis bureaucracy

179

C h ap te r 6

or any other feasible institutional arrangement. Factoring in the issue of expertise, we leave open the question of whether there are good substantive reasons for institutionalizing the trustee model. This is ultimately a question of institutional choice that is best made democratically. Nonetheless, the choice itself returns us to the question that animates this chapter: does bureaucracy instantiate the reflexivity necessary to satisfy the relevant second-­order tasks of institutional selection? We think it is clear that bureaucracy—­regardless of which broad conception of bureaucratic authority one adopts—­lacks the necessary reflexivity. In regard to the principal-­agent model, the institutional structure does not in principle preclude reflexivity, but it does not encourage it. The lack of autonomous choice vis-­à-­vis democratic decision makers reduces the role of bureaucrats to mere implementation. On each of the issues involved in the second-­order tasks of institutional choice, bureaucrats might serve best as an alternative source of information and as a mechanism of enforcement. But the constraint on independent decision making means that any feedback loop for assessing bureaucratic action ultimately will rely on democratic decision makers for monitoring and judgment. The ultimate distribution of authority rules out the reflexivity necessary for the second-­order tasks we have identified as central to processes of institutional selection and reform. In regard to the trustee model, the institutional structure allows reflexive thinking by the participants in the bureaucratic system. The attribution of autonomous choice creates the opportunity for bureaucrats to have an independent effect on how relevant populations manage the second-­order tasks. However, there are two primary reasons why we argue that a trustee approach to bureaucracy would not do as good a job as democratic institutions in accomplishing these tasks. First and foremost, a trustee bureaucracy will commonly lack both the diversity of perspectives and the institutional mechanisms necessary to adequately assess the relevant distributed knowledge in the populations as a whole.25 Here we find ourselves in limited agreement with Hayek, Coase, and other advocates of decentralization regarding the relative incapacity of bureaucracies to effectively aggregate relevant information and knowledge. As we pointed out in chapter 3, this criticism of agency decision making has been borne out in subsequent research on the question.26 Second, even if the trustee bureaucracy could surmount the obstacles necessary to solve the distributed-­knowledge problem, it will not be well situated to handle the Epstein (1997; 2000) provides a pointed example of how, in the 1980s and 1990s, AIDS activists in the United States challenged the federal health bureaucracy on just these grounds. 26   Farrell (1987); Bolton and Farrell (1990). 25  

180

R e finin g R e fle xivity

second-­order tasks as they relate to self-­monitoring and self-­assessment. In the end, the trustee model of bureaucracy fails to offer an adequate replacement for democratic institutions in these important tasks.

IV.  Hybrids and Social Norms Up to this point, we have focused primarily on formal institutions. Whether we have been discussing decentralized mechanisms like the market or centralized mechanisms like democratic institutions, we have conceptualized the institutional arrangement as a set of formal rules structuring the relevant social interactions. There has been good reason for doing so, given the fact that most of the normative justifications offered in support of these institutions place heavy emphasis on the formal legal institutions that form their basis. We would go so far as to say that any plausible normative account of social institutions must make this their primary emphasis. Nonetheless, this is not a complete account of the ways in which institutions coordinate social life. For it neglects the important role played by social norms. Social norms are ubiquitous. These informal rules coordinate significant aspects of many of our most important social relations. For example, they define gender roles and they structure relationships between members of different racial and ethnic groups. They delimit cultural boundaries both within and across societies. And, perhaps most importantly, they establish the basic terms of social community. Social norms instantiate commonly held beliefs about appropriate forms of behavior in relevant social situations.27 Their causal force derives, in large part, from the fact that they provide important information about the behavior of other social actors.28 By knowing the content of social norms, social actors can establish stable expectations about how others are going to act in common social situations. For example, when the content of the norms dictate cooperative behavior, social actors can use this information to develop expectations

See chapter 3 in Knight (1992) for a more developed account of the effects of social institutions (including social norms) on social expectations and beliefs. 28   By emphasizing the information effects of social norms we do not mean to diminish the significance of informal enforcement mechanisms for the efficacy of these norms. The information that social norms provide and thus the effects of norms on social expectations, while important to our general understanding as well as important for our analysis here, would not have the causal force they do if the information effects were not reinforced by the various kinds of informal sanctions that exist in most societies. 27  

181

C h ap te r 6

about the likelihood that others will cooperate and then make a decision to act accordingly. Social norms are particularly relevant to our analysis in two ways. First, one might ask whether social norms, as a decentralized coordinating mechanism, might offer an alternative to democratic decision making as a facilitator of the second-­order tasks of institutional choice. But we think that the limitations of social norms for these particular tasks are clear. Two of our arguments from previous chapters strongly support this claim. First, social norms, like other decentralized mechanisms, lack the necessary reflexivity. Social norms are most effective when social actors accept with confidence that they posit the commonly understood type of appropriate behavior in a given context. Much like the case of the market, reflexivity undermines the effectiveness of the norms. And a second argument relates to how social norms are established and maintained in a society. In chapter 3, we made an argument for a realist conception of how social institutions emerge in an unconstrained decentralized environment. As a practical matter, this closely approximates the kind of environment in which most social norms emerge. This has the following implication for the types of social norms that we would expect to emerge. They will be norms that favor in important ways the dominant groups in a society, giving them social and distributional advantage vis-­à-­vis other social groups. Therefore, these norms are unlikely to facilitate the kinds of social communication and interaction that are necessary to establish and maintain an effective framework for institutional experimentation and choice. However, by rejecting existing social norms as an independent candidate for the second-­order tasks of institutional choice, we do not rule out some role for them in this process. We might consider the secondary effects of norms in hybrid institutional arrangements. Here we have in mind the ways in which social norms interact with formal rules in most complex institutional arrangements. As we discussed in chapter 3, effective markets require formal institutional rules, but they also require a network of social norms that structure, for example, aspects of economic exchange. A similar example arises out of our consideration of the trustee model of bureaucracy. Advocates of this model must provide an account of why we might trust the trustee to act in ways consistent with the long-­run interests of the society. Some interesting research suggests that one source of confidence in the trustworthiness of bureaucrats can be found in the pervasive network of professional norms that informally structure bureaucratic organizations.29 Given these examples, one might ask if G. Miller (1992).

29  

182

R e finin g R e fle xivity

the interactive effects of formal and informal rules increase the likelihood that one of the basic forms of institutional coordination that we have previously rejected could be effective in these second-­order tasks. Again, we think not, for two primary reasons. The first is a mere reiteration of the kinds of social norms that will commonly emerge in decentralized environments. The kinds that commonly emerge are less likely to facilitate the right kinds of social interactions. The one caveat on this argument relates to the fact that, once we focus on what we have called hybrid institutional forms, we then have to take account of the constraining effects of existing formal institutions on the process of norm emergence. More on this point below. A second reason for doubting whether the hybrid forms might make the basic institution more appropriate as a general institutional framework relates to the actual effects of norms on formal institutional effectiveness. While it seems quite plausible that social norms can enhance the effectiveness of formal institutions, it does not seem likely that they can change their underlying structure. Social norms may well enhance what markets and bureaucracies are structured to do, but they cannot resolve the basic reflexivity deficit vis-­à-­vis these second-­order tasks.

V.  Democracy and Social Norms The possibility of interactive effects between formal and informal rules does, however, present one final question for this chapter: can democratic decision making as the framework for institutional experimentalism, choice, and maintenance in the formal domain have a positive long-­term effect on the kind of social norms that emerge in the society writ large? More specifically, can democracy facilitate social norms that are characterized more by socially beneficial effects than those that instantiate existing asymmetries in a society? Here we want to emphasize that our answer must be in terms of indirect effects because our claim is not that we would propose the establishment of formal procedures for norm selection. That would be a hugely unrealistic claim. On the other hand, any population might choose to create laws that offset the negative effects of informal norms, but that is a different issue. What we have in mind here are the potential indirect effects of effective democratic institutions. The desirable answer would be that the formal institutions of democracy can positively affect the emergence of social norms and enhance their role in facilitating cooperation in socially diverse societies. Such an argument may strike some as hopelessly misguided. On the one hand, the idea that formal

183

C h ap te r 6

mechanisms such as democratic institutions can either (1) substitute for informal mechanisms such as norms or (2) facilitate better social norms runs counter to many perspectives on the law. Brian Barry suggests that “[n]o society has ever got by with only legal rules.”30 Lon Fuller concludes that “[l]egal rules are not an effective device for directing human energies to those places where they can be most creatively and effectively applied.”31 Michael Taylor argues that “the state tends to undermine the conditions which make [stateless social order] workable, and in this way makes itself more desirable. It does this by weakening or destroying community. . . .”32 Such challenges are grounded in a conception of formal legal institutions as highly impersonal and routinized. On this conception, the primary effects of the legal institutions are to increase the dependence on impersonal formal mechanisms and to undermine the informal resources that foster creative and productive social interaction. This conception of formal institutions as highly impersonal and routinized is deeply embedded in modern thought on rationality and social institutions. In his classification of the bases of legitimacy, Weber emphasized that the main source of order in modern society rests on the fact that “the order has been established on a manner recognized to be legal.”33 This has commonly been interpreted to envision the impersonal and routinized conception of law that reinforces the earlier doubts about the potential salutary effects of legal institutions. On this interpretation, legal legitimacy is based on formal bureaucratic regulation and compliance is based on narrow instrumental rationality. Consider, for example, Cotterrell’s interpretation of Weber’s basic claim: The existence of law—­in particular conditions and in a particular form—­ provides its own ideological basis whatever its substantive content. And the action of the state, in accordance with law, derives legitimacy from law. Legal domination does not, therefore, depend on the law’s reflection of values to which those who accept its legitimacy are committed. As long as state action conforms to legal requirements, it can adopt any policies or reflect any values without disturbing the basis of its legitimacy.34 But these are not the only effects that formal legal institutions like democratic decision making might produce. We propose two types of salutary effects that a general framework of democratic decision making might have on Barry (1995, 34). Fuller (1981, 77). 32   Taylor (1982, 57). 33   Weber (1978, 36). 34   Cotterrell (1995, 136). 30   31  

184

R e finin g R e fle xivity

the decentralized evolution of social norms. Social norms evolve over time as the product of a decentralized process of repeated small-­n interactions. The emergence of a particular norm is a function of (1) the ways in which the small-­n coordination problems get resolved and (2) how a particular coordination solution is aggregated as the commonly understood way that such interactions are coordinated within the relevant population. While the democratic decision-­making environment will serve to define the general context in which decentralized emergence occurs, we need to identify specific mechanisms that might affect norm emergence in the desirable ways. Both of the effects we identify target the individual small-­n interaction that forms the basis of the aggregative effect. We will label these the by-­product and the spillover effects. Both might, and we emphasize might, diminish the influence of societal asymmetries on the resulting social norms, thus establishing norms that are less likely to instantiate these asymmetries. Consider first the possibility that democratic institutions envisioned as a general framework for institutional experimentation and choice might, as a by-­product, change social beliefs about the members of different social groups. This reverses the causal direction usually anticipated in these debates. This by-­product effect raises the possibility that participation in the democratic decision-­making process might change people’s beliefs about members of other social groups. That is, participation in a democratized process of institutional experimentation may have the secondary effect of diminishing negative beliefs about others, beliefs that can contribute to the long-­term evolution of asymmetric social norms. On this account, the changed beliefs about the characteristics and expected behavior of members of other, often competing, social groups might alter the dynamic of norm emergence in the direction of more egalitarian social norms. The relevant question here is, what are the features of the democratized process of institutional experimentation that might foster these changed beliefs? First, while we do not envision that the process will produce a social consensus, the requirements of participation suggest that the resulting outcomes will likely reflect a broader array of interests and commitments than would emerge in an unconstrained decentralized context. As Justice Holmes says in a famous passage, “The first requirement of a sound body of law is that it should correspond with the actual feelings and demands of the community, whether right or wrong.”35 To the extent that individual participants perceive that their views are receiving a fair hearing, whether or not their interests are actually instantiated in the substantive content of the rules, then they are Holmes (1881, 36).

35  

185

C h ap te r 6

more likely to perceive that the process is fair and legitimate.36 It is possible that the sense of fairness that people experience in the formal domain might persist in more informal settings. Second, a democratized process of institutional experimentation will, in principle, facilitate greater forms of communication across groups. Among scholars of law and legal institutions, Fuller is noted for his sustained attention to the relative appropriateness of legal processes for various types of problems and social situations. Much of Fuller’s work focuses on the circumstances under which various institutional mechanisms best resolve common social problems. The primary factor that influenced Fuller’s analysis was the importance of communication: “I believe that if we were forced to select the principle that supports and infuses all human association we would find it in the objective of maintaining communication with our fellows.”37 Fuller’s basic maxim is an appropriate guide for how legal institutions in general and democratic institutions in particular should be organized: “Open up, maintain and preserve the integrity of the channels of communication by which men convey to one another what they perceive, feel or desire.”38 Fuller anticipated that greater communication would facilitate beneficial relations among social groups beyond the specific circumstances of the particular cases. If communication can have these kinds of effect, then it offers a second reason for thinking that a democratized process would produce the salutary effects on social beliefs. We are focusing here on a very specific kind of social belief—­that the members of diverse social groups will cooperate in mutually beneficial ways in informal settings. Consider two descriptions of the types of interactions that are necessary to build informal cooperative beliefs across social groups. Rorty emphasizes a process of communication and reflection: “This process of coming to see other human beings as ‘one of us’ rather than as ‘them’ is a matter of detailed description of what unfamiliar people are like and of redescription of what we ourselves are like.”39 Sabel envisions a similar process for the extension of trust: “[T]he extension of trust in any particular setting depends in part on the actors’ reinterpreting their collective past, and especially their conflicts, in such a way that trusting co-­operation comes to seem a natural feature, at once accidental and ineluctable, of their common heritage.”40 In addition Tyler (1990). Fuller (1969, 185). 38   Fuller (1969, 186). 39   Rorty (1989, xvi). 40   Sabel (1992, 218). 36   37  

186

R e finin g R e fle xivity

to communication, Sabel highlights the importance of understanding shared interests. He argues that an acknowledgment of mutual dependence can lead to a redefinition of boundaries and an emergence of trust across previously conflicting groups. For the democratized process of institutional experimentation to produce the requisite beliefs about informal behavior, it should provide the opportunity for the types of interaction that Rorty and Sabel identify. A major strength of our pragmatist approach is that the principles and procedures that encourage formal cooperation in a process of experimentation are also the principles and procedures that are most likely to encourage these types of interaction. Especially important here is the analogy between interactions involving formal legal rules and the role of social norms in informal social interactions. On our conception of democracy, individuals work together at problem solving over the proper use of formal social institutions. This process provides the opportunity for communication among members of different social groups as to how formal rules should structure their common interactions. By analogy, these actors can translate their experiences in the legal sphere to their informal social interactions. If they do so, then the beneficial effects of their formal social interactions might influence the ongoing process by which social norms emerge and change. But it is important to note that this by-­product scenario represents only a possibility. Even Fuller, the advocate of conceptualizing legal processes as communication mechanisms, offers a skeptic’s perspective. In his work, Fuller distinguishes two basic principles of human association: shared commitment and legal principle. These roughly approximate the informal and formal mechanisms of social coordination. In his later writings, Fuller warns against the negative effects of “creeping legalism”: “As a matter of sociological observation we may therefore assert that as an association becomes increasingly dominated by the legal principle, the element of shared commitment—­ thought tacitly operative—­tends to sink out of sight; any attempt to secure recognition for its role is likely to stir anxieties and meet with strong resistance. This reaction will extend, not simply to what may be called the element of shared substantive commitment, but to that minimum commitment essential to make the legal system itself function properly.”41 There are obviously no guarantees that the reliance on legal institutions can be prevented from having such an effect. We offer the by-­product effect as only one possible reason for believing that a democratized process of institutional experimentation can have positive effects on social norms. Fuller (1981, 76).

41  

187

C h ap te r 6

The second effect, the spillover effect, does not rest on claims about changed social beliefs. Rather, it focuses on the potential effects of a formal democratic framework on the resources employed to resolve conflicts over how best to coordinate social interactions. This effect directly addresses the issue of the social context in which the interactions occur. Recall that on our realist account of unconstrained institutional emergence and change, the actors bargain over the outcome of their interactions. The conflict is commonly resolved in favor of the party who enjoys an advantage in the asymmetric distribution of resources in the society. And the resulting norms reflect those asymmetries. One of the primary second-­order tasks of democratic institutions would be the monitoring, assessment, and maintenance of the conditions necessary for effective institutional performance. This involves attention to the conditions under which other institutions work effectively as well as the conditions under which democratic institutions work effectively themselves. This last task applies also to the process of institutional experimentation. As we will argue in the next two chapters, these conditions require guarantees of free and equal participation in all aspects of democratic decision making. These guarantees will serve to prohibit political actors from taking advantage of the asymmetric distribution of resources in society within the domain of democratic decision making itself. And this is where the potential spillover effect becomes relevant. The question is, will the attention given to monitoring and maintaining free and equal participation in the democratized process of institutional experimentation spill over and thus affect the use of asymmetric bargaining power in the informal domain? It seems quite plausible that there would be such an effect. On the one hand, constraints in the formal domain against certain political strategies could raise the costs of these strategies in other aspects of social life. On the other hand, the reduced efficacy of these political strategies in the formal domain might change the perceptions of otherwise dominated groups in regard to their willingness to acquiesce to their use in the informal domain. Through resistance or collective action, they might alter the terms of bargaining and negotiation in the interactions that produce social norms. And this leads us to what we think is a striking methodological observation. If our conception of democratic decision making is institutionalized and if these proposed effects are real, then the process of spontaneous emergence of social norms will not be characterized by reflexivity but will be more likely to produce the types of norms that will be characterized by greater equality and less distributional asymmetry. If this change in social norms transpires, then the type of decentralized emergence of social norms that will obtain will be one that looks more like the convention or contract and selection models than

188

R e finin g R e fle xivity

the bargaining model that presently dominates. This last point reinforces the view of political theory that we endorse: scholars like Coase and North and others who share their views on decentralization mistake their normative visions for the appropriate explanatory mechanism. But we acknowledge that these are only possibilities. While it seems reasonable that the democratization of the process of institutional choice that would increase the egalitarian nature of the selection of formal institutions might also produce similar enhancements in the informal process, it is far from guaranteed. What we are highlighting is the possibility of a more general feedback between effective formal institutions, with their constraints on the nature of participation, and the informal domain of social life. The normative implications of such a possibility are numerous. For our purposes, the most important implication is the following: if democratic institutions can have a positive effect on the emergence and maintenance of social norms, then democracy can be an important positive condition for the fostering and maintenance of the benefits of social community. This follows from the importance of the right kinds of social norms for the effectiveness of community-­based communication and cooperation. These are issues that we have taken up in detail in chapters 2 and 3. To the extent that an effective social community requires social norms that minimize the influence of asymmetries of power, then a social context that favors the emergence of such norms can make a significant contribution to the overall effectiveness of such communities. Through this last argument, we hope to reorient two streams of contemporary thought about the relationship between the state and social community. The first relates to the commonly held view, even by many contemporary pragmatists, that the informal relationships that constitute community are a more important prerequisite for a successful democracy than are formal institutions. While we in no way mean to deny the relevance of a thriving community for effective democratic performance, we do mean to challenge the presumption that the causal path to a successful democracy runs necessarily through community first. At the very least, we want to introduce the possibility that a well-­functioning democracy—­in terms of effective legal institutions—­ can have the effect of fostering and maintaining informal social relationships that are characterized by productive communication and community-­based cooperation. The second stream of thought relates to the idea, highlighted at the beginning of this discussion, that increased reliance on law and legality will have a long-­term negative effect on the types of informal social resources, like interpersonal trust and shared social beliefs, that foster community-­based

189

C h ap te r 6

cooperation. Our argument in this section highlights the possibility that the legal institutions in a democratic society need not produce such negative effects and might, under the right conditions, actually foster the social resources relevant to informal community-­based cooperation. The views that dominate both of these streams of contemporary thought need to be reconsidered. They are grounded in incomplete, if not outright mistaken, assumptions about the possible consequences of formal institutional arrangements. When we talk about the effects of any institutional arrangement, we usually make a mistake if we assume that there is only one possible effect. Rather, in most instances there is a range of possible effects that depend on the conditions in which the institution operates. So, for example, when thinking about the effects of formal institutions on social norms and community, we need to take account of the fact that democratic institutions, like most types of institutional arrangements, can have both positive and negative effects and that the dominant effects in any particular circumstance will be, in substantial part, a function of the conditions under which they are operating. This fact reinforces one of the most significant features of the conception of pragmatism that we advocate throughout this book, the central importance of tempered consequentialism. Proponents of normative justifications of any type of institutional arrangement must attend to both consequences and conditions in making their case. What we have presented up to this point is an argument about the optimal operation of democratic decision-­making institutions. But the argument is not complete until we address the necessary conditions for the effectiveness of those institutions. It is to that aspect of our argument that we now turn.

190

Part three

This page intentionally left blank

Chapter 7 Formal Conditions: Institutionalizing Liberal Guarantees

I.  Introduction Our argument for the priority of democracy is based on a set of claims about the capacity of democratic institutions to facilitate the effective performance of social institutions. On our broadly consequentialist account, institutional effectiveness is the normative criterion for a justification of the legitimacy of democratic institutions. This normative criterion, in turn, is grounded in our commitment to pragmatism. Pragmatists assess the value of their choices and actions in terms of the consequences of those choices and actions. The relevant consequences in most situations are a product of the interdependent choices of many social actors and are affected in significant ways by the institutional context in which those choices take place. In this sense, good consequences are a product of effective social institutions. Thus, a pragmatist’s normative concerns will necessarily engage questions of institutional design and choice. The priority of democracy, as we depict it, reflects a two-­level analysis of institutional choice. At the first level, democratic arrangements are but one among a set of social institutions that a population might use to coordinate its ongoing social, political, and economic interactions. The members of any modern society are confronted by an ever-­increasing set of such coordination problems. What types of goods and services should be coordinated by market exchange and what types are better provided by a more centralized mechanism? What types of social disputes are best resolved by the courts and what types are best resolved by other forms of legislative or administrative procedures? How do we best coordinate our efforts to create and maintain public goods? Choices like these among alternative first-­order institutions will reflect the judgment of members of the population in question about which institutional alternative will allow them to produce the best consequences in a particular social situation. Democracy has no priority among these alternatives; no institution does. Institutional choice at this stage will often involve experimentation and will

C h ap te r 7

necessarily involve a collective assessment of institutional performance and consequences. And after such an initial choice is made, it will be followed by an ongoing process of monitoring performance and guaranteeing the conditions under which the institutions will perform effectively. Democracy enjoys a priority because of its unique capacity to facilitate these crucially important second-­order tasks of experimentation, assessment, monitoring, and sanctioning. But, as is the case with any type of institution, democratic decision making only performs these tasks effectively under certain specifiable conditions. This follows from the fact that any social institution commonly will produce both positive and negative effects. To facilitate the positive consequences and mitigate the negative ones, it is necessary to identify and maintain the set of conditions associated with the positive effects. For example, the interrelated mechanisms of voting and argument will produce their positive effects on information and knowledge acquisition, and thus on the quality of collective decisions, only under conditions that promote robust political participation. The same can be said for the positive benefits of democratic reflexivity. And the capacity of democratic institutions to promote the positive benefits of diversity are contingent in large part on the ways in which political competition among differing social groups is structured. This is the underlying basis of pragmatism’s commitment to tempered consequentialism. We cannot overemphasize the importance of this fact about the conditions necessary for effective institutional performance. Yet, strangely enough, many recent accounts of the beneficial consequences that democratic institutions produce fail to display appropriate concern with the conditions necessary for effective democratic decision making. Whether we are considering Ober’s argument about the superiority of Athenian democracy or Rodrik’s reliance on democracy as a meta-­institution or Page’s promotion of the epistemological benefits of collective decision making within diverse communities, just to name a few, we cannot help but conclude that the consequentialist arguments these thinkers offer for democratic decision making remain both analytically and normatively incomplete. Without a clear understanding of the conditions necessary to produce these positive consequences, we cannot make an adequate assessment of the overall merits of these claims. Analytically, we need to know more in order to assess the plausibility of the causal claims about institutional effects. Normatively, we cannot decide whether or not we should choose democratic decision making over the other institutional alternatives unless we know the full implications of that choice. That, in turn, entails knowing the costs and benefits of establishing and maintaining the necessary conditions of effectiveness.

194

Institu tio n a liz in g L ib e ra l G u a ra nte e s

Our argument for the priority of democracy presumes that it is possible to establish and maintain the conditions necessary for democracy to foster good consequences. This has the following implication for a pragmatist justification of democracy: the conditions of causal efficacy are the same as the conditions of normative legitimacy. For democratic institutions to perform effectively, a fundamental condition must obtain: the free and equal participation of all relevant individuals.1 To the extent that this condition is not satisfied, the arguments in support of the relative superiority and, thus, the normative legiti2 macy of democracy are called into question. In this chapter and the next, we discuss the free-­and-­equal-­participation condition and develop its implications for the institutionalization of democratic decision making. In doing so, we set out the particular form that freedom and equality must take in order to guarantee that the decisions that issue from the democratic process can be normatively justified. We must do this in some detail in order to adequately demonstrate how complex these conditions are and how difficult they may be to actually institutionalize. Consequently, we move beyond the standard ideal-­typical statements of abstract rights and attempt to identify the actual conditions that must be institutionalized in order to achieve the desired normative results. Here it is important to recognize both the narrowness and the breadth of our argument. On the one hand, we limit our analysis to the conditions of democratic freedom and equality. We offer an argument about the requirements of freedom and equality within the context of political decision making. We do not develop a general theory of freedom and equality. However, we do acknowledge that the conditions of legitimate political decision making may have implications for other aspects of social life and note them where

1   It is important to note that there is no conflict between this claim, which revolves around participation, and our earlier insistence on the representative character of democratic institutions. As Plotke (1997) cogently argues, participation and representation are not antonyms. Politically, the proper contrast to representation is exclusion. The extent and character of participation among those who are included in a given political association—­those, in other words, who are represented—­is a distinct matter. It is to that matter that we now turn our attention. 2   Here we should emphasize that our argument on behalf of equality and freedom is based on the importance of these conditions for the quality of democratic outcomes. In making this argument, we do not reject the possibility that arguments about the intrinsic quality of these conditions might be relevant and persuasive, yet we do not rest our case on them. In addition, our argument differs from other consequentialist arguments like utilitarianism in that, while it focuses on the consequences of actions, policies, institutions, etc., it offers an indirect justification of guarantees for freedom and equality as being necessary preconditions of these good consequences.

195

C h ap te r 7 3

relevant. On the other hand, we take as part of the task of justifying democracy the considerable burden of articulating how these conditions should actually be implemented in a democratic society. This requires us to address questions of institutional mechanisms as well as issues of public policy that largely determine whether or not institutional guarantees operate effectively. Before turning to specifics, we think that it would be helpful to offer a brief overview of the structure of our argument and its major claims. Throughout our argument, we maintain two basic ideas about how freedom, equality, and democracy relate to each other. First, since the justification for the free-­and-­ equal-­participation condition rests on its effects on the decision-­making process, the development of the implications of this condition must be directed toward the conditions that would actually achieve these effects. This is fundamentally a question of social causation: what kind of political participation is necessary to ensure that democratic decision making will produce the desired effects? Concern with this question requires us to make an important distinction between formal rights to participate and effective opportunities to participate. This distinction highlights an important difference between liberal and pragmatist conceptions of democracy. Liberalism is primarily concerned with formal rights, focusing on constitutional and other legal guarantees that protect individual citizens from illegitimate state interference in their affairs. These guarantees are deemed satisfied when they are successfully instantiated and maintained in the formal legal and political procedures of the society.4 Arguments that base their support for democratic institutions on the consequences that these institutions produce, including pragmatist accounts like the one that we present here, must go beyond classic liberal formulations and extend the analysis of political and institutional guarantees to the more substantive questions dictated by concerns with institutional effectiveness and democratic performance. We argue that liberal defenses of formal rights afford an insufficient basis for justifying democracy insofar as they are inattentive to matters of institutional performance. We further argue that we must also 3   In chapter 8 we briefly discuss the indirect implications of pragmatism for social life. In doing so, we consider the kinds of individual capacities and institutional arrangements that are necessary to live the life of experimentalism dictated by pragmatism. In this discussion, we consider: (1) how the dictates of democracy in the political sphere differ from the dictates of pragmatism in the social sphere and (2) how the actions, policies, and institutions of the political state may be a precondition of satisfying the dictates of pragmatism in social life. 4   The basic difference between our pragmatist approach and standard liberal accounts—­ exemplars would be Rawls, Dworkin, Habermas, and their intellectual descendants—­is that we are adopting a brand of consequentialism rather than some version of a deontological strategy of justification. This meta-­theoretical distinction is beyond our current concerns.

196

Institu tio n a liz in g L ib e ra l G u a ra nte e s

guarantee effective opportunity if we are to establish the necessary context in which the desired effects of democratic decision making will obtain. That is, the effective performance of democratic institutions is a function of the effectiveness of individual participation. Second, on the view we advance, the conditions of freedom and equality are intimately interrelated. Because of this, unlike many accounts of democracy, we integrate the analyses of the implications of the two participation requirements throughout the chapter. On our account, freedom and equality do have distinct implications for the nature of the decision-­making process. Freedom entails something about the nature and content of individual participation in the process; equality entails something about the interrelationship of the individual contributions to that process. At the same time, the freedom and equality conditions reinforce each other in ways that are necessary for the ongoing maintenance of these conditions and, thus, for the effective performance of democratic institutions. In the course of our argument, we defend the following basic claims about the institutionalization of the free-­and-­equal-­participation condition. First, individuals must be guaranteed freedom to participate in all aspects of the democratic decision-­making process. This freedom extends from the initial development of one’s own ideas, beliefs, and goals through the process of argument and debate (which includes the opportunity to offer one’s own proposals, to defend these proposals with argument, and to assess the proposals of others) to the final collective decision (which will generally involve some form of voting on the merits.) The requirements of effective institutional performance dictates that at each stage of the process, individuals enjoy a voluntary and uncoerced opportunity to influence political decision making. The emphasis on guaranteeing the opportunity for actual influence follows from the causal requirements of the institutional justification of democracy: democratic institutions are most likely to perform effectively when each individual has the opportunity to contribute to each aspect of the decision-­making process. In the process of institutionalizing such a guarantee, we must address such questions as (1) what types of access are required by this criterion? (2) what types of protections from interference does it imply? and (3) what types of public policies are necessary to ensure the opportunity for effectiveness? Second, this guarantee of freedom to participate must be distributed equally among the individual members of the society. A key element in guaranteeing effective institutional performance is the set of institutions and policies that preserve the beneficial effects of free participation. In this regard, equality of political participation, in the form of equal opportunity for political influence, is a necessary condition for the prevention of distortions in the decision-­making

197

C h ap te r 7

process. Institutionalizing political equality involves a task of neutralizing (or, in a weaker version, minimizing) the distorting effects of inequalities on democratic freedom, and thus on democratic participation and decision making. To see what this task requires, we must address a basic question: what factors must be equalized to guarantee equal opportunity for political influence? This entails analyzing the effects of formal political rules, the effects of social norms and custom, the effects of social asymmetries that may be relevant to political participation, and the effects of inequalities in individual capacities. Third, the process of applying the normative criteria of free and equal participation to real political institutions will involve difficult and sometimes controversial decisions. The institutional requirements do not always lend themselves to straightforward implementation and often confront us with trade-­offs with other things we may value. While we identify significant problems in efforts to quantify the concept of equal opportunity of political influence, the task is akin to a group maximization problem: maximize the opportunities for effective participation, given our other values. The emphasis throughout must be on effectiveness, and the focus must be on mitigating the factors that would adversely affect opportunities for influence. By our own assessment, formal freedom is the easiest to institutionalize, but it does raise some challenging issues. Formal equality is somewhat more difficult to achieve, especially when you take account of the causal dynamics of deliberation. Effective freedom and equality raise complex questions for both institutional and policy choice. Given both the trade-­offs that must often be addressed and the conflicts generated by the diversity of values in any society, the process of defining the normative boundaries of legitimate democratic decision making will inevitably be a political one. And the process may often produce counterintuitive recommendations. To establish these guarantees, the state may have to (1) interfere with the choices and opportunities of some persons to establish the requisite kind of democratic freedom and (2) treat different persons dissimilarly in order to establish the requisite kind of democratic equality. Therefore, it is important that there be procedures that (a) establish the legitimacy of these state actions and (b) protect against the illegitimate interventions of state actors. Here our pragmatist account accepts in a straightforward manner the considerable burden of justification inherent in normative debates over institutional choice. And in this way, we once again see the important difference between the dictates of liberal and pragmatic justifications of democracy. Pragmatism challenges liberalism to acknowledge the complexity involved in establishing the conditions of effective democratic decision making. In the end, there is a tension here that is subject to constant negotiation and reconsideration: a tension between the inherently political nature of the

198

Institu tio n a liz in g L ib e ra l G u a ra nte e s

institutionalization process and the social and institutional preconditions that must obtain for that very process to be considered normatively legitimate.

II.  Dimensions of Freedom Freedom of participation is a fundamental feature of the justifications of many of our basic social institutions. The legitimacy of these institutions is grounded in a requirement that individuals be guaranteed freedom to participate in the social, economic, and political activities of modern society: [T]here are many languages of legitimation present in the world of democratic politics today. One of the striking features of those languages, however, is that all of them share a common idiom of freedom or liberty. The language of economics directs us to the free market and to the freedom to make whatever contracts we will with one another; the language of rights focuses on rights of free thought, free expression, free movement, and the like; the language of welfare and fairness and equality, or of poverty and exploitation and subordination, claims to articulate the requirements necessary for enjoying freedom or for making freedom effective. And the language of democratic legitimation harps on the legitimacy of what a free people freely decide, and on the way in which individual persons share in that collective freedom.5 In the specific context of political decision making, democracy has always been closely linked to the idea that individuals freely participate in the decision-­making process. The freedom-­to-­participate requirement applies to a wide range of political activities. Sen identifies the broad array of protections that such a requirement seeks to guarantee: “the opportunities that people have to determine who should govern and on what principles, and also . . . the possibility to scrutinize and criticize authorities, to have freedom of political expression and an uncensored press, to enjoy the freedom to choose between different political parties, and so on” as well as “the political entitlements associated with democracies in the broadest sense (encompassing opportunities of political dialogue, dissent and critique as well as voting rights and participatory selection of legislators and executives).”6

Pettit (1997, 6). Sen (1999, 38).

5   6  

199

C h ap te r 7

Such ideas have been instantiated in the basic constitutional protections of democratic societies. While there is striking similarity in the form that these protections take in most democratic societies, there is considerable disagreement among democratic theorists as to their institutional and policy implications. This disagreement is, in large part, a function of the different justifications offered for the normative legitimacy of the democratic process. From our perspective, the controversy is substantially a disagreement over what kind and how much individual participation is necessary to produce and maintain legitimate collective decisions in a society. Here we offer an account of political freedom in terms of the guarantees necessary to achieve and maintain the type of participation required for effective institutional performance. In doing so, we argue that the freedom of participation protections should take account of whatever factors might negatively influence the quality of an individual’s contribution to the collective decision-­making process. In order to analyze these factors, we employ a simple sketch of how individuals participate in democratic decision making. Each individual has the opportunity to contribute to collective decision making at various stages of the process. These include opportunities to contribute their ideas and beliefs in the various input stages of the process (e.g., all of those points at which relevant participants engage in interpreting needs or interests, advancing proposals, setting the agenda, assessing and amending policies, and so forth) as well as in the final stages of determining the collective outcomes. At each of these stages, individuals make choices whether or not to act and, if so, how. Their choices may be motivated by many things, but we assume that they are concerned about being effective—­that is, about affecting the collective decision in a way consistent with these motivations. Thus, they base their choices on their own goals and on their expectations about the implications of their choices for the nature and content of the collective decision. Their expectations are a function of a number of factors relevant to the social context in which they find themselves. This fact reinforces the idea that political participation is a social act and, thus, that there are necessarily social aspects to any 7 adequate concept of political freedom.

On this idea, consider this comment by Sen (1999, 31) on the social aspects of freedom: “Individual freedom is quintessentially a social product, and there is a two-­way relation between (1) social arrangements to expand individual freedoms and (2) the use of individual freedoms not only to improve the respective lives but also to make the social arrangements more appropriate and effective. Also, individual conceptions of justice and propriety, which influence the specific uses that individuals make of their freedoms, depend on social associations—­particularly on the interactive formation of public perceptions and on collaborative comprehension of problems 7  

200

Institu tio n a liz in g L ib e ra l G u a ra nte e s

The relevant analytical question is, what factors affect individual choice in such a way as to negatively influence the potential effects of participation on the collective outcome? Any answer to this question must involve an analysis of the individual participant and the context in which her participation takes place as well as a set of claims about the effects of various factors on the exercise of political choice. And here it is important to note the connection between our answer to the analytical question and our normative claim about the necessary conditions for legitimate democratic decision making. Given that our normative justification for democracy is grounded in the effectiveness of institutional performance and that the analytical question focuses on causal effects on effective performance, the answer to the analytical question constitutes the first cut in the process of specifying the necessary normative conditions. That is, those factors that negatively affect the nature of individual participation are the factors to which we must attend in an effort to institutionalize political freedom. However, we need to emphasize the preliminary nature of this connection. As we will point out later in this discussion, once the relevant factors have been identified, we must still consider questions of the feasibility of neutralizing the negative factors as well as questions of the potential trade-­offs between the requirements of freedom and other desirable goals. Only after assessing these issues can we offer a comprehensive account of the normative requirements of political participation. Not surprisingly, the analysis of the effects on political choice turns out to be a fairly complex one. To illustrate the layers of this complexity, we want to highlight the factors that are identified by three alternative conceptions of 8 freedom. These three conceptions characterize freedom as noninterference, nondomination, and nonlimitation, respectively. First, freedom as noninterference is probably the most common conception of political freedom in contemporary debates. Most classical liberals—­ Benjamin Constant, John Stuart Mill, Isaiah Berlin—­endorse a variant of this conception of liberty. On this conception, freedom entails the absence of direct interference in political choice. This interference can take many forms. The most intrusive ones consist of complete barriers to access to the institutional mechanisms of collective decision making. Such barriers effectively prevent not only individuals but entire classes of individuals from participating in the political process. They may take the form of some kind of and remedies. The analysis and assessment of public policies have to be sensitive to these diverse connections.” 8   Here we draw on Pettit’s (1997; 2001) framework for analyzing political freedom. We do, however, reach different conclusions about the requirements of legitimate political participation.

201

C h ap te r 7

incarceration or imprisonment or, more commonly, they may be the product of constitutional or legal restrictions on eligibility for voting or other forms of participation. Less intrusive, but no less causally effective, further forms of interference involve intentional acts of coercion against individual participants. In such cases, the interference does not actually prevent participation but does restrict the full enjoyment of the opportunity to influence. Put in analytical terms, direct interference of this type serves either to remove certain alternatives from the feasible set or alters the values of some of the alternatives. Such coercion usually consists of threats, which have the effect of altering the range of options available to the participant, causing her to choose an option different from the one that she would prefer absent the expected harm triggered by the 9 coercive act. This form of coercive interference can be directed at the actual choices among the political outcomes or at the strategies necessary to pursue such choices. If directed at specific political outcomes, the interference will generally involve legal constraints on certain policies or political programs, for example, the constitutional prohibition of the Nazi political agenda in Germany after World War II. If directed at political strategies, the interference may take the form of constitutional or statutory constraints on certain ways of advocating political preferences, for example, restrictions on speech or the press or political association. All of these constraints can be interpreted as barriers to the opportunity to influence political outcomes. Or this type of strategic interference may take the form of direct threats of harm or injury if someone chooses to pursue certain alternatives. These direct threats characterize the political oppression of ethnic and racial minorities in many societies. Freedom as noninterference treats any of these forms of interference as violations of political freedom. The primary characteristic of each of these forms of interference is a direct act by one individual with the intent of negatively influencing another individual’s choice. In the liberal tradition, the primary focus has been on governmental interference, on interference in political choice by state actors. However, as Mill made clear when he sought to protect individual liberty from the “despotism of custom,” concern with this form of interference need not be limited to state actors. Direct interference can also be the product of the actions of other individuals in the decision-­making process. Second, freedom as nondomination expands the range of potential effects on political choice that might call for special protections. This conception emphasizes the variety of ways in which one individual might negatively affect Note that we discuss the coercive effects of promises below.

9  

202

Institu tio n a liz in g L ib e ra l G u a ra nte e s

the political choices of other actors. On this conception, a person can be said to be free from domination when she is “not exposed to an arbitrary power of interference on the part of others.”10 For our purposes, the importance of this conception rests in the idea of “exposure to an arbitrary power.” Relative to the noninterference conception, the domination approach alters the focus of analysis in two significant ways. On the one hand, it expands the range of potential negative factors from instances of direct, intentional interference to 11 cases of anticipated reaction. On this conception, individuals may be considered unfree if they alter their choices in anticipation of the reaction of another who they believe has the power to negatively affect them. This does not require any intentional act of interference on the part of the powerful actor.12 It merely requires the existence of the power asymmetries that produce the reasonable belief that the dominated individual will suffer harm or injury if she chooses a political option contrary to the interests of the dominating actor. So, for example, a worker who alters his political choices so as not to incite economic retaliation from his employer suffers the effects of anticipated reaction. This expands the range of factors potentially eligible for freedom protections from the interpersonal interactions identified by the interference conception to the social factors that allow for the domination effects in the anticipated reaction case. Note further that, by implication, nondomination also expands the range of factors to include actual instances of indirect interference—­that is, interference in nonpolitical spheres as a way of affecting choice in the 13 political sphere. On the other hand, the nondomination conception restricts the range of acts of direct interference that would be considered constraints on political freedom. This conception places the emphasis on those instances of interference that are arbitrary and capricious.14 This allows for the possibility that some forms of direct interference and some instances of anticipated reaction are consistent with the attribution of political freedom. On this conception, interference is nonarbitrary when the interferer’s actions “track the interests

Pettit (2001, 138–­39). Bachrach and Baratz (1970); Lukes (2005). 12   Pettit (2001, 138). 13   Pettit (1997, 274) offers the following account of the way that we might interpret the state’s responsibilities here: “The natural way to cast freedom as non-­domination is in the role of a value that the state should try to promote, not in the role of a constraint that it has to honour; this, moreover, is the way in which it is generally cast in the republican tradition: the tradition is consequentialist in character.” 14   Pettit (2001, 138–­39). 10   11  

203

C h ap te r 7

and ideas of those who suffer the interference.”15 While freedom as noninterference would rule out any form of intentional interference with political choice, freedom as nondomination will rule out only those forms of intentional interference that are arbitrary according to this definition. This has the following institutional and policy implication: assuming that nondomination is something that a society values and for which it seeks constitutional protection, state interference in the choices of political actors is justified when that interference is intended to alleviate the negative effects of domination.16 For the purposes of our analysis, this highlights the importance of assessing the net effects of both direct interference and the sources of anticipated reaction for the determination of appropriate guarantees of political freedom. Third, freedom as nonlimitation emphasizes the broadest range of potential negative effects on political decision making. It identifies a large number of factors that might limit the choices and, thus, the potential outcomes available to political actors. These factors can be divided into three main categories.17 The first are the set of constraints emphasized by the interference and domination conceptions: cases of intentional action. Here we can include direct and indirect interference as well as instances of anticipated reaction. The second category includes informal societal factors that can constrain the choices of political actors. Here we can include social norms and conventions, the informal rules that structure social life. For example, a social norm that prohibits women from certain kinds of public activity that are necessary for full political participation can constitute a limitation on political freedom. The causal effect of such norms is somewhat similar to the mechanism at work in the anticipated reaction case: individuals refrain from making a choice in the political context in anticipation of the negative reaction of others in social contexts to the violation of the norm. Put simply, social norms can constrain political choice if, after taking account of the likely reaction to the norm violation, an individual deems the net effect of such a choice a negative one for her overall welfare. The third category of limitations adds the circumstances of the individual participant to the mix. It includes factors that might prevent the individual from realizing the full enjoyment of her political opportunities. Here the focus is on the capacity of the individual to participate in the decision-­making process. Such limitations may range from an individual’s basic cognitive capacities and abilities to such factors as disability and poverty.

Pettit (1997, 272). Pettit (1997, 273–­74). 17   Pettit (2001, 129–­30. 15   16  

204

Institu tio n a liz in g L ib e ra l G u a ra nte e s

On the conception of freedom as nonlimitation, an individual’s political freedom is a function of the degree to which he or she is not subject to these limitations. On the strongest version of this account, all of these limitations are treated similarly and the role of the state is to protect individuals from the negative effects of these limitations. In this way it differs from the other conceptions in that they make a distinction between intentional and interpersonal effects, on the one hand, and the institutional and individual effects, on the other. In doing so they contend that the interpersonal factors are the most harmful to political freedom and thus are the ones that warrant constitutional protection. The nonlimitation conception would potentially extend state protections to the institutional and individual effects. This disagreement raises important issues about the nature of state involvement in guaranteeing freedom, issues that we address in greater detail later in the chapter. For now, this consideration of the three alternative conceptions of freedom provides us with a framework of potential effects on political choice. From the perspective of an individual participant, her choices can be potentially affected by factors in each of these categories: •  direct interference by the state •  direct interference by other citizens •  indirect interference •  social factors that form the basis of anticipated reactions •  social norms and conventions •  individual capacities and circumstances. Each type of effect must be considered to see if it warrants inclusion in a normative account of the prerequisites of democratic legitimacy.

III.  Institutionalizing Formal Freedom We begin our discussion of the normative preconditions of democracy by distinguishing between formal and effective participation. We invoke this distinction in our discussion of both freedom and equality. It is analogous to a distinction offered by Sen in his discussion of the role of freedom in a development 18 strategy for nations. He suggests that there is an important difference between the notion of freedom that is instantiated in formal procedures and a concept of freedom that incorporates a consideration of the real opportunities that individuals actually enjoy in a society. Sen argues that an adequate development Sen (1999). We discuss the implications of this distinction in greater detail below.

18  

205

C h ap te r 7

strategy will employ the latter concept and take account of both the institutional protections and the policies necessary to guarantee such conditions. We think that a similar distinction applied to political freedom helps to organize the analysis of the normative preconditions of democracy. Formal freedom refers to the procedural rules that define the various methods of participation in the decision-­making process. Effective freedom refers to the nature and quality of an individual’s participation. Attention to issues of formal political freedom characterizes all major justifications of democracy. Whether or not democracy entails consideration of the factors relevant to effective political freedom is a more controversial question. Here we confine our focus to issues of formal freedom and equality, reserving the analysis of effective participation until chapter 8. The primary problem to which formal guarantees of freedom are addressed is one of direct interference in individual participation. Whether the interference takes the form of complete barriers to access or the form of intentional acts of coercion, it is easy to see how they violate any reasonable understanding of the role of political freedom in a democratic society. Choice in the face of such interference is inevitably distorted and undermines the idea of an individual’s voluntary contribution to the collective decision-­making process. The effective performance of democratic institutions is a function of the diversity of ideas that originates in the contributions of individual participants. It is also a function of the diversity of challenges that these same participants bring to bear as they test the merits of various alternatives. Direct interference in political choice can diminish this diversity and, thus, negatively affect institutional performance. Freedom to participate requires, at the very least, protection from such forms of direct interference. In principle, the task of institutionalizing these formal protections is a fairly straightforward one. The goal is to guarantee uncoerced access for all individuals to all arenas of political participation. Such guarantees are achieved by establishing and maintaining basic rules governing access and prohibiting direct interference. These rules will generally consist of some mix of constitutional provisions and statutes. The key to the effectiveness of these guarantees is the capacity of the state to monitor and enforce these rules. The state’s enforcement role raises an additional layer of complexity to the implementation of formal freedom. The very act of enforcement constitutes direct interference by state actors in the choices of individual participants. In a broad sense, the enforcement of formal guarantees of political freedom constrains the use of certain strategies, like threats of private retaliation, that some political actors may use to their individual advantage. Nonetheless, the normative preconditions for formal democratic freedom justify the state’s

206

Institu tio n a liz in g L ib e ra l G u a ra nte e s

authority to preclude such strategies. Here state interference in the political choices of individual participants introduces a basic tension between freedom and enforcement that can only be resolved as part of a broader process of monitoring and assessing effective institutional performance. State enforcement can only be justified when it is in the pursuit of goals that are the product of normatively legitimate collective decisions. Such decisions, on our account, will only emerge within a democratic decision-­making setting. Formal freedom requires us to monitor and regulate three kinds of interactions in the political arena. We can distinguish these interactions in terms of whether they involve vertical or horizontal relationships. The first, to which we have already alluded, is the relationship between individual participants and the state. This interaction raises issues of vertical freedom in the sense that the actors are situated in a hierarchical relationship, with the state actors having the legal authority to monitor the behavior of participants in the democratic process. Individual participants must be protected from those forms of direct interference by state actors that are not justified by the normative requirements of free and equal participation. When state actors move beyond their monitoring and enforcement responsibilities and seek instead to negatively influence the substantive choices of individual participants, the formal freedom of these participants has been violated. In contrast to the concerns of vertical freedom, the other two kinds of relevant interaction raise issues of horizontal freedom. Both involve relationships among the individuals who participate in the political process. One kind of horizontal interaction involves actors whose political preferences conflict. Direct interference by political opponents is a classic violation of formal political freedom. Many forms of political coercion are clear, especially when a threat is coupled with an action that carries out the threat. Yet, when we are confronted with mere verbal communication, the line between illegitimate coercion and acceptable political strategy is often blurred. Here the primary issue is, what is the content of the political communication? If the communication focuses exclusively on the substantive issues under consideration, that is, claims about the implications and consequences of implementing the policy at issue, then the statement would be a legitimate political strategy. An example of this form of communication is, “If we adopt these economic policies, the market will collapse and we will all lose our jobs.” If, on the other hand, the content of the statement deviates from the issues under consideration and focuses instead on consequences unrelated to the policy at issue, questions of coercion and illegitimate interference emerge. An example of this form of communication is, “If you support these economic policies and they are enacted, I will fire you!”

207

C h ap te r 7

The second kind of horizontal interaction involves relations among political actors who share political goals. At first blush, direct interference in free political choice among political allies may seem unlikely, but the collective-­ action problems inherent in political associations make coercion an often attractive strategy for political leaders. Direct interference can be a problem among allies who seek to coerce others in order to maintain group and party discipline. This is especially a problem for political minorities who may need 19 group discipline to achieve any form of political success. Instances of intra-­ group interference in political decision making violate a strict conception of political freedom. And it could be a violation under our effective performance justification. Our account is grounded in the value of each individual’s contribution to the collective decision-­making process. So, intragroup interference in order to maintain group solidarity will, in principle, undermine diversity and, in doing so, undermine the potential positive effects of these contributions. Our default view is that intragroup interference violates formal freedom. But this default may be open to rebuttal if a persuasive argument can be made that group solidarity is necessary for any member of the group to have a causal effect on the decision-­making process. The effective institutional performance justification of democracy recommends a strong presumption against this 20 exception. The preceding discussion suggests the following about the institutionalization of formal protections of freedom: Formal protections of vertical freedom are easy to establish but may at times be difficult to maintain. Formal protections of horizontal freedom raise challenges in monitoring both allies and opponents.

IV.  The Importance of Equality Equality plays a significant role in standard justifications of democracy. This is true for those who conceive of democracy primarily in terms of voting as well as those who place their emphasis on the deliberative aspects of the decision-­ making process. The general importance of political equality becomes clear if we consider one recent definition of democracy: “By a democratic procedure I mean a method of determining the content of laws (and other legally binding See, for example, the discussions of social movements and political associations in Gutmann (1998). Rosenblum (1998); Polletta (2004); and Rosenblum and Post (2002). 20   We return to this issue below in our discussion of indirect interference and intra-­group social norms. 19  

208

Institu tio n a liz in g L ib e ra l G u a ra nte e s

decisions) such that the preferences of citizens have some formal connection with the outcome in which each counts equally.”21 At a minimum, democracy requires that each individual vote count equally. Given that democratic processes will generate losses as well as winners, the equal treatment of final votes is essential. But advocates of deliberative democracy usually extend the requirements of equality beyond voting. This follows from the fact that they tie the legitimacy of democratic outcomes to the deliberative features of the pro22 cess: “a legitimate decision . . . is one that results from the deliberation of all.” Indeed, most defenses of democratic deliberation insist that this criterion of 23 legitimacy requires a strong form of equality. Some critics, however, remain unpersuaded. “Explorations of deliberative or communicative democracy” one such critic observes, “often refer rather grandly to a principle of equal access to decision-­making assemblies or substantive equality in resources and power, but they do not give much consistent attention to how these conditions would ever be achieved.”24 We find this criticism persuasive and seek to 25 respond to it. In doing so, we address the question, what kind of equality is necessary for the effective performance of democratic institutions? Given our stress on the uncertainty of outcomes produced by democratic arrangements, such arrangements obviously cannot require equality of outcomes, at least insofar as equality of outcomes might mean that all relevant interests are equally manifest in collective outcomes. Real equality of influence is unachievable under democratic procedures because the very nature of the process makes the outcomes uncertain and subject to the exigencies of political debate and deliberation.26 Equality of influence would, on our account, spell the end of democratic politics. Democracy, then, requires some version of equality of opportunity, of which there are many conceptions on offer.27 We limit the analysis here to equality of political opportunity because that is all the discussion of the role of equality in democratic theory is about. But this limitation hardly diminishes the complexity of the issues at hand. Barry (1991, 25). Recall too our earlier comments on Schumpeter and the “minimalist” theories of democracy that purportedly descend from his work. Even these theorists concede that effective political competition entails an equal distribution of rights and liberties. 22   Manin (1987, 352). See also Cohen (1989a). 23   Again, see Cohen (1989a; 1989b). 24   Phillips (1995, 154). 25   On this point, see the important contributions by Brighouse (1996) and Christiano (1996), each of whom directly addresses the relation of deliberation and equality. We note convergences and differences between our respective arguments at several points below. 26   Dworkin (1987). 27   Roemer (1993, 146–­47). 21  

209

C h ap te r 7

On our account, democracy requires a criterion of equality that focuses on the factors that may affect the quality of the participation of individual citizens. Democracy requires that we equalize whatever opportunities exist for political influence in the decision-­making process. In this way, the requirements of equality track the normative requirements of free participation. The institutional goal is to neutralize (or, in a weaker version, minimize) those inequalities that negatively influence an individual’s contribution to the collective decision-­making process. This is what we mean by equal opportunity of political influence. The institutional task is to find a way to treat each participant equally at those stages of the democratic process where participation is deemed relevant to effective institutional performance. Here it is helpful to invoke the conceptual distinction between formal and effective political equality. Democracy presupposes formal procedural guarantees that afford equal access to relevant deliberative arenas at both 28 agenda-­setting and decision-­making stages. These are the types of formal constitutional protections that are endorsed in classic liberal accounts. This is a difficult problem of institutional design any defensible solution to which is necessary but not sufficient to establish the sort of political equality of opportunity that democratic deliberation requires. Effective institutional performance requires more than this. Because deliberation revolves centrally around the uncoerced give-­and-­t ake of reasoned argument, it also suggests the need for a more substantive notion of equal opportunity of political influence. This raises a question of how far the requirement of equality of political participation extends. Our analysis of the dimensions of free participation suggests that equality may readily extend beyond institutional access to the basic circumstances of social life. We can distinguish two aspects of the question. First, effective institutional performance calls attention to the horizontal relationships among political actors. This suggests that political equality may have implications for both the distribution and the use of resources, at least to the extent that they are relevant to ensure that an individual’s political choices are indeed uncoerced. Here we have in mind such factors as material wealth and educational treatment. Second, effective performance raises questions about the capacity of individual participants to advance persuasive claims. Here the questions relate to whether political equality requires us to accommodate and remedy the asymmetrical distribution in any political constituency of relevant deficiencies and faculties (e.g., in the ability to reason, to articulate ideas, and so forth). 28   Dahl (1989, 107, 112–­14) stresses the importance of ensuring access at both stages. Christiano (1996, 262–­68) stresses the complexity and contestability of deliberative agenda-­setting.

210

Institu tio n a liz in g L ib e ra l G u a ra nte e s

V.  Institutionalizing Formal Equality We restrict the analysis in this section to questions of formal equality, leaving questions of effectiveness to a later discussion. Formal equality entails equal treatment in all democratic procedures. Here we analyze some of the ways in which formal rules, either in the form of constitutional provisions, statutes, or informal organizational norms, establish procedural equality. Social choice theory provides a systematic analysis of the normative and 29 analytical properties of voting procedures. No comparable analysis exists for institutions (as opposed to the ideals) of deliberative democracy. In this section and the next, we take social choice analyses as a counterpoint. We state three conditions that social choice theorists impose in order to ensure formal equality in voting. We then explore some difficulties involved in establishing analogous criteria for ensuring formal equality among parties to political argument. Such criteria are necessary just insofar as we have insisted that not just any form of political argument or debate constitutes democratic deliberation. Social choice theorists show that, under specifiable conditions, all known aggregation procedures suffer from important, unavoidable, endogenous problems. Most famously, Arrow demonstrated that there exists no aggregation mechanism that simultaneously conforms to a set of several relatively unobjectionable normative conditions and generates coherent collective decisions. Three of Arrow’s conditions aim to ensure one or another aspect of procedural equality. First, unrestricted domain disallows any prior constraint on the content of the preferences or interests that a proposed aggregation procedure must accommodate. It simply requires that the aggregation procedure itself not impose ex ante filters on the substantive views of relevant constituencies. In this sense, unrestricted domain governs what we will refer to as “conditions of entry.” Second, anonymity requires that all voters be treated equally by the 30 voting procedure. Third, neutrality requires that the voting procedure not be 31 biased toward one or another alternative. So, where anonymity requires that voters be treated equally, neutrality requires that the alternatives over which For extensive discussions see Sen (2003). This condition can be weakened to demand only nondictatorship; this ensures that there is no individual who can unilaterally determine the social choice. Arrow’s result holds even with this weaker requirement. 31   This condition can also be weakened with no damage to Arrow’s result. All that is required is nonimposition, which simply means that, for any pair of alternatives X and Y, there is some array of preferences within the relevant constituency such that X defeats Y and another array such that Y defeats X. 29   30  

211

C h ap te r 7

they vote be treated equally. In this sense, anonymity and neutrality govern different aspects of what we will call the “internal workings” of democratic arrangements. We first argue that for political argument to approximate democratic deliberation, entry to relevant institutions must be governed by normative criteria very much like unrestricted domain. We then turn to the conditions analogous to anonymity and neutrality that govern the internal workings of those institutions. Unrestricted Domain. Because “the people” who populate any plausible democratic arrangement are heterogeneous across multiple dimensions, advocates of deliberation must subscribe to some principle very much like unrestricted domain.32 Deliberative democracy requires the most expansive possible conditions of entry to formal or official political arenas.33 As noted earlier, such entry must be available at both agenda-­setting and final decision-­ making stages. In order to provide a sense of what such expansive conditions of entry require, we explore two useful examples of how not to proceed here. We draw both examples from the work of John Rawls. First, Rawls insists that parties to deliberation must subscribe to “precepts of reasonable discussion.”34 Second, he claims that democratic deliberation need be responsive not to “the fact of pluralism” per se but only to the less expansive “fact of reasonable pluralism.”35 We examine these examples in turn. Precepts of reasonable discussion, according to Rawls, enjoin parties to political deliberation from accusing “one another of self-­or group interest, prejudice or bias, and of such deeply entrenched errors as ideological blindness and delusion.” Charges such as this, according to Rawls, amount to “a declaration of intellectual war.” We must instead be prepared to countenance deep, perhaps insurmountable, disagreement while at the same time attributing to others “a certain good faith.” There are at least three reasons why one might object to Rawls’s precepts. First, political actors may, in fact, be driven by self-­interest, blinded by prejudice, or For a more fully developed argument on this point, see J. Johnson (1998). The remainder of this section draws on this essay. 33   Note that we are here concerned with formal or official decision-­making arenas. There may be good reason to restrict entry to the sort of secondary associations necessary to a robust deliberative democracy. This is because such secondary associations—­if they are to afford a secure milieu within which particular, especially disadvantaged, constituencies might articulate interests and perspectives—­may need to exclude nonmembers (Fraser 1992). Any such practice of exclusion, however, may be challenged and hence require justification within the formal or official institutions within which secondary associations operate and from which they derive legal standing. 34   Rawls (1989, 238–­39). 35   Rawls (1993, 36–­37, 58f). 32  

212

Institu tio n a liz in g L ib e ra l G u a ra nte e s 36

deluded by ideology. It very plausibly is among the desirable features of democratic deliberation that it allows participants to raise this possibility, to challenge those to whom they believe the charge applies, to do so publicly, and thereby 37 to invite those so challenged to respond. Thus, if in a deliberative decision-­ making process, you press claims that, knowingly or not, are prejudiced, ideologically biased, or unjustifiably self-­interested, others must be allowed not only to contest those claims but, Rawls’s precepts notwithstanding, to characterize 38 them as prejudiced, ideological, or selfish as part of their reason for so doing. Second, there surely are points when seemingly “unreasonable” factors—­ such as, for instance, anger, frustration, humor, fear, joy, or humiliation—­ quite reasonably and justifiably enter political argument. Should political actors motivated by such emotions be disallowed, for that reason, from participating in democratic deliberation? There seems to be no good reason why we should ground our vision of deliberation on the sort of conceptual gerrymandering that draws hard-­and-­fast boundaries between reason and emotion when such boundaries likely would banish not only obstreperous demands 39 and angry shouts but tears and laughter from the deliberative arena. Finally, Rawls’s precepts in all likelihood would disallow important political practices. Civil disobedience, for example, seems, and indeed often is designed to be, paradigmatically “unreasonable.” But it surely is plausible to see civil disobedience as part of an ongoing process of political deliberation. “Outrageous” and “unreasonable” acts of civil disobedience might, by demonstrating the depth of grievances or of outrage, prompt relevant political actors 40 to reconsider and perhaps revise an otherwise binding collective decision. Consider now the claim that deliberative procedures need only accommo41 date the fact of “reasonable pluralism.” This claim prejudges in an unjustifiRawls (1993, 58) admits as much, but claims that these are “sources of unreasonable disagreement” and so, in keeping with his views on “reasonable pluralism” (which we address below) would not emerge in democratic deliberation. 37   This benefit potentially accrues to both sides in any such exchange. For deliberation on this view does not only allow participants to challenge selfishness, prejudice, and ideology. It also provides a check from which any “reasonable” party to deliberation might derive, if not exactly enjoyment, at least benefit. It might prompt such parties to reexamine their views and commitments in order to ensure that the charges are unsound. Christiano (1996, 259) makes a similar point. 38   Part of the “reason” here will require that the accusing party establish why she believes that her characterization is justified. 39   For qualms regarding such gerrymandering, see Rorty (1985). 40   See Habermas (1985). It should be noted that in making his case Habermas draws directly and sympathetically on Rawls’s own earlier account of civil disobedience. 41   Rawls (1993, 281f) attributes the distinction between pluralism per se and “reasonable” pluralism to Joshua Cohen. 36  

213

C h ap te r 7

able way the question of which sorts of argument or value legitimately are admissible to the process of political deliberation and debate. Here, in contrast to the precepts of reasonable discussion, Rawls does not simply demand that the parties to deliberation adopt a civil demeanor. Rather, he demands some prepolitical normative criteria to which parties to deliberation subscribe and that enable them to recognize as reasonable some range of possible sorts 42 of claim or position that they do not merely tolerate but treat with respect. Other claims and views are inadmissible. Rawls, of course, claims that his “political liberalism” requires an “overlapping consensus” on the principles of justice that govern the “basic structure” of society and that such a consensus need accommodate only “the fact of reasonable pluralism.” In this sense, his conception of justice would emerge from deliberation over “matters of constitutional essentials and basic justice.”43 In a democracy, however, deliberation would address more mundane policy issues as well. Defenders of Rawls suggest ways in which deliberation informed by a concern to accommodate “reasonable pluralism” might operate at this second level. On their account, “actual deliberation” consists in “the give and take of argument that is respectful of reasonable differences.”44 Mutual respect entails in particular ways, both “integrity” in advancing one’s own position and “magnanimity” in characterizing the positions of others.45 This, however, leaves poorly defined the criteria we should use to determine what constitutes a position that is “reasonable” and hence deserving of respect. At this juncture there are two dangers that are not simultaneously avoidable. On the one hand, any set of substantive criteria for distinguishing reasonable from unreasonable arguments risks unwarranted arbitrariness and introduces an odd circularity into the defense of deliberation. First, imposing substantive standards of reasonableness on the sorts of views that are admissible to deliberative arena risks being unacceptably arbitrary.46 The same “burdens of judgment” that, according to Rawls, ensure an irreducible plurality of interests, attachments, and commitments in any democratic constituency would make it at least difficult, probably impossible, to discern with confidence, whether 42   On Rawls’s account, “reasonable pluralism” appears as prepolitical insofar as he sees it as a way of accommodating his conception of justice “not so much . . . to the brute forces of the world but to the inevitable outcome of free human reason” (Rawls 1993, 37). 43   Rawls (1993). 44   Gutmann (1993, 197). 45   Gutmann and Thompson (1990, 78f). 46   Thus, even Rawls concedes that his theory of justice, because it accommodates only the “fact of reasonable pluralism”—­in contrast to the “fact of pluralism” per se—­”runs the danger of being arbitrary and exclusive” (Rawls 1993, 59).

214

Institu tio n a liz in g L ib e ra l G u a ra nte e s

the views advocated by any particular actor are reasonable or otherwise.47 Second, we encounter the problem of circularity. From a deliberative perspective, political decisions are legitimate, when they are, because they have survived a process of reasoned argument. Imposing substantive criteria of reasonableness as an ex ante filter on admissibility would preempt that very process of reasoned argument. Instead of generating outcomes that are legitimate because they emerge from reasoned debate, substantive criteria would circumscribe in advance the range of views on offer in deliberative arenas. If only reasonable views enter the deliberative process, how can the view that ultimately emerges be otherwise than reasonable? On the other hand, one might adopt fairly minimalist, largely formal criteria for differentiating reasonable from unreasonable views. Consider one such effort that asserts that “reasonable is defined . . . in terms of a willingness to entertain and respond to objections” and that “to be unreasonable” is, by contrast, to “favor institutions and policies that cannot be justified to others.”48 The problem is that this criterion of reasonableness may prove too weak and so may generate perverse results. Consider, for example, Americans who reject the theory of evolution and subscribe instead to “creationist” views and who, moreover, wish to implement policies that mandate the teaching of “creation science” alongside of, or instead of, the theory of evolution in public schools. A minimalist approach would not exclude such views from the deliberative arena because creationists, in fact, are reasonable in precisely the sense that the minimalist conception requires. They entertain and respond to 49 objections, and they revise their views accordingly. Conversely, it is possible 47   These “burdens of judgment” derive, according to Rawls, from six unavoidable features of political decision-­making. So, in the context of reaching a decision on almost any issue, one or more of the following conditions will hold: (a) empirical evidence will be complex and conflicting, (b) different parties will accord differential weight to relevant considerations even when they agree on which are ‘relevant,’ (c) moral and political concepts will be ambiguous and their range of applicability is contestable, (d) different actors will bring different perspectives and experiences to bear on the assessment of evidence and arguments, (e) the multiple normative considerations adduced for or against an issue by participants will have differential force, and (f) the competing values raised in the decision-­making process will not be fully reconcilable (Rawls 1993, 54–­58). 48   Cohen (1994a, 1537–­38). 49   “If the claims of modern day creationists are compared with those of their nineteenth-­ century counterparts, significant shifts in orientation and assertion are evident. One of the most visible opponents of creationism, Stephen Gould, concedes that creationists have modified their views about the amount of variability allowed at the level of species change. Creationists do, in short, change their minds. Doubtless they would credit these shifts to their efforts to adjust their views to newly emerging evidence, in what they imagine to be a scientifically respectable way” (Laudan 1996, 224).

215

C h ap te r 7

to view the opponents of creationism as unreasonable in precisely this minimalist sense because they advocate policies (e.g., the teaching of evolution 50 rather than creation science) on rigid and hence unjustifiable grounds. In the end, both of the types of ex ante Rawlsian constraint on admissibility that we have examined seek to establish as a condition of entry into deliberative institutions what are more plausibly seen as potential products of deliberation. Thus, where Rawls makes adherence to his “precepts of reasonable discussion” a condition of entry into relevant deliberative arenas, we suspect that deliberation, governed by the sort of expansive conditions of entry that we endorse, might, where it is successful, engender “good faith” by enabling participants to develop greater understanding of and trust in both one another and the deliberative process itself. Similarly, deliberative procedures subject to the widest possible terms of entry would make “reasonable pluralism,” where it were pos51 sible, an outcome of, rather than a precondition for, democratic deliberation. However, it is not sufficient to ensure expansive conditions of entry to deliberative arenas. It also is necessary to ensure that once various participants and the competing positions they endorse have gained admission to deliberative institutions, the internal workings of those institutions do not accord differential advantage either to particular participants or to their favored positions. Recall that in their analyses of aggregation mechanisms, social choice theorists impose the conditions of anonymity and neutrality in order to accord equal protection to voters and alternatives respectively. Now we explore analogous conditions for deliberative arrangements. We consider the analogues to anonymity and neutrality in turn. 50   Laudan (1996, 223–­30). Characterizing the ruling of Judge William Overton in McLean v. Arkansas, which portrayed creationism as nonscientific and, on that basis, excluded it from the public school curriculum in the state, Laudan remarks: “His obiter dicta are about as remote from well-­founded opinion in the philosophy of science as creationism is from respectable geology” (Laudan 1996, 227). 51   Indeed, in comments on an earlier version of this argument, several individuals independently suggested that we might interpret unrestricted domain to require simply that while all views must be admissible to relevant deliberative arenas, some admissible views might be accorded little or no weight by parties to deliberation. This is a plausible interpretation. But note that on this interpretation, “reasonable pluralism” no longer operates as an ex ante filter. It instead emerges within the process of deliberation itself. And admissible views can only, in that context, be accorded little or no weight for reasons. In this sense, our insistence that deliberative arrangements must meet something like the condition of unrestricted domain provides a vantage point from which to clarify the conception of reasonableness operative in defenses of democratic deliberation. Note also that this puts additional pressure on defenders of deliberation to identify the mechanisms at work in the exchange of reasons in deliberative interactions (Johnson 1998). Bohman (1996b) makes some important progress on this latter task.

216

Institu tio n a liz in g L ib e ra l G u a ra nte e s

Anonymity. An aggregation procedure treats voters anonymously when it operates “blindly” in the sense that while (in order to identify a winner) it differentiates the content of votes cast, it does not discriminate among voters on the basis of characteristics such as socioeconomic position or religious affiliation. This is especially important to the goal of treating political actors equally because it minimizes the chance that they can be identified as targets of coercive interference. Historically, something like this view has been institutionalized in the practice of the secret ballot.52 Yet because deliberation centrally involves debate over social and political practices and policies, secrecy seems a particularly inappropriate condition to impose on deliberative arrangements. We want to ensure, after all, that decisions are actually informed by and result from debate rather than simply imposed by one or a few well-­placed parties. To this end, deliberative procedures rely on public contest of reasons as a way of checking power and, thereby, ensuring that participants are treated equally.53 Deliberation justifiably discriminates among divergent views based on their content in the sense that parties to deliberation will identify some views as more defensible or justifiable than others. Deliberative arrangements, however, “do not single out individuals.”54 Instead, deliberative arrangements seek minimally to dampen, and optimally to eliminate entirely, any arbitrary inequalities among participants to any interaction. They do so by ensuring that, in articulating and defending their views, participants rely not on asymmetries created, for example, by socioeconomic resources or political power 52   Obviously, voting does not entail such secrecy on all accounts. In his Considerations on Representative Government, John Stuart Mill, for example, opposes the secret ballot. He argues that in order to ensure that, when participating in an election, each citizen would “consider the interest of the public, not his private advantage,” all “voting, like any other public duty, should be performed under the eye and criticism of the public.” And he defends this view in part by claiming that secrecy no longer is needed to protect voters from powerful sources of external influence, claiming that in most countries of Europe “the power of coercing voters has declined and is declining” (Mill 1861/1991, 355–­56). Mill is not simply overly sanguine here. He arguably is also inconsistent. Just two years earlier in On Liberty, he had insisted that liberty most needed protection not from “political despotism” (however important that threat might be) but from the ubiquitous “social tyranny” of prevailing opinion (Mill 1861/1991, 8–­9). Contemporary critics of the secret ballot (e.g., Brennan and Pettit 1990) are no more convincing than is Mill on this score. 53   As we note below, however, the substantive demands of our conception of equal opportunity of political influence may, under some circumstances, require political institutions and policies at some variance with the sort of procedural equality depicted here. Put otherwise, the sort of substantive equality required by democratic deliberation may well demand policies that treat participants differently in ways that procedural equality seemingly prohibits. 54   Cohen (1989a, 22).

217

C h ap te r 7 55

but instead only on what Habermas calls “the force of the better argument.” This means that the procedures that govern the deliberative phase of democratic decision making protect equality by ensuring that all claims and counterclaims are subject to critical public scrutiny and that, when challenged, 56 any participant must defend her proposal or back her objection with reasons. This portrait will come as little surprise to those familiar with the literature on deliberation even if the connection between equality and the requirement of public scrutiny typically is not drawn out in the way we depict it here. Consider two things that this commonplace view of deliberation entails or, more precisely, two things that it does not entail. First, this commonplace view of deliberation does not require that individuals actually participate equally. Thus, for example, it does not prejudge the relation between deliberative democracy and representative institutions. It does not, in other words, presume that deliberation requires direct democracy. Among the “principles” that historically have animated the practice of representative government since the eighteenth century are the following: (1) representatives are elected by and therefore accountable to constituents, and they thereby must be able to defend their actions to those constituents; (2) representatives retain discretion and hence are called upon to exercise judgment where this may require that they set aside parochial interests of their constituents; (3) freedom of expression allows for constituents to prevail upon and express opinions to officials; and (4) legislation is enacted only after discussion and debate.57 Thus, at each point, representative institutions presuppose the sort of public debate that characterizes deliberation. The relation between deliberation and equality simply means that representative institutions must be organized in such a way that they do not single out individual representatives 58 or their constituents in unjustifiable ways. Similarly, the commonplace view of deliberation does not presume that citizens are literally equal in the sense that each has the requisite interest, experience, or expertise to participate in every decision that affects her life. It does 55   Brighouse (1996, 125) “embraces the idea that when inequalities on influence have their source in the persuasive presentation of good evidence and argument they are acceptable.” We would agree—­with the very important caveat that we explore below—­that deliberative arrangements presuppose policies to ensure that parties to democratic deliberation have the capacities they need if they are to be effective participants. 56   It is important to reiterate here that we understand democracy in the narrow sense laid out in the introduction. 57   Manin (1994, 136–­47). 58   The same, it should be noted, is true of the various secondary associations that constitute the environment for “official” deliberative institutions. There, members cannot be singled out.

218

Institu tio n a liz in g L ib e ra l G u a ra nte e s

not, for instance, preclude authority relations. This is especially important in complex, functionally differentiated societies.59 What the egalitarian thrust of public argumentation requires is that claims to authority are subject to challenge. Thus, for example, on contested issues, those who invoke special expertise can be compelled to provide reasons for a given decision when, in the standard course of events, their claim to authority might well have sufficed.60 Second, the view of deliberation as consisting in a public process through which policy proposals are advanced, challenged, and defended and where, at each stage of that process, participants must rely solely on reasons does not categorically exclude either self-­interested claims or the conflicts that such claims might generate from the range of topics admissible to relevant deliberative arenas. Members of previously excluded groups, for example, typically demand entry into relevant decision-­making arenas precisely because, so long 61 as they remain excluded, their interests are not adequately considered. There is no reason to think that this should be any less true of deliberative arrangements than, for instance, of electoral systems. It similarly is easy to envision situations where self-­interested claims are 62 justified even among long-­established participants in deliberation. Consider a common problem of public-­goods provision modeled in simple game-­ theoretic terms as a “chicken” interaction. Here two or more actors are in a situation where any one of them can supply the requisite level of a public good (say maintenance of flood-­control levees along a stretch of river). If no one provides the good, the consequences are potentially disastrous for all (say massive flooding in the spring). Each player nevertheless prefers that someone else perform the required task. Indeed, the equilibria in such a game involve outcomes in which some actor provides the public good while others exploit her cooperative activity. Faced with a strategic situation of this sort, any party to deliberative proceedings aimed at resolving it clearly could, with justifica63 tion, object to being exploited in this way. This sort of objection surely represents a justifiable response to predictable demands that some particular actor simply should sacrifice in the common interest. And if the group were to try to arrange some “fair” resolution (e.g., rotation of burdens over time), any such Warren (1996, 46–­48). Warren (1996, 58–­59). 61   See, e.g., Phillips (1995). 62   This example is taken from Johnson (1998). 63   We assume here that the relevant actors are equally endowed with the resources needed to provide the public good. That is, there is no obvious asymmetry of, for example, wealth, experience, or power. Obviously, inequalities on any of those dimensions (or others) might well justify the sort of exploitation depicted in the text. 59   60  

219

C h ap te r 7

resolution would have normative force precisely because self-­interest-­based claims provide, as it were, a significant part of the normative scaffolding in terms of which fairness can be defined. Neutrality. An aggregation mechanism treats alternatives neutrally when it is not biased, ex ante, for or against any particular alternative. The results from social choice theory are instructive with regard to this condition. These results suggest, among other things, that (1) different methods of counting votes (e.g., simple majority, majoritarian methods such as the Condorcet rule, or positional methods such as plurality voting) can generate dramatically different outcomes from the same initial profile of preferences and (2) as a result, not only are the outcomes of aggregation especially vulnerable to manipulation, but we typically are unable to differentiate outcomes produced by agenda control or strategic voting from those that are not.64 Aggregative arrangements thus have a difficult time conforming to the demand for neutrality. Unfortunately, deliberation encounters very similar problems. Although the literature on this issue is significantly less extensive, there is very good reason to suspect that the outcome of political debate depends heavily upon factors such as the sequence in which participants speak and the point at 65 which debate is terminated. This leaves those concerned with the problem of designing plausible deliberative institutions with an unenviable, perhaps insurmountable, task. If the different procedures that might govern political argument generate widely different outcomes from the same initial range of views, deliberation is susceptible to objections analogous to those that have been leveled at aggregation mechanisms. The outcome of deliberation is then hostage to precisely the sort of arbitrary factors for which aggregation has re66 peatedly been criticized.

Riker (1982). For a brief review, see Austen-­Smith (1995) and, more recently, Landa and Meirowitz (2009). The termination point is an important parameter here insofar as we view deliberation as a form of political decision-­making; a decision requires that at some juncture, argument will cease and a choice will be made among feasible alternatives. And although any such choice is revisable, deliberation is not merely ceaseless, aimless conversation. 66   This arbitrariness is most evident in noncooperative bargaining models. But note that it does not presuppose either that relevant agents are driven primarily by narrow interests or that the interests of parties to discussion conflict in extreme ways. As Austen-­Smith (1995) makes clear, matters of sequence and termination also are important in so-­called cheap-­talk models where communication is not directly payoff dependent and where players typically are concerned to coordinate their efforts, even if they differ on just how to attain coordination. Farrell and Rabin (1996) review this game-­theoretic literature. Johnson (1993) draws some initial connections 64   65  

220

Institu tio n a liz in g L ib e ra l G u a ra nte e s

Advocates of deliberation, of course, might view this apprehension as entirely misplaced. After all, they might claim, deliberative arrangements are intended to prompt participants to revise their views in light of reasons that they encounter in the course of public political argument. From this perspective, whether or not deliberative procedures generate different outcomes from the same initial range of views is irrelevant. Yet this response is troubling. For it suggests that we have no reliable criteria for determining whether any given deliberative arrangement treats alternatives equally. That hardly provides a robust basis for asserting, for instance, that the “alternatives to deliberation” represent “less moral or more authoritarian ways of dealing with fundamental moral conflicts concerning social justice.”67 In short, the condition of neutrality raises issues of institutional design about which it is difficult to say very much with any great confidence. What we do know, however, leads us to suspect that the sort of formal equality required to structure political argument into democratic deliberation may be difficult to implement precisely because it will be extremely difficult to determine whether any particular institutional arrangement in fact embodies anything like the requisite sort of formal equality. between cheap-­t alk games and the sorts of mechanisms that might sustain the force of reasons in democratic deliberation. 67   Gutmann (1993, 202).

221

Chapter 8 Substantive Conditions: Pragmatism and Effectiveness

I.  Introduction: Democratic Effectiveness Institutional guarantees of formal participation are a necessary feature of any adequate account of the requirements of freedom and equality in a democratic society. This is about as far as most liberal justifications of democracy take us. Protection from direct interference and procedural equality are clearly important aspects of these requirements. However, these protections alone will not produce the kind of political participation we envision in our pragmatist account of the priority of democracy. This priority rests on the idea that socially beneficial outcomes are a product of democratic processes in which a diverse array of ideas and beliefs compete for collective acceptance. The only test of the ultimate quality of a proposal is whether it is freely accepted over time by the members of the society who themselves enjoy equal freedom. Guarantees of free and equal participation, like those we depicted in the preceding chapter, are intended to maintain the integrity of the competition and to constrain distorting influences on the process. For democratic competition to be effective, individual participants must be free to (1) contribute their ideas and express their beliefs in the input side of the process and (2) make uncoerced choices as participants in determining the outcomes of the process. Constitutional and statutory protections establish the formal institutional context in which deliberation and elections take place. While necessary, however, they are not sufficient to sustain the effective performance of democratic institutions. For there are other types of factors that might infringe on the types of individual contributions necessary for the democratic process to have the desired effects. Effective institutional performance trades on a causal relationship between individual participation and normatively legitimate collective outcomes. The requirements of this causal relationship dictate that we establish conditions necessary to create an institutional environment that allows for the highest probability of achieving the desired effects. Here the requirements of effective institutional performance dictate that we attend to substantive

P rag ma tism an d E ffe c tive n e ss

aspects of effective individual participation and its preconditions—­freedom and equality.

II.  What Does Effective Participation Entail? Sen offers some guidance in answering this question in his discussion of the role of freedom in the context of social and economic development.1 He argues that “[t]he success of a society is to be evaluated, in this view, primarily by the substantive freedoms that the members of that society enjoy.”2 He identifies freedom as “a principal determinant of individual initiative and social effectiveness.”3 In doing so, he clearly introduces a connection between effectiveness and social outcomes. Sen distinguishes two concepts of freedom: freedom to act and freedom to achieve. Freedom to act is defined as “a person’s autonomy in the form of being able to do what she wants, and her immunity from interference by others.”4 This is basically what we have described as formal freedom, protections from direct interference with an individual’s actions. Freedom to achieve is a more encompassing concept, relating freedom “to what a person is free to have or achieve—­on the basis of her own actions and those of others.”5 Sen introduces this second alternative as a way of getting at the importance of having a real, and not merely a formal, chance of achieving one’s goals. This emphasis prompts him to argue for a conception of freedom that incorporates two basic elements: “the processes that allow freedom of actions and decisions, and the actual opportunities that people have, given their personal and social circumstances.”6 On this conception, individuals can be deemed unfree if they suffer from either inadequate processes or inadequate opportunities. This conception places the burden of determining effectiveness on the analysis of relevant opportunities. Sen poses the task of the analysis as follows: “In assessing opportunities, attention has to be paid to the actual ability of a person to achieve those things that she has reason to value. In this specific context the focus is not directly on what the processes involved happen to be, but on what the real opportunities of achievement are for the persons involved.”7 Sen (1999, 2002). Sen (1999, 18). 3   Sen (1999, 18). 4   Sen (2003, 596). 5   Sen (2003, 597). 6   Sen (1999, 17). 7   Sen (2003, 10). 1   2  

223

C h ap te r 8

What is clear from Sen’s discussion is the importance he accords to the close connection between freedom and the ability to achieve one’s preferences. Some argue that Sen’s conception of freedom entails a causal relationship 8 between preferences and outcomes. In principle, this is consistent with our own desire to maintain a causal connection between individual participation and effective institutional performance. However, the connection to preferences raises a question as to how far the analyses of opportunities will lead to an assessment of actual achievement. Sen asks, “[I]f the opportunity aspect of freedom is judged in terms of preferences (no matter whether one’s preference is confined to one’s self-­interest or not), can the assessment of opportunity differ in any way from achieving simply what is preferred? Can there be a dichotomy between achievement and opportunity, even when both are judged in the light of the relevant person’s preferences?”9 In the context of an analysis of social and economic development, Sen offers somewhat contradictory interpretations of how far he would draw the connection between effectiveness and achievement. On the one hand, he wants to avoid any conception of freedom that will equate it with success. He explicitly distinguishes freedom to achieve from actual achievement.10 On the other hand, he claims that an individual could be considered free in the relevant ways if she were able to achieve certain outcomes, like a healthy and nutritional diet, even if the outcomes were brought about by the actions of others on her behalf. For the purpose of arriving at a conception of freedom relevant to democratic decision making, it is imperative to retain the distinction between opportunities and achievement. We have earlier defined democracy fundamentally in terms of the uncertainty of collective outcomes. These outcomes are the product of the interdependent choices of all of the actors in the decision-­ making process. No one individual can ever be guaranteed that she will get her interests instantiated in democratic outcomes. This suggests that an adequate conception of effective democratic participation cannot be grounded in the actual achievement of any particular individual’s goals. Thus, effectiveness in the democratic process cannot entail successful outcomes. It can, however, relate in important ways to real possibilities of influence. And the real possibility for political influence is a function of the opportunities enjoyed by individual political actors. The goal of effective institutional performance makes strong demands on our criterion of political participation. Participants in democratic processes Pettit (2001). Sen (2003, 15–­16). 10   Sen (1999, 18). 8   9  

224

P rag ma tism an d E ffe c tive n e ss

actively engage in political argument. The task for any participant in such a process is to advance and entertain proposals—­reasons for action—­that will influence others to endorse her preferred collective outcomes. In other words, not only does voting hardly exhaust the avenues of possible influence, it does not exhaust the modes by which participants can influence democratic politics. The relevant difference is captured in the following helpful distinction between influence and impact: “The intuitive difference is this: someone’s impact in politics is the difference he can make, just on his own, by voting for, or choosing one decision rather than another. Someone’s influence, on the other hand, is the difference he can make not just on his own, but also by leading or inducing others to believe or vote or choose as he does.”11 If an individual participant, for whatever reason, is denied the opportunity to influence others, she will be unable to affect the collective decision-­making process. Through this failure, her contribution to the democratic process is diminished, as might be the quality of the collective outcome. Her interests and goals will likely go unaccounted for, thus violating the fundamental notion that democratic outcomes are the product of the freely articulated interests and concerns of equal citizens. It is important here to note how the requirements of effective participation highlight the interrelated nature of the ideals of freedom and equality in a democratic society. Effectiveness entails that each individual must have effective opportunities of political influence and that these opportunities must be distributed equally. Hence our focus on the necessary criterion of equal opportunity of political influence. In the next two sections, we explicate and defend this conception. We organize our discussion around the dimensions of freedom set out at the beginning of chapter 7. We argue that effective participation entails attention to issues of (1) indirect interference in horizontal relationships, (2) social norms and customs, and (3) the capacities of individual participants. With each of these factors, we assess the implications for voluntary and uncoerced participation. In doing so, we mean to call attention to the increasingly controversial implications of including such factors in the debates over the requirements of political freedom. We also address what is necessary to guarantee that all citizens experience such participation equally. To adequately secure equal opportunity of political influence, we must attend to the substantive aspects of political equality. Cohen identifies the substantive dimension of equality for deliberative democracy as follows: “The participants are substantively equal in that the existing distribution of power and resources does not shape their chances to Dworkin (1987, 9).

11  

225

C h ap te r 8

contribute at any stage of the deliberative process, nor does that distribution play an authoritative role in their deliberation.”12 This claim rightly highlights the fundamental connection between the social distribution of power and resources and the achievement of real political equality. The connection is a complex one. To assess the requirements of political equality for democracy, we must explore the various ways this social distribution of power and resources influences effective participation in the deliberative process. This, however, raises important questions about the limits to which the state can go to guarantee these forms of protection.

III.  Institutionalizing Effective Freedom and Equality I: Social Factors Horizontal Relationships. Democracy’s normative force is derived in part from the basic premise that individual participants will treat each other as free and equal political actors. From this it does not necessarily follow that all citizens must be substantively equal or even that they be treated equally in the nonpolitical aspects of social life. There may be other arguments that justify a broader criterion of equality in social and economic spheres, but such a criterion does not follow directly from the justification of political democracy. Nonetheless, the circumstances of social and economic life do become relevant to our concerns here insofar as they significantly affect the interactions among participants in the political arena. In democratic politics, participants assess arguments offered in the course of political debate and then decide how they want to vote on the various policy questions. Effective participation requires that when the time comes to make their final decision on a question, individuals base their choice on what they think is the force of the better argument and not on other extraneous factors. To be able to do this, individuals must be protected from various forms of interference. The guarantees of formal freedom address direct interference by state actors as well as by other participants. Yet there are other ways in which social interactions among participants can interfere with the decision-­making process. Here we have in mind two forms of behavior that might negatively influence an individual participant’s contribution to the effectiveness of collective decision making: (1) indirect interference with political choice by means of threat to injure in nonpolitical contexts and (2) self-­censorship due to the anticipated negative reaction of other participants to one’s political choices. Cohen (1989b, 33).

12  

226

P rag ma tism an d E ffe c tive n e ss

We discussed these factors earlier. In both instances, the key to understanding the nature of the negative effect is found in a focus on the expectations that individuals develop about what other people will do to them and how they might do it. The relevant causal issue is, what factors make the threats that produce these expectations credible? That is, why do individual participants believe that others will actually inflict costs on them if they make certain political choices? The answer lies in the asymmetries in the distribution of power and resources that exist in nonpolitical spheres of social life. These asymmetries lend credibility to such threats by increasing the likelihood that the threatening party will be successful in inflicting the anticipated costs if the threat is in fact carried out. When these asymmetries are clear and thus the expectations are stable, the threat of interference will seldom have to be employed because the constraining effects of self-­censorship will be all that is necessary to produce the desired effect.13 In order to guarantee voluntary and uncoerced participation, it is necessary to find ways to neutralize the causal effect in the political arena of these nonpolitical asymmetries. Here effective institutional performance requires institutional arrangements that constrain any actor who might seek to exploit an advantageous asymmetry for purposes of coercing other participants. There are, in principle, many factors in social life that might fall within the category of relevant powers and resources. The primary question here is which of these factors should be governed by the conception of equal opportunity of political influence. To fully answer this question, we must first determine which factors are politically relevant and then assess which asymmetries in the distribution of those factors warrant corrective measures to address inequalities. Arguably, the primary factor threatening the prospects for uncoerced political decision making is pervasive asymmetries of material resources and the social power that follows from them. Individuals who enjoy an advantage in the distribution of material resources can affect the democratic process through the ability to both make promises and issue threats that this material advantage affords them. Promises work through the benefits that they can provide other citizens in exchange for favorable support on particular policy decisions. Explicit bribery is generally precluded by law, but more implicit forms of trading votes for benefits are the subject of standard criticisms of most modern representative democracies. These implicit trades violate the notion of equal opportunity of influence.14 To constrain such trades, we must establish rules Pettit (2001, 136). There is a clear resemblance between our argument here and Walzer’s (1982) discussion of the importance of devising mechanisms that will block certain sorts of exchange. 13   14  

227

C h ap te r 8

prohibiting the use of material advantages in the political process. Proposals range from campaign finance restrictions on private contributions to complete public financing of political competition to, conversely, measures that would place all campaign contributions behind a veil of secrecy that would ensure that they remain anonymous, thereby depriving contributors of the ability to 15 claim credit and candidates to sell access. While any constraints would diminish the effects of resource asymmetries on deliberation, equal opportunity of influence requires a significant wedge be driven between resources and influence. From a pragmatist perspective, the best way of pursuing this goal will be experimental; it would examine various institutional mechanisms and determine which among them most effectively diminishes the impact of money in democratic processes. However, even this form of extensive public regulation of political competition will not adequately address the effects of threats supported by resource asymmetries. Explicit threats seek to intimidate citizens by invoking economic or social sanctions should they vote contrary to the threatener’s interests. Such explicit threats are usually precluded by law. But asymmetries in material resources can have a profound effect on democratic politics through what might best be conceived of as implicit threats. The best example of these effects is found in some analyses of democratic capitalist societies.16 On these accounts, capitalists have a disproportionate effect on democratic policy making because of their control of the material resources that determine the ultimate success of those policies. In terms of our analysis of the deliberative process, the effect is manifest as an implicit threat by those who control the material resources necessary for the success of democratic outcomes. Here citizens are disproportionately inclined to endorse the interests and views of capitalists because they—­the citizens—­anticipate that failure to do so will elicit an economic response from capitalists that would have severe negative consequences for the broader society. On the extent of political economic inequality in the United States, see, e.g., Piketty and Saez (2003), Mishel, Bernstein, and Shierholz (2009), and Wolff (2001). On the impact of that inequality on the workings of American democratic processes, see Bartels (2008). The empirical literature here is expansive, rich, and largely beyond the immediate scope of our enterprise. For arguments that campaign finance restrictions enhance free and equal participation in the democratic process, see Sunstein (1995) and Rawls (1993). For an example of an argument that political equality would be better encouraged if the state would assume the costs of party competition, see Cohen (1989b, 40). For the proposal that secrecy is a more effective tool than publicity for protecting democratic processes, see Ackerman and Ayres (2004). 16   Lindblom (1980); Przeworski and Wallerstein (1988). For evidence that financial markets respond in a variety of negative ways to the prospect of the electoral success of left-­leaning political parties, see Herron (2000). 15  

228

P rag ma tism an d E ffe c tive n e ss

And here it becomes apparent how demanding and extensive the conditions for an effective democratic process may be. To effectively diminish the effects of such implicit threats, political institutions might well have to place 17 significant constraints on the nonpolitical use of material resources. As we discuss below, a society seeking to offset the effects of material asymmetries on effective political participation will confront the conflicts between the requirements of political equality and their commitment to freedom of action in the economic sphere. Social Norms and Customs. The protections of formal freedom are intended to guarantee voluntary and uncoerced choice throughout the democratic process. To secure these protections, we construct formal institutions whose rules define the terms of uncoerced political activity. But, as with any set of formal institutions, these democratic institutions are embedded in a social context dense with other institutional arrangements. These other institutions, both formal and informal, place their own demands on the choices of social actors. The plurality of such institutions complicates the task for an individual decision maker. Under such circumstances, individual choice must take account of the set of interrelated incentives and sanctions created by the various institutional rules. For democracy, the fact of institutional pluralism raises questions about the extent to which other social institutions might influence the effectiveness of individual participation. We can identify the effects of institutional pluralism at many stages of the democratic process. When it involves an interaction with other formal institutions, the task is one of reconciling the dictates of the different institutional rules. Although this may involve trade-­offs among societal goals, the nature of the task is straightforward: reconcile the different dictates so as to neutralize the negative effects on effective participation. We will say more about the trade-­off issue below. When institutional pluralism involves an interaction with informal institutions—­the norms and customs in a society—­the task is more complex. Sen highlights the complex relationship between social norms and freedom in his discussion of economic development. He argued that an adequate conception of freedom in the development context would have to “acknowledge the role of social values and prevailing mores, which can influence the freedoms that people enjoy and have reason to treasure. . . . The exercise Cohen (1989b) argues that these effects of material resources justify proposals for public ownership of the means of production. Short of that, it is possible to imagine both policies that place less sweeping constraints on capital mobility and measures analogous to current labor laws that prevent workers from being dismissed for union-­organizing activities. 17  

229

C h ap te r 8

of freedom is mediated by values, but the values in turn are influenced by public discussions and social interactions, which are themselves influenced by participatory freedoms.”18 The types of social norms that Sen has in mind are characterized by asymmetries in status among the members of various social groups. Examples are numerous, affecting the relative status of men and women in many societies as well as the relative status of members of different racial, ethnic, or religious groups. The potential problem for the effort to guarantee effective political participation is analogous to the problems involved in horizontal relationships. Indirect forms of horizontal interference rest on a single causal process: asymmetries of resources and power in nonpolitical spheres are used to influence the choices of actors in the political sphere. A similar causal process underlies the ways in which social norms can negatively influence political participation: the costs incurred from violating a norm in a nonpolitical sphere alter the relative value of the choices available to a participant in the political decision-­ making process. The main concern for the effectiveness of democratic institutions is that the asymmetries in status in the social domain will adversely affect the quality of individual participation in the political domain. There are many alternative definitions of social norms in the social science 19 literature. Most of these definitions have two features in common that are relevant to this analysis: norms define appropriate behavior in various social contexts, and they create expectations about the consequences of violating the standards of appropriateness. When individuals violate the behavioral dictates of social norms, they expose themselves to the costs generated by various forms of informal sanctions.20 It is in the costs of sanctioning that the indirect effects on political participation can be found. That is, the enforcement of norms can negatively influence political participation when the norms prohibit forms of behavior that would be strategically beneficial in the political decision-­making process. In such cases, social norms serve as a limitation on political participation. Adherence to such norms alters the nature of what an individual would contribute to the collective decision-­making process if she were otherwise acting freely. The cumulative effect of such limitations is to diminish the diversity of ideas and beliefs contributed to the decision-­making process and, thus, to undermine effective institutional performance. So, the fact of the matter seems Sen (1999, 9). For an informative review of the major alternatives, see the various contributions in Hechter and Opp (2001). 20   Taylor (1987). 18   19  

230

P rag ma tism an d E ffe c tive n e ss

to be established. The remaining question is, does this type of effect warrant protection by the state? Some theorists, associated with the conception of freedom as nonlimitation, argue that individuals do require protection from the effects of social norms and customs. On this account, freedom “requires the absence of noncompromising factors—­say, natural obstacles—­that condition the extent to which people can make uninterfered-­with choices; it requires, not just a capacity for acting without interference, but an environment of opportunity in which there is ample scope for the exercise of that capacity.”21 But even some who are sensitive to the need for an extensive conception of freedom would stop short of addressing these effects. They justify their resistance by distinguishing between interpersonal and impersonal effects.22 As this argument goes, interpersonal effects are more serious than impersonal ones because they are motivated by imminent conflicts of political interests and directly reflect asymmetric relationships among the participants. On the other hand, these impersonal effects, while clearly constituting a limitation on political choice, are more diffuse and less identifiably associated with specific conflicts of interests. Given the fact that these limitations are unconnected to any particular political conflict, they are sufficiently unrelated to political debate to justify their exclusion from the protections of effective participation. This counterargument highlights the difficulty inherent in drawing the appropriate boundaries on protected effects once we move beyond instances of intentional interference. The interpersonal-­impersonal distinction seems intended to draw the line in such a way as to exclude those effects that cannot be directly manipulated for immediate political advantage. At first blush, social norms would appear to fall on the impersonal side of this distinction. However, the effects of social norms readily illustrate the problems with this criterion just to the extent that two central features of those effects reflect intentional efforts to gain strategic advantage.23 First, social norms are the product of a decentralized process of prior social interactions. The most robust theory of how these norms evolve suggests that they instantiate the asymmetries in power that systematically characterize the prior interactions. In doing so, they create social advantages for certain groups in society that get reinforced through ongoing compliance with the norms. To Pettit (2001, 133). Pettit (2001, 132). 23   Our argument here parallels the case Satz (1992; 1995) makes—­in her assessment of whether to allow markets for certain sorts of “goods”—­for attending to the ways background norms influence parties to exchange. 21   22  

231

C h ap te r 8

the extent that this is the case, the effects of social norms on political participation will be neither random nor arbitrary but will rather have a systematic distorting effect on the democratic process. Second, these negative effects actually are, in a significant way, the product of intentional action: specifically, the sanctioning behavior that supports compliance with the norms. Either directly, through the implementation of the sanction, or indirectly, through the result of anticipated reaction, the interpersonal process of monitoring and sanctioning is part of the causal mechanism that produces the constraining effect of norms on political participation. It is easy enough to see how these two effects combine to establish and maintain social asymmetries that define relationships, for example, among people of different gender, race, religion, or ethnicity. This, in turn, suggests that social norms share enough of the features of interpersonal constraints to warrant inclusion among the factors against which individuals require protection. But, more importantly, direct manipulation for immediate political advantage is not the relevant issue for our analysis. Rather, the relevant issue is whether the constraining effects of social norms negatively influence individual contributions to the collective decision-­making process. And here the conclusion seems to be clear. The limiting effects of many social norms distort individual contributions and undermine effective institutional performance. Therefore, the effects of social norms on political choice should in principle be included among the factors that threaten effective political participation. Yet, the practical task of incorporating protections against these effects among the guarantees of equal opportunity of political influence raises difficult implementation questions. We might use two primary mechanisms to try to neutralize the effects of social norms. On the one hand, we can create mechanisms that impose limited prohibitions against the ways individual agents enforce norms on others and are directed at eliminating the sort of informal sanctioning that make noncompliance a costly choice. Such prohibitions would be limited to those cases in which norm enforcement infringes on political participation. For example, we could prohibit members of the community from enforcing social norms that discourage others in the community from participating in public activities. This should be an effective mechanism for dealing with the more substantial forms of sanctioning, those in which the enforcing party commits some explicit act that inflicts significant injury or harm. But such prohibitions will be less effective in counteracting some of the less obvious and explicit forms of sanctioning, acts like shunning or gossip that primarily impose psychological and emotional costs on transgressors. On the other hand, we could simply make illegal the counterproductive social norms. In one sense, this approach resembles the task of reconciling

232

P rag ma tism an d E ffe c tive n e ss

conflicts that arise among formal institutions. If one rule aims to allow free political choice and another indirectly constrains that choice, then the appropriate response would be to legally abolish the rule that produces the constraining behavior. But in most instances, it will not be as easy as this. The abolition of social norms raises an issue fundamental to the general project of establishing constitutional protections for democratic freedom and equality. Put simply, where do we draw the line between issues subject to constitutional protection and issues subject to consideration in the democratic process? This is, of course, among the most controversial issues in any constitutional democracy, leading to ongoing debates about the relative authority of the different governmental branches in a democratic society. Although we postpone a general discussion of this question until later in the chapter, note how the question of whether efforts to abolish offending social norms should themselves be constitutionally proscribed will complicate and inflame this controversy. This is a primary area where the concern to guarantee necessary preconditions of democratic legitimacy will impact the substance of social life. This will be an issue for any norm that covers all segments of society. Yet it is an especially complex matter in cases where the norm obtains only among some social groups. Intragroup norms play an important role in maintaining group identity and social solidarity.24 They often enjoy the support of the large majority of the members of those groups. At the same time, the enforcement of such norms within a social group can have the same effect on political participation as that of direct interference in horizontal relationships. As such, they undermine effective individual participation. But as with other cases of interference among political allies, many have argued that such intragroup norms may be entitled to special protection if they are necessary to maintain solidarity within politically oppressed groups. Some have even argued that due to their special endangered status, these groups should be granted a group veto over efforts of the population as a whole to abolish such norms.25 There is an obvious tension here within our account of democratic effectiveness. On the one hand, it seems inconsistent with our general approach to democracy to make the substantive abolition of these norms part of the necessary preconditions of the process. On the other hand, the norms often have the indirect effect of constraining free political participation. A more appropriate approach would be to adopt the more limited mechanism of prohibition of norm enforcement when it infringes on political participation, while acknowledging that there are types of informal sanctioning that may elude its Hardin (1995). Young (1990).

24   25  

233

C h ap te r 8

purview. This would leave the more general question of the relative merits of the norm itself firmly in the realm of democratic politics. Within this realm, however, we see no good argument to support the idea of a group veto over the abolition of the norm.26 Absent such an argument, ideas like the group veto strike us as merely another potential institutional constraint on effective political participation.

IV.  Institutionalizing Effective Freedom and Equality II: Capacities of Individuals The Relevance of Individual Capacities. Our pragmatist defense of democracy envisions active participation by citizens in processes of mutual discussion and persuasion. Such participation requires that each citizen be able to advance arguments that others might find persuasive. Thus, equal opportunity of political influence must attend to the conditions under which all citizens would be able to engage in discussion at this level. The depth of the anticipated participation highlights the importance of the effects of individual-­ level capacities for effective institutional performance. Once we begin to focus on the capacities of individual participants, we significantly expand the range of factors against which effective political participation may require protection. The nonlimitation conception of freedom suggests that this would require us to take account of the negative effects of “the natural limitations imposed by disability, illness, poverty and the like.”27 For our purposes, the primary analytical question is: what effect do these limitations on capacities have on effective institutional performance? There are two main ways to think about this. First, lack of capacity to participate diminishes the diversity of inputs necessary for effective performance. Second, asymmetries in capacity distort the testing of ideas by giving disproportionate influence to those who enjoy advantages in the relevant capacity. This would seem to establish a prima facie case for including protections against the effects of what often are seen as “natural” limitations on capacity among the normative preconditions of democracy. In the limit, this “would allow the state to do everything possible to cope with limitations of freedom due to handicap and illness, lack of education and information, insecurity and poverty, and on that account it may be found appealing; it would look for a Johnson (2000; 2002). Pettit (2001, 130).

26   27  

234

P rag ma tism an d E ffe c tive n e ss

very rich environment of opportunities.”28 But this envisions quite a pervasive involvement by state actors: “the ideal would permit the state to coerce or manipulate or force people in this or that manner—­even short of emergency circumstances where people’s life or health is at risk—­just so long as this was beneficial in the overall struggle against limitation: just so long as those interpersonal assaults on freedom promised to emancipate people from somewhat larger impersonal obstacles.”29 This vision conjures up the understandable concerns about an all-­powerful, intrusive state that is a prominent feature of standard liberal accounts of democracy. Thus, while the answer to the analytical question about the causal effect of natural incapacities seems clear, it is considerably more complicated to resolve the normative problem of establishing the necessary protections without creating other unwarranted constraints on democratic decision making. The task of implementing the conditions of effective participation has, to this point in our analysis, involved two steps: (1) identify a negative effect and (2) devise an institutional mechanism to regulate or constrain that effect. The institutional constraints, in turn, have involved variations on a single mechanism: protections against the use in the political arena of power and resource advantages that derive primarily from nonpolitical contexts. But this would be an inadequate way to address the problem on the dimension of the capacities of individuals. For, while the requirements of political equality may be merely to neutralize the negative effects of asymmetries in capacities, the criteria of effective individual participation require more than this. Equal opportunity of political influence also requires that each individual have a real opportunity for influence, and this places a stronger burden on us. Because of the requirements of effective political freedom, the task here is to address the issue of the adequacy of capacity as well as the issue of the equality of capacity. Therefore, the necessary protections on the dimension of individual capacities involve a combination of institutional protections and public policies that will achieve an adequate level of capacity for all individuals and then guarantee that equality of opportunity for political influence be maintained. To accomplish such a task, we first need to determine exactly what it is that we need to provide people and then figure out how we go about equalizing it. Rawls proposes a resource approach to equality of opportunity in a political society.30 On his account, each citizen will be guaranteed a minimum Pettit (2001, 131). Pettit (2001, 131). 30   Rawls (1993). Christiano (1996, 255) also explicitly endorses a resource-­based conception of political equality. Likewise, although he seeks to distance himself from resource-­based conceptions, 28   29  

235

C h ap te r 8

threshold of primary goods that she can use to further her political and economic goals. Justice as fairness requires nothing more for political equality. Rawls accords no explicit attention to issues of equality of capacity. He assumes only that citizens “do have, at least to the essential minimum degree, the moral, intellectual, and physical capacities that enable them to be fully cooperating members of society over a complete life.”31 He acknowledges that if this assumption about capacities is not satisfied, then the just distribution of primary goods may not accomplish the goal of political equality. Rawls emphasizes the importance of guaranteeing that political liberties be secured by their “fair value”: “this guarantee means that the worth of the political liberties to all citizens, whatever their social and economic position, must be approximately equal, or at least sufficiently equal, in the sense that everyone has a fair opportunity to hold public office and to influence the outcome of political decisions.”32 Rawls’ account of political equality does not appear—­ on his own interpretation—to satisfy the requirements of equal opportunity of political influence in a deliberative democracy. On the one hand, he conceptualizes the fair value of political liberties in terms of the “fair opportunity . . . to influence the outcome of political decisions.” On the other hand, he limits analysis of fair opportunity to the ownership of the minimum threshold of primary goods and merely assumes that actors possess the capacities needed to effectively use these resources. The differences between Rawls’ conception of equality and conceptions that revolve around equal capacities is made clear by his discussion of the way that justice as fairness would treat asymmetries in capacities that arise above the minimum threshold of primary goods. On the dimension of moral and intellectual capacities and skills, Rawls concludes that any variations above the minimum threshold are acceptable and consistent with the principles of justice as fairness.33 Ultimately, Rawls treats as an assumption what equality of capacity treats as a fundamental feature of political equality.34 Brighouse (1996, 127–­31) apparently endorses resource-­based remedies for political disabilities. Although we do not argue this here, we believe that our focus on the need to ensure that parties to deliberation enjoy equal capacities needed for effective participation rectifies the weakness that Brighouse (1996, 126–­27) attributes to opportunity-­based conceptions of political equality. 31   Rawls (1993, 183). 32   Rawls (1993, 327). 33   Rawls (1993, 184). 34   Note that Rawls’s account of political equality allows for inequalities in politically relevant capacities and resources, but any such inequality will operate in favor of the socially advantaged groups in society.

236

P rag ma tism an d E ffe c tive n e ss

Not surprisingly, Sen has attempted to offer an alternative that would capture the contingent nature of the distribution of capacities in a society. On his account, an adequate conception of equality must focus on “a characterization of freedom in the form of alternative sets of accomplishments that we have the power to achieve.”35 He gives priority to the size of the feasible set open to an individual rather than to the resources or means that she possesses. He argues that a Rawlsian focus on primary goods is inadequate because it fails to address the important variations in individuals’ abilities to effectively use these goods. The task becomes one of finding a way to equalize the size of individual feasible sets. While this alternative is an advance over the primary goods approach, there is still the problem of how Sen treats the relationship between freedom and achievement. As we discussed in the section on effectiveness, in developing his capacities approach, Sen distinguishes between freedom and control. On his account, “effective freedom” does not imply actual control: “As long as the levers of control are systematically exercised in line with what I would choose and for that exact reason, my ‘effective freedom’ is uncompromised, though my ‘freedom as control’ may be limited or absent.”36 This distinction, while defensible in some spheres of social life, seems particularly inappropriate as a conception of political equality for democratic deliberation. Here a more active conception of freedom seems more consistent with the underlying premises of deliberation. What is needed is a capacities-­based conception that retains the spirit of Sen’s argument but advances a different conception of “effective freedom.” Consider Bohman’s attempt, which is explicitly concerned with the capacities relevant to participation in a deliberative process.37 According to this argument, a focus on capacities highlights the fundamental importance of effective freedom: “[F]reedom is, on this account, the capacity to live as one would choose; it is the capacity for social agency, the ability to participate in joint activities and achieve one’s go als in them. For political liberties, the issue is effective use of public freedoms, which may be absent even in the absence of coercion or prohibitions.”38 This focus is justified on grounds similar to Sen’s: equality of resources is an insufficient remedy for deficiencies in effective participation because people differ in the capacities necessary to effectively use available resources.39 Sen (1992, 34). Sen (1992, 34). 37   Bohman (1996b). 38   Bohman (1996b, 130). 39   Bohman (1996b, 128). 35   36  

237

C h ap te r 8

This capacity-­ based conception seems better suited than narrower resource-­based conceptions to the job of capturing the effective participation requirement of democratic deliberation. It focuses efforts to equalize opportunity on those capacities that are relevant to political agency. But that is as far as it gets us. In order to fully satisfy the requirement of the inclusion of capacities in our conception of equal opportunity of political influence, we must determine which capacities are politically relevant, identify methods by which their relative distribution can be measured, and explore policy prescriptions for offsetting the effects of unequal capacities on democratic deliberation. Unfortunately, it turns out to be quite difficult to carry out each of these tasks. Consider first the task of identifying politically relevant capacities. What kinds of capacities are relevant to democratic deliberation and thus to a conception of political equality? There are a number of potential candidates. We want to highlight three. The first is the capacity to formulate authentic preferences. As we noted earlier, asymmetrical distribution of power and resources in a society undermines democratic deliberation. In part, this is because such asymmetries can induce participants to embrace “adaptive preferences.”40 On this account, the legitimacy of the democratic process rests on the idea that people act on free and voluntarily established preferences. To the extent that the preferences of citizens reflect adaptation to the diminished possibilities that, in turn, result from being disadvantaged by an asymmetric distribution of resources, then these preferences may undermine the idea of equal opportunity of political influence. The second relevant capacity relates to the effective use of cultural resources. The problem has been characterized as one that “cultural imperialism” poses for minorities.41 This concept emphasizes the fact that minorities are required to express their ideas and needs in the language of the dominant groups in society. To the extent that these minority groups are less adept or, in the extreme, wholly ineffective in using the language and concepts of the dominant groups, they will lack an equal opportunity to influence the members of the deliberative body to adopt policies that will address their particular needs. The third, and we think most important, kind of relevant capacity relates to basic cognitive abilities and skills.42 In his account of capacities, Sen highlights the problems and difficulties in acquiring the information necessary to Elster (1983). Young (1990, 133–­34). 42   Christiano (1996, 255) notes the importance of such abilities to the discussion of political equality. 40   41  

238

P rag ma tism an d E ffe c tive n e ss

diminish uncertainty and thus to make effective decisions.43 We take this to be one aspect of a more general category of cognitive capacities that is central to the deliberative process. Unless each participant has the cognitive capacities and skills necessary to effectively articulate and defend persuasive claims, then there will be no real equality of opportunity for political influence.44 Each of these three factors—­autonomous preferences, command of cultural resources, and cognitive capacities—­are relevant to a citizen’s ability to effectively influence other citizens to support her own preferred outcomes. However, an adequate account of effective political participation must take seriously the observation that “we cannot deduce what is politically fair from abstract principles of political equality: we have to draw on empirical judgments of what is likely to happen as well as what seems in principle to be fair.”45 This requires that we now consider two issues: (1) the measures of equal opportunity of influence (as understood in terms of capacities) and (2) the mechanisms by which we might institutionalize equal opportunity, taking account of these measures. The Problem of Measurement. Sen suggests that the analysis of equality in “freedom to achieve” relates to the available data on actual achievement.46 Thus, while he generally distinguishes between achievement and the freedom to achieve, he notes that any measure of such freedom rests necessarily on a consideration of the actual outcomes of relevant collective decision-­making processes. To rely on other sources of information risks delving into counterfactuals about possible outcomes that are difficult, at best, to justify. Part of Sen’s reliance on outcomes and actual achievement may be explained by his analytical focus on effective freedom and not control. Data on achievement may be sufficient to give a reliable measure of effective freedom.47 But, for reasons given earlier, Sen’s relatively simple conception of “effective freedom” does not adequately capture the active dimension of political equality. And, unfortunately, the alternative modification of Sen’s conception of effective Sen (1992, 148–­49). We also take note of Cohen’s suggestion of the importance of self-­worth: “Claims to equal political standing are fueled, too, by the connection between such standing and a sense of self-­ worth, a connection rooted in the public recognition associated with equal standing” (Cohen 1994b, 613). While we agree with the basic thrust of his suggestion, we think that successful attention to the three kinds of capacities we address here will most likely have the effect of effectively addressing the question of self-­worth. 45   Phillips (1995, 38). 46   Sen (1992, 5). 47   Cohen (1995) challenges the informational requirements of Sen’s capabilities-­approach for nonpolitical spheres. We take up this argument below. 43   44  

239

C h ap te r 8

freedom as applied to democratic deliberation makes measurement issues even more complex. Bohman offers an alternative conception grounded in the idea that political equality requires a guarantee of a minimum threshold of “effective freedom.” He argues that “[a] good empirical indicator of such a deliberative capacity is whether or not citizens or groups of citizens are able to initiate public deliberation about their concerns. This ability to initiate acts of deliberation thus provides a measurable threshold for political equality and reasonable cooperation.”48 He suggests that we look for empirical indicators that help define this minimum threshold: “Poverty in this sense is a measure of minimal political equality in a democracy: it sets the threshold requirement of publicity in deliberation in terms of the equal capacities to participate effectively. The development of such abilities is the ‘floor’ of civic equality, since they offer citizens greater possibilities of deliberative uptake for their differing reasons, some of which may not yet be publicly recognized as worthy of consideration.”49 The virtue of this account is that it establishes a standard that will recognize unacceptable inequalities in capacities in those cases in which citizens are completely ineffective, completely incapable of initiating public deliberation on their concerns: “Deliberative democracy must fulfill demands for equality in the means for effective participation at least enough so that no citizen is so poor as to fail to influence outcomes or to avoid exclusion.”50 At first blush, this capacities approach does not appear to be calibrated to actual outcomes. Nevertheless, on closer examination, it is hard to see how we can apply Bohman’s threshold measures without reference to the causal effect of individual participation on the collective outcomes of the deliberation process. An argument in support of this conception insists, “This standard does not require that particular citizens or groups of citizens can ever expect to determine the outcomes of debate and deliberation. However, it does require that whenever citizens engage in deliberation, they reasonably expect that their reasons could ultimately be adopted by their fellow citizens.”51 Without reference to prior causal effect, how would they develop those expectations? Both of these capacity-­centered conceptions illustrate the importance of examining outcomes to determine the effectiveness of participation in the deliberative process. How do we use evidence of outcomes to establish a measure of equal opportunity of influence? This involves a determination of the causal Bohman (1996b, 128). Bohman (1996b, 128). 50   Bohman (1996b, 148). 51   Bohman (1996a 14). 48   49  

240

P rag ma tism an d E ffe c tive n e ss

effect of one’s participation on the collective outcomes of the democratic process. This is exceedingly complicated in the analysis of deliberative processes. To appreciate why this is so, it again is helpful to compare the ideal of deliberative democracy with a purely aggregative democratic mechanism. With a purely aggregative decision-­making procedure, citizens vote by announcing their individual preference or judgment, and the collective outcome is the manifestation of the aggregation of these preferences. There is in this scheme a presumed relationship between initial preferences and the preferences instantiated in the collective outcome.52 The greater the deviation between an individual’s preferences and those of the collectivity, the lesser the presumed causal effect. With deliberation, however, this relationship is more complex. To the extent that democratic deliberation involves political argument under conditions of freedom and equality, one is always open to the possibility of reassessing one’s own position or views.53 As a result, one’s eventual stance as an individual member of a deliberative body may diverge in significant ways from one’s initial position. Indeed, the final collective outcome may not reflect the initial position of any member of the community in a straightforward way. Thus, the distance measure of causal effect available under the pure aggregation approach may be inapplicable for an analysis of deliberation. This highlights a striking feature of democratic deliberation: if operating effectively, it can on its own undermine our ability to assess equality of opportunity of influence by reference to the outcome of deliberative processes. Bohman’s proposal is sufficient to capture the most egregious consequences of unequally distributed capacities. Most people would agree, we think, that the dictates of political equality are violated in those cases in which the interests of a particular social group never are reflected in collective outcomes. This alternative suggests that such failure manifests the extent to which the excluded bear a disproportionate share of the costs of democratic policies: “Asymmetrical exclusion and inclusion succeeds by constantly shifting considerable burdens on the worst off, who lack the resources, capabilities and social recognition to mount a challenge to the conditions which govern As our earlier discussion of voting mechanisms makes clear, the presumed relationship need not be entirely straightforward, but the presumption holds nonetheless. And our argument for majority rule is premised to a significant extent on the fact that it is reasonable over a broad range of initial inputs (preferences or judgments) in the sense that it is responsive to those initial inputs. 53   Cohen (1989b, 34). In fact, even some social choice scholars acknowledge the possibility that such change is a fundamental feature of the democratic process. For example, very early on, James Buchanan (1954, 120) insisted that “government by discussion” is a basic element of democracy and that “individual values can and do change in the process of decision-­making.” 52  

241

C h ap te r 8

institutionalized deliberation.”54 But an adequate understanding of political inequality must capture more than the complete absence of influence in the deliberative process. We need a measure of equal opportunity of influence that meaningfully captures less egregious, intermediate cases. Cohen offers a different kind of response to Sen, arguing that because a Rawlsian focus on primary goods makes fewer informational demands, it is more tractable and hence more attractive than a capacities-­based conception of equality.55 We might ask whether this argument might help to resolve the problem of measuring political equality. Addressing the debate in the context of considerations of social justice, Cohen argues that the capabilities approach sets excessive and unachievable information requirements in terms of assessing the nature of an individual’s capabilities set. He suggests that when we take a serious look at the types of information actually available to us in making these determinations, we will find that the best measure we are able to construct in most cases is the more assessable primary-­goods measure: “One way to make the required simplifications would be to specify certain especially severe and informationally transparent cases of limited capabilities, to focus on capability assessments in those cases, and to rely on primary goods for interpersonal comparisons elsewhere.”56 In the context of democratic deliberation, then, this recommendation requires that we first determine a “minimally acceptable threshold of . . . functioning” in the deliberative sphere and then identify some package of politically relevant primary goods. The former determination presumes that we can ascertain with confidence those capacities that are distributed in an “especially severe and informationally transparent” way. This proposal will resolve our measurement problem if the politically relevant asymmetries in capacities that we must capture in order to assess intermediate cases match the capacities that are necessary for a “minimally acceptable threshold of . . . functioning.” Otherwise, we are left with the primary-­goods measure for the equal opportunity of influence. For the reasons that we discussed above, we think that this measure is inadequate to the task of guaranteeing substantive political equality. Unfortunately, the proposed criterion for identifying capacities that are unequal in an “especially severe and informationally transparent” way will be either controversial and difficult to satisfy or inadequate to the task of capturing the capacities for effective participation. Rawls himself acknowledges the difficulty in reaching agreement on such matters: “Whether the constitutional essentials covering the basic freedoms Bohman (1996a, 5). Cohen (1995). 56   Cohen (1995, 285 [emphasis in original]). 54   55  

242

P rag ma tism an d E ffe c tive n e ss

are satisfied is more or less visible on the face of constitutional arrangements and how these can be seen to work in practice. But whether the aims of the principles covering social and economic inequalities are realized is far more difficult to ascertain.”57 In the end, there seems to be no reason to believe that this approach to issues of substantive political equality will be more tractable than the effective-­capacities conception. At this point, although we concede here that there is no easy answer to the measurement question for intermediate cases, we do not see this as an insurmountable obstacle to the goal of incorporating individual-­level capacities into the guaranteed preconditions for effective institutional performance. We agree with Sen that “[f]reedom is an irreducibly plural concept.” He concluded from this that “[w]hile we can attempt to combine the different aspects and sub-­aspects in some integrated formulation, the most important task is to be clear about the different facets of freedom—­how and why they differ, and in what way they have their respective relevance.”58 We find this good advice in thinking about the problem of incorporating individual-­level capacities in the requirements of effective political participation. Ultimately, it is important to acknowledge that effective participation has many dimensions. With regard to the question of individual capacities, rather than relying on the creation of a single index, we must undertake specific analyses of the relative effects of the different politically relevant capacities. In assessing proposals for how to address a particular capacity, we should be guided by a basic question: does this enhance or diminish the opportunities of all individuals to influence collective outcomes?59 While we neutralize the effects of inequalities on the one hand, we must also pursue policies that increase the probability that each individual will have what it takes to effectively engage in the give-­and-­t ake of the deliberative process. Here we tend to agree with those who suggest that “greater means will generally imply a more expansive capability set.”60 But effective political participation requires more than a guarantee of a basic minimum of these resources. Mechanisms for Enhancing Capacities. Clearly, societies must accept trade-­offs if they hope to guarantee the type of political participation that effective democratic institutions require. The desire for political freedom and Rawls (1993, 229). Sen (2003, 585). 59   It is important to recall that for Sen (1999, 174–­81), the weighting and evaluation of capabilities is an essentially political undertaking, best addressed in public argument under democratic conditions. We agree. 60   Cohen (1995). 57   58  

243

C h ap te r 8

equality must be weighed relative to other societal goals. For example, Sen argues that “[a]n attempt to achieve equality of capabilities—­without taking note of aggregative considerations—­can lead to severe curtailment of the capabilities that people can altogether have. . . . [T]he import of the concept of equality cannot even be adequately understood without paying simultaneous attention also to aggregative considerations—­to the ‘efficiency aspect’, broadly speaking.”61 Our defense of effective political participation identifies additional potential trade-­offs in both the private and public spheres. In the private sphere, political equality might require restrictions on the way individuals can use material resources generated by interactions in nonpolitical realms.62 This draws attention to potential conflicts between political equality on the one hand and freedom of economic activity on the other. In the public sphere, political equality, we now argue, might require that the state treat citizens differently. The policy mechanisms required to reduce such inequalities will be most important during the period when citizens who are disadvantaged in ways relevant to political equality are being incorporated into the democratic process. Presumably, as remedial policies and institutional arrangements enhance the influence of such citizens, the democratic outcomes in which they effectively participate will mitigate some of the societal factors that cause their disadvantages in the first place.63 In the interim, however, when citizens are disadvantaged due to past histories of inequality, they are not treated equally when they are treated the same.64 Consider the following proposal from Roemer for a way of distinguishing those asymmetries in politically relevant capacities that deserve to be corrected from those that society should accept and, thus, allow to affect the democratic process. The distinction rests on the notion of personal responsibility: “I say that equality of opportunity has been achieved among a group of people if society indemnifies persons in the group against bad consequences due to circumstances and brute luck, but does not indemnify them against the consequences of their autonomous choices. Thus, an equal-­opportunity policy must equalize outcomes in so far as they are the consequences of causes beyond a person’s control, but allow differential outcomes in so far as they Sen (1992, 7–­8). Walzer (1982). 63   This particular claim touches on a point that Christiano (1996, 257) makes more generally. He notes that the relation between democratic deliberation and political equality is reciprocal. While deliberation presumes equality, it also can promote equality by prompting participants to reflect on their interests and convictions and, thereby, render their participation more effective. 64   Sunstein (1992). 61   62  

244

P rag ma tism an d E ffe c tive n e ss

result from autonomous choice.”65 In order to extend this idea to the question of the equality of politically relevant capacities, we must assess two issues: (1) how do we determine which factors that might threaten politically relevant capacities are beyond the control of individuals, and (2) how do we rectify the inequalities in the context of a democratic process? In answering the first question, we assume that any distinction between autonomous choice and uncontrollable circumstances is political in nature. The determination of which politically relevant capacities are beyond the control of individual citizens will be based on extant conceptions of personal responsibility. These will often be controversial and subject to considerable debate. It is important to note here that as long as the members of a society determine that some asymmetries in politically relevant capacities are due to autonomous choice, and so fall within the control of individual citizens, equality of opportunity of influence will not converge into equality of outcomes (in terms of influence). Unfortunately, this fact does not resolve the measurement problem. It does, however, reinforce the notion that measurement issues will remain, in part, a political question to be addressed by the members of that society. Roemer offers an answer to the second question that reinforces our concerns about the difficulty of adequately guaranteeing equality of opportunity of influence, including attention to relevant political capacities, without requiring equality of outcomes.66 In addressing broader questions of social justice, he focuses on outcomes and recommends the restructuring of outcomes through indemnification as the necessary policy mechanism for achieving equality of opportunity. We propose a different approach and focus on two policy mechanisms that might alter the close relationship between the social distribution of power and resources and the development of politically relevant capacities. The first, and potentially most important, mechanism involves the relationship between material resources and cognitive capacities. The fundamental legitimacy of democratic institutions rests on the effective contributions of individual citizens. Effective participation, in turn, is contingent on the cognitive capacities of these citizens. The standard policy recommendations relevant to the development of cognitive capacities minimally involve government support for education, especially economic guarantees for poor and materially disadvantaged citizens. Yet, while government support of education is essential, it remains an insufficient policy response. This follows from the fact that the effects of unequal income and wealth on asymmetries in the Roemer (1995, 4). Roemer (1998).

65   66  

245

C h ap te r 8

development of cognitive capacities extend well beyond formal educational opportunities. There is substantial scientific evidence demonstrating that intellectual development is significantly affected by childhood poverty and especially by even relatively small levels of malnutrition.67 Recent research shows that the lack of material resources negatively affects diet and environment in such a way that jeopardizes the cognitive development of individuals from impoverished backgrounds. The detrimental effects of poverty and malnutrition on cognitive development first appear in the formative years of childhood, but their effects extend well beyond the early childhood years.68 The main implication of these studies is that the development of cognitive capacities necessary for effective participation in democratic deliberation may require government expenditures to guarantee the social and economic prerequisites of effective participation. These policies imply that government intervention aimed at developing effective participation must disproportionately favor socially disadvantaged groups. It is important to note that we are not proposing the redistribution of income and wealth primarily as a remedy for problems posed by the efforts of advantaged actors to exploit resource asymmetries in order to coerce others.69 Rather, we endorse such redistribution as a remedy for the more fundamental problem that citizens must possess a certain level of income and resources if they are to develop the basic capacities necessary to be effective participants in democratic politics.70 A second proposed policy mechanism aims to avert potential difficulties arising from the relation between the social distribution of power and resources and the development of preferences and of the ability to use cultural resources. According to this proposal, states should create financial incentives to establish secondary associations that foster effective democratic decision making.71 Such secondary associations can encourage the development of various capacities related to deliberative activity and, thereby, can enhance equal opportunity of influence.72 See, e.g., Brown and Pollitt (1996); Mendez and Adair (1999). This suggests both the importance and the difficulty in distinguishing between conditions over which individuals have no control and those for which they can be held responsible. 69   Redistribution might well help mitigate such problems as a by-­product. 70   We are aware, of course, that arbitrary assessments of “competence” have historically been used as exclusionary devices precisely in order to prevent individuals and groups from actively participating in politics. See on this point Smiley (1999). 71   Cohen and Rogers (1995). 72   Christiano discusses several problematic aspects of secondary associations. For example, he points out that claims by such associations to “represent” some broader constituency typically are 67   68  

246

P rag ma tism an d E ffe c tive n e ss

Cohen and Rogers argue that government resources should be distributed in ways that encourage other-­regarding behavior.73 They thus suggest that this policy would favor groups that encompass larger segments of the population over more narrowly defined groups. This would be an appropriate standard of resource allocation in a society in which basic thresholds of effective participation have already been satisfied. However, until such levels have been attained, a more effective policy, designed with an eye toward the goal of establishing substantive political equality, would be to have the state disproportionately favor the groups least likely to have developed the capacities necessary for effective participation.74 The use of these policy mechanisms to enhance substantive political equality entails that governments treat citizens differently in their implementation of economic and social policies. But these policies alone might not correct the adverse effects of distributional asymmetries on equal opportunity of political influence. As part of a transitional policy to incorporate politically disadvantaged groups into democratic processes, governments might also implement procedural reforms that differentially favor these groups. To the extent that democratic deliberation will be incomplete and not result in consensus, voting will remain a necessary component of the democratic process. Can alterations in electoral rules offset some of the problems raised by inequalities in the substantive area? Here we have in mind some of the procedural mechanisms that have been proposed to ensure the adequate proportionality of representation. One proposal would give previously oppressed groups weighted control over the agenda concerning these issues. This proposal is justified on the grounds that mechanisms such as these will enhance the influence of minorities: “As opposed to distributing representation as a proportional resource, a cumulative voting scheme enables a minority to build up coalitions and public associations cutting across the barriers contestable. And he notes that promoters of such associations risk circularity when they both see them as necessary to robust democratic deliberation and hold that deliberation will generate rich associational life (Christiano 1996, 265, 280–­83). 73   Cohen and Rogers (1995). 74   Bohman (1996b, 138) makes a point similar to this regarding the goals of campaign finance reform and public campaign expenditures. Phillips (1995, 180–­82) suggests that there is a tension between her focus on formal representation and Cohen and Rogers’s focus on voluntary associations. We think that arguments such as ours in favor of substantive political equality suggest that this tension is misperceived. On these accounts, the voluntary associations advocated by Cohen and Rogers can be seen as a precondition for equality at the level of formal representation. Fraser (1992) makes a convincing case for the need, within an overarching public sphere, for relatively exclusive arenas within which disadvantaged groups can securely articulate interests and perspectives.

247

C h ap te r 8

that made a particular group’s inequalities persistent. In such schemes, the development of capacities is encouraged by changes in the political economy of how minorities achieve representation: these changes in incentives make it more difficult for the majority to ignore their views and interests.”75 A similar justification is offered for another procedural mechanism that treats citizens in a differential way.76 According to a proposal for “threshold representation,” representatives in a democratic body are chosen according to a criterion that would ensure that all groups have effective representation of their interests. For some groups, this might entail a threshold number of seats that is lower than their actual proportionate numbers in the population. For others, especially disadvantaged groups, this threshold requirement might entail that the number of seats allocated to these groups exceed the number that would be required by a strict proportional scheme. This follows from the facts that (1) if there are only a few members of disadvantaged groups in a deliberative body, they can be easily ignored, and (2) effective influence in a deliberative body requires, in part, the ability to coordinate one’s arguments and activities with other members of the group. It is important to note that these procedural mechanisms place the demands of procedural and substantive political equality in apparent conflict. There are instances in which a mechanism intended to foster substantive political equality may actually generate procedural inequalities within the political realm itself. This, of course, is simply a particular manifestation of the general difficulty posed by the circumstances of politics.

V.  The State: Protections, Interventions, and the Distribution of Power Effective democratic participation is a complex conception, consisting of both procedural and substantive requirements. The degree of this complexity is best seen when we move from vague generalizations about free and equal participants to a more fine-­grained analysis of how we might actually institutionalize freedom and equality in a democratic society. Our analysis identifies a comprehensive framework of guaranteed protections, protections that serve as the necessary preconditions for a normatively legitimate democratic system. We can restate the main conclusions as follows. On our justification of the priority of democracy, the normative requirements of democratic freedom and Bohman (1996b, 134). Kymlicka (1995). Compare some of the institutional devices that Guinier (1995) advocates as well as some of the mechanisms of submajority rule that Vermeule (2007) explores. 75   76  

248

P rag ma tism an d E ffe c tive n e ss

equality are defined by the necessary causal conditions for the effective performance of democratic institutions. Effective performance is contingent on the effectiveness of individual participation. Effective participation requires that all individuals be guaranteed equal opportunity for political influence. In terms of political freedom, this requires that participation be voluntary and uncoerced at all stages of the decision-­making process. Furthermore, it requires that mechanisms be established to protect individuals from those factors in the decision-­making context that may negatively affect opportunities to influence the outcomes of that process. Among these mechanisms, we must include institutions and policies that guarantee that each individual will have sufficient capacity to engage in the process of mutual persuasion that is at the heart of the democratic process. In terms of political equality, this requires that the opportunity for political influence be distributed equally among individual participants. Institutionalizing such equality involves a task of neutralizing (or, in a weaker version, minimizing) the distorting effects of inequalities on democratic freedom and, thus, on democratic participation and decision making. Our account introduces a number of instances of state intervention in the lives of individual participants. Some take the form of state interference in the choices of some political actors in order to establish and maintain horizontal freedom. Others involve issues related to the question of guaranteeing adequate individual capacity. Of special concern to many will be the proposals for treating different people dissimilarly in order to minimize relevant inequalities. These are the types of interventions that cause advocates of standard liberal theory to lose sleep at night. But they are nonetheless dictated by the requirements of effective institutional performance. And here it is important to note that these are not merely the requirements of our pragmatist account of the priority of democracy. We would argue that these approximate the kinds of conditions that are anticipated by the arguments offered by Ober and Rodrik, among others, about the superior features of democratic decision making.77 Arguments such as these are unsustainable without reference to the conditions that are necessary to make the good consequences possible. One cannot make proper sense of their consequentialist claims about democratic institutions without doing so. 77   Both authors indeed gesture in this direction. Rodrik (2007, 113, 231), for instance, repeatedly notes the crucial importance for “deliberative councils” in policy making for developing countries. And he insists on the inclusiveness that such enterprises would require. Similarly, Ober (2008, 24) laments the “limits to Athenian fairness and openness.” As we noted earlier, given their substantive preoccupations, it is unsurprising that they do not dwell on this theme.

249

C h ap te r 8

These forms of state intervention are something that we do not take lightly. Our pragmatist commitments always lead us to prefer a free and open environment in which individuals are able to pursue their interests and engage in experimentation. In such an environment, state intervention in the lives of others is something that should be employed only when necessary. However, when it comes to the case of fostering and maintaining these basic conditions, there is not really an effective institutional alternative. This suggests that the basic guarantees must involve a substantial role for formal laws and governmental enforcement. Given the obvious concerns about the effects of state power, we should be understandably concerned about the legitimacy of this state intervention. On our account, the task of determining the kinds of activities that would constitute violations of the free-­and-­equal-­participation protections rests with the relevant population itself. Such decisions are an important part of the overall responsibility for monitoring and assessing the performance of social institutions. Therefore, it follows that such decisions are best made democratically. This places an important constraint on the actions of those persons responsible for the actual enforcement of these basic protections: to the extent that state interference fails to further the goals established through the democratic process, the interference by state actors does itself constitute a violation of the free-­and-­equal-­participation requirement. And this has the following institutional implication: the task of guaranteeing the legitimacy of state interference is but one aspect of the general task of self-­monitoring attributed to democracy. Both democratic and judicial institutions must be established as mechanisms of redress for individual claims about such illegitimate interference.78 At the same time, institutional mechanisms must be established to monitor the enforcement behavior of state actors. A pragmatist’s approach to this problem would focus on mechanisms for distributing the enforcement authority of government among the full range of societal interests and commitments. Three types of mechanisms come to mind. First, the traditional separation-­of-­power mechanisms would serve to distribute formal governmental authority. These include not only constitutional principles separating the primary functions of government but also institutions of federalism. Second, procedures for selecting government officials (elections for political office, judicial appointments, bureaucratic appointments, etc.) could be drawn in such a way as to facilitate a substantive separation of powers along the lines envisioned by Montesquieu in his discussion of the need to distribute governmental authority among the 78   In the next chapter, we briefly address what our pragmatist account implies in regard to the issue of democratic legitimacy and its relationship to questions of legal and political obligation.

250

P rag ma tism an d E ffe c tive n e ss

array of societal interests.79 The underlying notion is that governmental authority can be best checked if the distribution of government officials reflects the general distribution of power and interests in society. This raises the issue of quotas and institutionalized diversity, but it does so in support of a constraint on governmental authority and not on what are for many the politically contentious grounds of affirmative action in support of minority interests. Third, the development of what Elinor Ostrom has called “polycentric” governance systems could place governmental enforcement procedures within a network of institutional mechanisms, allowing for more effective monitoring of government officials. She describes the general structure of these systems as follows: By polycentric, I mean a system where citizens are able to organize not just one but multiple governing authorities at differing scales. Each unit may exercise considerable independence to make and enforce rules within a circumscribed scope of authority for a specified geographical area. In a polycentric system, some units are general-­purpose governments, whereas others may be highly specialized. Self-­organized resource governance systems, in such a system, may be special districts, private associations, or parts of a local government. These are nested in several levels of general-­purpose governments that also provide civil equity as well as criminal courts.80 In such systems, a combination of centralized and decentralized decision-­ making procedures are employed by a network of state and private actors. One of the overall goals of these polycentric systems is to more effectively distribute the burden of enforcement among formal and informal mechanisms.81 Through mechanisms such as these, the interventions of the state in the choices of political actors can be tempered by attention to the ways in which the power over these interventions is distributed. This proposal is grounded in the notion that the best way of constraining the illegitimate interventions of state actors is to establish monitoring procedures that (1) employ multiple and overlapping monitors and (2) reflect the diversity of interests in society. Here it is important to note that this is yet another way in which freedom and equality are intertwined in a democratic society: equality in the distribution of power over monitoring works to guarantee freedom of political choice.

Montesquieu (1989). Ostrom (1999, 528). 81   Ostrom (1999, 508). 79   80  

251

C h ap te r 8

VI.  The Conditions of Democratic Effectiveness and the Inevitability of Politics Our analysis demonstrates that the task of assessing the existence and extent of democratic freedom and equality is more complex than typically is recognized. Since our justification of democracy presupposes effective participation, such assessment presumes that we can determine with some confidence the effectiveness of individual participants within democratic arrangements. Here effective participation is calibrated in complex ways with the opportunities that people have to influence the outcomes of the democratic process. As we have seen, effective participation has a number of dimensions and, as we move from the more obvious issues to the more indirect and subtle ways in which effective participation can be undermined, the assessment of effectiveness becomes more difficult. No single measure of participation is available to us. The task of guaranteeing the normative preconditions of democratic decision making becomes not a global enterprise but rather an incremental one. And on each dimension and with each assessment, the relevant question is, does the factor under consideration enhance or diminish the opportunities of all individuals to influence collective outcomes? This has the following implications for the robustness of democratic decision making and thus for claims of the superiority of democratic institutions for any of the basic tasks of social coordination. Just as questions of robustness come into play in assessing the appropriateness of decentralized mechanisms like the market, they are equally relevant to the assessment of the appropriateness of democratic institutions. And as in the case of the market, the effectiveness of democratic institutions declines if we fail to sustain the conditions necessary for that effectiveness. But the implications of the robustness of democracy differ in assessments of the various institutional tasks. In the case of first-­order questions, basic questions of which institutional arrangement we judge best for a particular task of social coordination, the robustness of democratic decision making will be a major factor in those decisions. Such questions require a comparative assessment, and the robustness of democratic institutions will affect the relative merits of democracy vis-­à-­vis other institutional forms. To the extent that we are unable to maintain the conditions of effectiveness in a particular situation, democratic decision making will be a less attractive choice. However, in the case of the second-­order question, to which type of institution do we want to entrust the fundamentally important task of coordinating all other decisions about institutional choice, the robustness of democratic institutions has a much different implication. For the second-­order priority

252

P rag ma tism an d E ffe c tive n e ss

of democracy, for which we argue here, is based on the fact that democratic institutions are uniquely qualified, because of their specific combination of institutional features, to assume those tasks. No other available institutional arrangement will perform those tasks effectively even when they are operating optimally. Thus, when we fail to foster and maintain the conditions for democratic effectiveness, there is no readily available institutional alternative to which to turn. In this second-­order case, problems with the robustness of democracy do not merely undermine its relative value but also fundamentally challenge a population’s capacity to make the necessary decisions on basic institutional questions. This fact highlights the importance and thus increases the burden of finding effective mechanisms for monitoring and maintaining conditions of effective political participation. But here it is difficult to resist the conclusion that whatever form the actual mechanism for assessing effective participation takes, the requirements of democratic freedom and equality will themselves remain fundamentally political questions.82 This probably comes as no surprise to anyone familiar with the long-­running debate about the relative merits of markets and democratic institutions. Consider the tension that explicitly characterizes Hayek’s own justification of the market. As we highlighted in chapter 3, Hayek emphasizes that the effectiveness of the competitive process is contingent on the equal participation of all members of the society: “The true basis of this argument is that nobody can know who knows best and that the only way by which we can find out is through a social process in which everybody is allowed to try and see what he can do.”83 And he explicitly argues that the participation must be free and voluntary: “From the awareness of the limitations of individual knowledge and from the fact that no person or small group of persons can know all that is known to somebody, individualism also derives its main practical conclusion: its demand for strict limitation of all coercive or exclusive power.”84 But, as even Hayek acknowledges, the earliest advocates of the market realized that the free and equal participation of all members required formal institutional guarantees: “They were more than merely aware of the conflicts of individual interests and stressed the necessity of ‘well-­constructed’ institutions where the ‘rules and principles of contending interests and compromised advantages’ would reconcile conflicting interests without giving any one group power to make their views and interests always prevail over those In this sense, our arguments here support Christiano (1996, 256) who insists on “the inherent contestability of deliberative equality.” 83   Hayek (1948, 15). 84   Hayek (1948, 16). 82  

253

C h ap te r 8

of all others.”85 He attributes the distortions of market competition to misuses of power, either by the government or by economic actors: “All we need to consider is how difficult it is in a competitive system to discover ways of supplying to consumers better or cheaper goods than they already get. Where such unused opportunities seem to exist we usually find that they remain underdeveloped because their use is either prevented by the power of authority (including the enforcement of patent privileges), or by some private misuse of power which the law ought to prohibit.”86 For Hayek, this places government in the rather paradoxical position of using coercive power as a way of limiting coercive power in the market: “True individualism . . . does not deny the necessity of coercive power but wishes to limit it—­to limit it to those fields where it is indispensable to prevent coercion by others and in order to reduce the total of coercion to a minimum.”87 And this would seem to lend support to our claims about the importance of state enforcement of the conditions necessary for effective institutional performance. Nonetheless, in the case of the market, Hayek resists the obvious implications of his own normative justification. He primarily focuses on the relationship between the individual and the state and emphasizes the importance of governmental noninterference in individual choice. While acknowledging the potential negative effects of inter-­actor power relations, he explicitly rejects the types of interventions that could be necessary to maintain the conditions on which his justifications of decentralized mechanisms depend. For example, he argues that the individualism that he endorses “is profoundly opposed to all prescriptive privilege, to all protection, by law or by force, of any rights not based on rules equally applicable to all persons, it also denies government the right to limit what the able or fortunate may achieve. It is equally opposed to any rigid limitation of the position individuals may achieve, whether this power is used to perpetuate inequality or to create equality.”88 From our perspective, the fact that Hayek accords insufficient attention to the maintenance of the conditions of free and equal participation in inter-­ actor relations undermines his basic justification of decentralized mechanisms. Hayek’s failure to endorse one of the most significant implications of his theory of effective markets is characteristic of many, if not most, proponents of decentralized social coordination. In response we can imagine Hayek Hayek (1948, 13). Hayek (1984, 260). 87   Hayek (1948, 16–­17). 88   Hayek (1948, 30). 85   86  

254

P rag ma tism an d E ffe c tive n e ss

arguing about the impossibility of adequately monitoring and addressing these inter-­actor relations.89 Or he might ultimately admit that he holds the value of individual liberty more dearly than his commitment to effective markets. This would not be uncommon given the fact that we are all confronted at times by trade-­offs among our most deeply held social and political values. But the basic point would remain the same: the process of establishing and maintaining the necessary conditions for effective institutional performance is complex, open to conflict, and inevitably political. There is, in the end, a more general tension here that is subject to constant negotiation and reconsideration: a tension between the inherently political nature of the institutionalization process and the social and institutional preconditions necessary for that very process to be considered normatively legitimate. In the process of these ongoing negotiations, the members of any political society will realize that a commitment to democratic freedom and equality involves potential trade-­offs with other societal goals.90 This follows primarily from a consideration of the institutional prerequisites of the substantive dimension of effective participation. In order to guarantee that each citizen will enjoy equal opportunity of political influence, society must take the steps necessary to guarantee that each citizen has the capacity to effectively participate in the deliberative arena. Under some conditions, this will entail some redistribution of power and relevant material resources as well as an acceptance of inequalities in the treatment of citizens by the state. This will involve hard choices. But the dictates of freedom and equality in a democracy require no less. Hayek (1984, 255). In a discussion of the related problem of creating an index for evaluating economic development policy, Sen acknowledges that such decisions are best handled through democratic decision-­making: “There is an interesting choice here between ‘technocracy’ and ‘democracy’ in a selection of weights, which may be worth discussing a little. A choice procedure that relies on a democratic search for agreement or a consensus can be extremely messy, and many technocrats are sufficiently disgusted by its messiness to pine for some wonderful formula that would simply give us ready-­made weights that are ‘just right.’ However, no such magic formula does, of course, exist, since the issue of weighting is one of valuation and judgment, and not one of some impersonal technology” (Sen 1999, 79). 89   90  

255

Chapter 9 Conclusion

In this final chapter, we do four things. We begin by summarizing our pragmatic justification of democracy. Then we turn to three questions that might be asked about this account. First, to what extent does our argument for the normative priority of democracy provide support for a more general pragmatist theory of legitimate authority and political obligation? Second, to what extent might our argument offer greater appeal to reluctant participants who might otherwise adopt violent or coercive strategies? Finally, to what extent is our conception of pragmatist democracy practical? Conversely, is our argument utopian? In answering these questions, we offer provisional replies aimed more at indicating the contours of our thinking than at providing fully formulated views.

I.  Summary Our argument presumes that we occupy the “circumstances of politics,” a state of affairs characterized by irreducible diversity, inescapable interdependence, and therefore persistent disagreement. Diversity across multiple dimensions is a fact of modern social life; as a result it necessarily informs every aspect of our analysis and justification of democracy. The implications of diversity, in its various forms, are substantial and, given unavoidable interdependencies, pressing. Its effects on our collective endeavors can be both positive and negative. In such circumstances, we face the collective task of maximizing its positive benefits while minimizing its negative costs. But, in accepting this challenge, we must acknowledge that disagreement and conflict are ongoing and inevitable. Because it takes place against this background, politics is, most importantly, about institutional choice. Consensus on substantive policy matters is seldom if ever possible. Nonetheless, we need various types of institutional arrangements that help to coordinate our ongoing interactions generally and facilitate continuing debates over our substantive disagreements in particular. The task of institutional choice is complex. It is not merely a process of selecting a set

C o n clu sion

of rules and practices. It also requires a focus on the conditions under which the various institutional arrangements operate. This is a question of institutional effectiveness. In any particular context, the task of institutional creation involves not only choosing from among several possibilities but also identifying mechanisms needed to monitor whether the conditions necessary for the effective performance of that institutional form are being adequately fostered and sustained. Any modern socially diverse society faces this task. The two of us face it as pragmatists. Our fundamental argument is that pragmatism offers an original and compelling justification of democracy in terms of what we take to be the unique contributions democratic mechanisms can make to processes of institutional choice. We build our justification on a particular conception of pragmatism that emphasizes three basic commitments—­fallibilism, anti-­ skepticism, and consequentialism. On our account, pragmatism endorses a particular approach to political choice in general and institutional choice in particular, what we call tempered consequentialism. This approach emphasizes that in the pursuit of good consequences, we must be equally attentive to the conditions and circumstances under which these good consequences are most likely to be realized. Our conception of pragmatism differs from others within the pragmatist tradition in at least three ways. First, we argue that pragmatism, with its emphasis on consequences and conditions, has important political implications. These implications manifest themselves most significantly in terms of the procedural prerequisites of institutional choice. Second, we reject the view of many pragmatists who endorse a radically democratic perspective and who argue that all aspects of social life are best organized on democratic principles. We argue that no institutional form enjoys this kind of first-­order priority among the various alternatives. Third, we argue that the fundamental role assessments of institutional effectiveness play in pragmatist thinking entails a central place for formal institutions in any pragmatist justification of democracy. In modern, diverse societies, institutional choice is most appropriately seen as an ongoing process of inquiry and experimentation. The task of deciding exactly which institutions work best, where, and to what end is, for a pragmatist, context-­specific and a subject for ongoing experimentation. Because of this, concerns about the effectiveness of the experimental process enter the analysis. Experimentalism will generate useful knowledge only insofar as the conditions are in place to let experimental processes operate in effective ways. Institutional experimentalism requires policies, practices, and, most important, institutions that allow relevant individuals and groups to foster, monitor, assess, and maintain the kinds of social conditions under which

257

C h ap te r 9

experimentalism will thrive in the society at large. This highlights the fact that experimentalism is itself an institutional choice. And it calls our attention to the single most important task for any process of institutional choice: we must identify some institutional arrangement (or set of arrangements) that can serve as a mechanism for (1) coordinating effective institutional experimentation, (2) monitoring and assessing the conditions of effective institutional performance for the range of institutions available in any society, and, most important, (3) monitoring and assessing its own ongoing performance. Which of the available institutions will best perform these tasks? To properly address this question, we have to first consider a number of common claims that in the process of creating or reforming institutions, we should accord priority to one or another form of institution. On these accounts, the priority works to create a default presumption that we should rely on some particular institution to coordinate our social interactions unless it is clearly shown to be inadequate for the substantive task. Many social and political theorists grant such a priority to the market. Our analysis of the effectiveness of markets provides a textbook example for how the process of institutional justification should take place. The selection of any institution raises a burden of justification that involves a process of reconciliation. Advocates of any institutional arrangement like the market advance, at least tacitly, a set of analytical claims about its effective performance. These will include claims about the relationship between the conditions under which the institution operates effectively and the positive consequences that it will produce. In the case of the market, these are claims about the relationship between individual choice under circumstances of decentralized competition and social efficiency. The degree of difficulty involved in satisfying the burden of justification is a function of the magnitude of the discrepancy between these analytical claims and the realities on the ground. That is, our explanatory accounts of the actual conditions under which the institution operates will often diverge, frequently in substantial ways, from the analytical ideal type. When they do, then the normative task focuses on the question of robustness. The more that an institution can continue to perform effectively as the conditions under which it operates diverge from the optimal, the more persuasive will be the normative claim about the value and appropriateness of relying on that institution to coordinate ongoing interactions in a given domain. For the market, the question of robustness hinges on how well it continues to produce socially efficient outcomes as the conditions of decentralized competition wane. In addition to markets, we examined a range of other decentralized institutional mechanisms. And in each instance, robustness emerges as a significant

258

C o n clu sion

concern. The difficulty manifests itself differently in different instances—­ market failure, problems and even outright failures of coordination, or the intrusion of politics in the form of persistent distributional conflict. In the end, each of the decentralized institutional forms we examined encounters challenges to its capacity to operate effectively absent a relatively demanding set of initial conditions. These challenges arise when either the conditions under which it in fact is typically situated or the conditions likely to obtain when it emerges in the first place depart from the conditions that formal analysis reveals as necessary to its effective performance. And there is little if any evidence that the decentralized institutions themselves are effectively self-­regulating, that they have the capacity to monitor and maintain the conditions necessary for their own effectiveness. Thus, in each of these cases, there is pressure toward greater centralization of the tasks of coordination and, therefore, of the structure of decision making. Yet advocates of democratic institutions face a similar burden of justification. They, too, must reconcile the gap between their analytical claims about the benefits of democracy and its realization in everyday life. Standard justifications of democracy tend to rest their case on one of two grounds: (1) they claim that democracy instantiates an important normative value (like equality), or (2) they claim that democracy produces good first-­order effects on collective decision making. Concentrating for a moment on claims of the second sort, we note that these accounts commonly defend democracy primarily as a beneficial means of facilitating collective decision making about substantive policy questions. They emphasize the extent to which democracy can bring individuals together in pursuit of agreement, if not on a common good, then at least on generally shared goals. Two things are striking about justifications of this sort. First, they fail to provide a theoretical framework for assessing the gap between analytical benefits and actual democratic institutions.1 Second, even if a gap did not exist between the performance promised by the analysis and the actual effects, these justifications typically provide no reason to presume that the benefits for democracy would extend across social domains in ways that warrant granting a general first-­order priority for democracy. While these accounts would justify the choice of democratic institutions in a number of domains of interaction, none can support a claim that democracy deserves a first-­order priority among For instance, Shapiro’s (2005) survey of the state of democratic theory reveals that those laboring in the field pay scant attention to institutions. Conversely, in his recent examination of various mechanisms of democratic decision-­making, Vermeule (2007) proceeds by explicitly setting broader questions of theoretical justification to the side. 1  

259

C h ap te r 9

the institutional alternatives across all cases. Put simply, democracy is not the best way for us to coordinate all of our ongoing social, economic, and political interactions. We are often better served by relying on other institutional forms for that purpose. Our pragmatist justification of democracy locates the normatively attractive feature of democracy elsewhere. We defend the priority of democracy by reference to the way in which democratic institutional arrangements facilitate the selection of effective institutional mechanisms to provide ongoing first-­order coordination in disparate domains of interaction. This shift in focus highlights the important role that democracy plays, not in achieving consensus or commonality, but rather in addressing the ongoing conflict that exists in modern society. Indeed, we suggest that democratic politics is perhaps best seen less as a way of reaching consensus or agreement than as an effective way of structuring the terms of such persistent disagreement. The priority we accord to democracy reflects its usefulness in approaching the crucial second-­order tasks involved in the ongoing process of selecting, implementing, and maintaining effective institutional arrangements. Democracy is especially, almost uniquely, conducive to these undertakings. This priority derives from three fundamental features of democracy—­voting, argument, and reflexivity. Each institutional feature contributes to the positive effects of democratic processes on collective decision making. Together they create both the appropriate environment and the proper decision-­making mechanisms to undertake the important second-­order tasks of institutional experimentalism and choice. First, under the appropriate conditions, political argument can serve to undermine the challenges of social choice theory to the legitimacy of democratic outcomes. On our account, political argument operates to mitigate the instability that plagues voting by allowing participants to try to persuade one another, where persuasion contributes to the formation of beliefs and preferences in a quite specific sense. Here persuasion operates less to directly influence the substantive content of individual preferences than to impose conceptual restrictions on the structure of individual preferences. It need not always succeed in so doing, but when successful, this amounts to establishing agreement among the relevant population regarding the dimensions of political conflict. This agreement can then serve to make coherent aggregate decisions about the competing institutional alternatives possible. Second, by emphasizing the importance of political argument, democratic decision-­making structures competition in such a way as to take special advantage of the benefits of diversity. It does so primarily in terms of the particular way in which it employs distributed knowledge in a society. Unlike

260

C o n clu sion

decentralized approaches that would test the value of a particular institutional arrangement through an ongoing set of small-­n interactions, democratic decision making constructs a more comprehensive test for any proposed social institution. The collective nature of the decision-­making process forces any claim about a social institution to face the scrutiny of an expansive array of perspectives while simultaneously giving each perspective the opportunity to challenge any other feasible alternative. In doing so, democratic institutions attempt to guarantee that resulting institutional choices will be the product of a thorough and comprehensive testing process. In this way, they seek to give social diversity the maximal effect on the decision-­making process among the feasible set of institutional alternatives. Third, the reflexivity of democratic arrangements derives from the fact that political argument requires relevant parties to assert, defend, and revise their own views and to entertain, challenge, or accept those of others. It derives, in other words, from ongoing disagreement and conflict. The irreducible diversity of interests, commitments, and attachments that gives rise to politics in the first place means that members of a population have an interest in monitoring the institutions that coordinate their ongoing interactions. Political argument not only allows members of a democratic polity to collectively revisit past substantive decisions, it also allows them to collectively reconsider and revise the terms of their ongoing interactions. Democratic politics, in this way, structures disagreement not just over policy decisions but also over the operation and reform of social and political institutions. Moreover, because democratic arrangements structure political argument so as to allow the emergence of new ideas, perspectives, and interests—­and, thereby, new constituencies and oppositions—­this reflexivity has an endogenous, dynamic dimension. This last feature is especially important for the ongoing task of monitoring and assessing the effectiveness of social institutions. We contend that relative to the other existing institutional alternatives, democracy is better at (1) facilitating experimentalism on institutional choice, (2) monitoring and maintaining institutional effectiveness (particularly in regard to appropriate conditions), and (3) reflexively monitoring its own effectiveness. The capacity of democracy to better satisfy these fundamental requirements of modern, socially diverse societies provides an important reason for endorsing democracy as the best means of collective governance. It grounds our pragmatist case for the priority of democracy. But here it is important to note that as with other institutional arrangements, democratic decision making only performs these tasks effectively under certain specifiable conditions. The priority we accord democracy presumes that it is possible to establish and maintain the conditions necessary for

261

C h ap te r 9

democracy to foster good consequences. This has the following implication for the pragmatist justification of democracy: the conditions of causal efficacy are the same as the conditions of normative legitimacy. This is the lesson of tempered consequentialism. For democratic institutions to perform effectively, a fundamental condition must obtain: the free and equal participation of all relevant individuals. To the extent that this condition is not satisfied, our arguments in support of the relative superiority and, thus, the normative justification of democracy are called into question. The specific criterion that the condition establishes is one of equal opportunity of political influence. This criterion places a substantial burden on any democratic society, as it has significant implications for how that society institutionalizes freedom and equality for its citizens. Concerns with practical questions of how we might institutionalize this criterion require us to make an important distinction between formal rights to participate and effective opportunities to participate. We argue that both formal rights and effective opportunity are necessary to establish a context in which the desired effects of democratic decision making will obtain. That is, the effective performance of democratic institutions is a function of the effectiveness of individual participation. On the view we advance, the conditions of freedom and equality are tightly interrelated. Yet freedom and equality do have distinct implications for the nature of the decision-­making process. Freedom entails something about the nature and content of individual participation in the process; equality entails something about the interrelationship of the individual contributions to that process. At the same time, the freedom and equality conditions reinforce each other in ways that are necessary for their own maintenance and, thus, for the effective performance of democratic institutions. One of the interesting features of our account is that it provides a distinctive justification for freedom and equality in a democracy. As we previously observed, one of the standard approaches to justifying democracy is to link it to an important normative value. For example, some theorists justify democracy as the best means of collective governance because it instantiates equality among participants. In this way, democracy is justified indirectly because we directly value equality. Our pragmatic account offers a different approach. Tempered constitutionalism justifies freedom and equality on the grounds that free and equal participation enhances the effectiveness of the democratic process. Here democracy justifies freedom and equality and not the other way around. The actual process of applying the criteria of free and equal participation to real political institutions will involve difficult and sometimes controversial

262

C o n clu sion

decisions. The institutional requirements do not always lend themselves to straightforward implementation, and they often confront us with trade-­offs with other things that we may value. While we identify significant problems in efforts to quantify the concept of equal opportunity of political influence, the task is akin to a group-­maximization problem: maximize the opportunities for effective participation, given our other values. The emphasis throughout must be on effectiveness, and the focus must be on mitigating the factors that would adversely affect opportunities for influence. Effective freedom and equality raise complex questions for both institutional and policy choice. This complexity highlights the last important feature of our pragmatist justification of democracy—­the social process of experimentation and institutional choice will remain unavoidably political. The political nature of the process is evident at every stage. Disagreement and conflict will be a common feature of the choice of particular social institutions. There is room for disagreement over both institutional outcomes and initial conditions. Social actors cannot reasonably expect a neutral, apolitical metric that might be applied to such decisions. Even under ideal conditions, there is no reason to expect consensus on such matters as how to weight competing normative criteria or how to judge whether empirical conditions sufficiently approach those necessary to ensure normatively attractive outcomes. The fact of social diversity suggests that there will always be space for legitimate disagreement about such matters. And, most importantly, politics extends to the comprehensive task of institutionalizing the experimental process itself. Given both the trade-­offs that must often be addressed and the conflicts generated by the diversity of values in any society, the process of defining the normative boundaries of legitimate democratic decision making will inevitably be a political one. In the end, there is a tension here that is subject to constant negotiation and reconsideration: a tension between the inherently political nature of the institutionalization process and the social and institutional preconditions that must obtain for that very process to be considered normatively legitimate.

II.  Legitimate Authority and Political Obligation In this book, we offer an account of how democracy enjoys an important priority vis-­à-­vis other institutional arrangements and in doing so have offered a normative justification for democracy. But how far does this justification extend? Does it enjoy normative force beyond the relative considerations of governance and other institutional choices? More specifically, what are the

263

C h ap te r 9

implications, if any, for a more general pragmatist theory of legitimate authority and political obligation? Issues of legitimacy and obligation are at the core of perennial debates in political and legal philosophy. Legitimate authority and political obligation are conceptually distinct. Questions of legitimacy focus our attention primarily on the actions of the state, seeking to understand when, and under what conditions, the use of coercive force is normatively justified. Questions of obligation focus our attention primarily on the actions of individual citizens, seeking to understand when, and under what conditions, individuals have a reason for action that requires them to obey the law. But while they are conceptually distinct, questions of legitimacy and obligation are interrelated in a fundamentally important way. In combination, they seek to establish a comprehensive normative justification for the fostering and maintenance of social order writ large. Together they justify an institutional scheme for guaranteeing compliance with law and the dictates of government action. Any such scheme is best considered from the perspective of an individual citizen who faces a decision about whether or not to comply with a law that compels behavior contrary to her self-­interest. On the one hand, if we argue that individuals have an obligation to comply with the law, we are advancing—­tacitly or explicitly—­a reason for action that is independent of her self-­interest. The obligation creates a duty to comply with the law regardless of its effect on her narrow self-­interest. On the other hand, arguments for legitimate authority justify the use of sanctions and other forms of punishment in cases of where individuals do not comply with the law. Such sanctions affect the self-­interested reasons a citizen might consider in assessing whether or not she should comply with the law. When effective, the sanctions will alter the relative value of her choices and make compliance the preferred choice. In tandem, these two kinds of argument seek to create a normative basis for legal compliance and therefore social order, regardless of the type of motivation that governs the compliance decision in a particular case. In arguing that the questions about legitimate authority and political obligation are best understood as two aspects of the same normative project, we want to emphasize three common characteristics of the many justifications that have been offered through the years and that are particularly relevant for our purposes. We already have highlighted the first of these, that theorists aim to provide justifications that extend beyond individual self-­interest and support broader claims about the normative force of coercive authority and political obligation. Second, in their arguments, theorists seek to justify not merely the substance of any particular law or action but the state as a general purveyor of law and policy. Political obligation is conceived of as a problem about the

264

C o n clu sion

general source of law in a society and not about any particular law produced by that source. Third, questions of legitimacy and obligation have generally been conceived of as moral questions. An adequate answer is commonly characterized in moral terms. Justifications of legitimate authority are cast in terms of a moral relationship between the state and its citizens.2 Justifications of political obligation are commonly characterized as moral reasons for obeying the law.3 These common characteristics underlie the minimal criterion for an adequate justification that has emerged in the literature. For our purposes we can state this, in terms of reasons for action, as follows. An adequate justification of legitimate authority must establish a non-­self-­interested, moral reason why individual citizens will accept the general use of coercive force by the state. An adequate justification of political obligation must establish a moral reason why individual citizens must comply with the laws of the state that is independent of concerns about the personal implications of government sanctions. Political theorists have proposed many different types of justifications. We find it most useful to distinguish them in terms of the specific mechanism (or mechanisms) on which the normative force of the argument is grounded. Standard justifications highlight the following range of primary mechanisms: from a natural duty owed to humanity writ large, to an obligation owed to the members of particular groups of which you are a member, to individual consent, to specific processes that instantiate important values (like fairness or equality), and finally to instrumentalist accounts (including, but not limited to, consequentialist accounts of government action). By invoking these various mechanisms, each approach seeks to provide a reason for action that is binding on the individual members of a society. But the task of providing such a justification in a modern world, characterized as it is by diversity across multiple dimensions, has proven quite daunting.4 Given the diversity of moral and ethical commitments that the various members of a society might hold, there is little reason to believe that any single justification will prove compelling and persuasive to them all. As pragmatists, we accept the fact that, as with other issues, such general questions of justification will be controversial and subject to ongoing debate. For this reason alone, we question the emphasis that has been placed on attempting to resolve such issues at some abstract level in terms of universal principles or See, e.g., Christiano (2004). See, e.g., Dagger (2000). 4   The concept of an “overlapping consensus” on a conception of justice that Rawls elaborated is likely the best-­known effort to finesse this problem. 2   3  

265

C h ap te r 9

values. Yet, at the same time, we acknowledge the fundamental importance of social order writ large for our ability to enjoy productive and rewarding social lives. And, thus, we understand the desire to find normative justifications that provide us with compelling reasons for complying with laws and other dictates of the state, most especially in those circumstances in which the consequences of such action run counter to our narrow self-­interest. Through a brief consideration of how a pragmatist might best answer these questions of legitimacy and obligation, we hope to offer insights into both the possibilities and the limitations of such justifications. Various approaches to these issues connect democracy in some way to legitimate authority and obligation. Some propose an indirect connection by way of consent theory. On these accounts, individual citizens consent to the democratic process and thus to the actions and dictates of the democratic state. The nature of that consent can take a number of forms familiar in the consent theory literature. In fact, Putnam attributes such a position to Dewey and “the other classical pragmatists.”5 We find consent justifications unpersuasive for many of the same reasons raised in the critical literature. The most cogent of these criticisms challenges the plausibility of claims that infer tacit consent from the “voluntary” nature of the various types of actual behavior. We think that consent is the wrong approach for pragmatists to take on these matters.6 Others propose connections between democracy and legitimacy that are more explicitly related to the nature of democratic decision making. We have highlighted these in our earlier discussion of democracy. One variant focuses on procedures and justifies democratic outcomes to the extent that the procedures instantiate important moral values, like equality and fairness. A second variant focuses on instrumental effects and justifies democratic outcomes in terms of the good consequences that democracy produces. The epistemological arguments that we discussed in chapter 5 fall into this category. Our proposal for the best pragmatist approach would similarly connect legitimacy directly to democratic decision making. The justification is grounded in a commitment to tempered consequentialism. It combines elements of both procedural and instrumental accounts. To develop this argument, we will focus on the problem of political obligation.

Putnam (2004, 103); Dewey (1927). Having said this, we would add that if a pragmatist is still inclined to pursue a consent-­based justification, she might find support for a kind of continuing consent account in our discussion of the reflexivity of the democratic process. Under the proper conditions, this reflexivity allows for the kind of ongoing reassessment and reaffirmation of democratic decisions that a consent-­based account would have to demonstrate. 5   6  

266

C o n clu sion

The basic structure of a pragmatist justification of political obligation would begin as follows: •  Given the conditions of diversity in modern society, we, individually and collectively, need, at a minimum, a set of social institutions that will coordinate our common social interactions. These institutions are a necessary, but not sufficient, condition for achieving one’s own life plans. Any successful society must find a way of establishing and maintaining that set of institutions. •  Pragmatists are committed to tempered consequentialism for the reasons that we set out in chapter 2. •  Given this commitment, pragmatists are also committed to fostering and maintaining effective social institutions. •  And, given the commitment to effective institutional performance, pragmatists are also committed to the democratic choice of institutions on the ground that it is the best approach to establishing effective institutions. This is our argument for the priority of democracy vis-­à-­vis other social institutions. The next step in the argument for political obligation requires us to make the connection between the commitment to democratic decision making and some form of general political obligation. We believe that tempered consequentialism offers a ground for this connection. The pragmatist commitment to democracy is not merely a commitment to the choice of institutions; it is a commitment to effective choice. And thus there is a commitment to fostering and maintaining the conditions of effectiveness. It is in the commitment to effectiveness that an obligation to obey the laws of the state lies. Among the factors that determine the effectiveness of any social institution is the level of compliance that the set of institutional rules induces among the members of the relevant population. At first blush, this may seem trivially obvious since one of the basic measures of institutional effectiveness would be compliance with the rules by the members of the group. But stated as merely a measure of effectiveness, we miss an important way in which compliance at one point in time affects compliance in the future.7 To see the importance of the ongoing causal effect of compliance, it is best to consider the decision problem from a purely strategic, self-­interested perspective. The strategic decision as to whether or not to comply with an institutional rule is based on a comparison of the expected value of compliance versus that of noncompliance. Now assume that an institutional rule exists 7   For a more general discussion of the determinants of institutional compliance, see Knight (1992, 48–­66).

267

C h ap te r 9

that could, if effective, coordinate our social interactions toward mutually beneficially outcomes. Knowledge of the existence of the rule and the possible positive benefits of coordination will not be enough by itself to guarantee our compliance. In addition, we need to know that others with whom we interact will also follow the rule. That stems from the fact that one of the important determinants of the expected value of compliance with such an institutional rule will be our belief about the probability that others with whom we interact will also follow the rule. If our expectation is that others will be compliant with the rules, then it enhances the expected value of our own compliance: the greater the probability that others will comply, the higher the expected value of compliance and thus the greater the probability that we, too, will comply. This same logic applies to the strategic choice of every member of the group. In this way, the present effectiveness of any social institution is, in important part, a function of stable, reciprocal expectations regarding the positive compliance behavior of social actors. And these expectations are themselves a function of past compliance behavior. When we think about this as a dynamic process, it is easy to see that it is not trivial to say that institutional effectiveness requires compliance to produce ongoing compliance. Without such stable expectations, coordination in terms of the dictates of institutional rules fails. There can be no clearer form of institutional ineffectiveness. The causal importance of stable expectations about compliance informs the pragmatist argument for political obligation. For a pragmatist, a commitment to democracy is a commitment to effective democracy. And the commitment to effective democratic choice of social institutions entails a commitment to fostering and maintaining the social and institutional conditions of that effectiveness. Among the necessary conditions for the effectiveness of democratic decision making is a set of stable expectations that the members of the relevant group will participate in the decision-­making process and accept the outcomes of that process. These stable expectations will be produced through experience, through observations that the other members of one’s group do, in fact, comply with the dictates of the democratic process. Because of this, the effectiveness of the institutions of democracy is contingent, in part, on the level of ongoing participation and compliance in the community.8 8   We recognize that assessing compliance is, in many cases, a complicated task. In part, this is because formal institutions are shot through with informal rules that themselves never actually form a seamless whole. In part, it reflects the inevitably contingent character of social and political interaction and the “unforeseen contingencies” it generates. In part, it simply reflects the vagaries of political judgment on the part of political agents who, often, are passionately or opportunistically engaged in pursuing or protecting relative advantage within an established

268

C o n clu sion

If we commit to effective democratic decision making in the choice of social institutions, then we make a commitment to act in ways that foster the effectiveness of the process. Compliance with the dictates of the process is at the core of this last commitment. Continued institutional effectiveness requires such an ongoing commitment. And it can serve, we think, as the basis of a political obligation to comply with the dictates of democracy. The obligation on a pragmatist account follows ultimately from the commitment to tempered consequentialism.9 We would characterize the extent of the obligation as follows: Tempered consequentialism and its commitment to effective institutional performance sustains an obligation (1) to accept and comply with the procedural dictates that support the effectiveness of the experimental process, (2) to accept and comply with the outcomes of that choice process, and thus (3) to accept and comply with the substantive outcomes of the democratic process in those instances in which it has been selected to perform first-­order tasks of coordination. To better understand the nature of the obligation, it might be helpful to distinguish it from two other arguments about institutional compliance. First, it is not an argument about the prudence of legal compliance. The strategic account discussed above supports an argument for compliance grounded in narrow self-­interest. On this argument, if the institution is working effectively and if everyone else is complying with the rule, then it will be in our self-­interest to comply with the rule because it will produce greater benefits than noncompliance. The reason for action is a purely self-­interested one. The pragmatist argument for obligation offers a different reason for complying with the rule. It contends that a pragmatist has a non-­self-­interested reason for compliance: the commitment to the effectiveness of the democratic process. Second, the pragmatist argument is unrelated to arguments for political obligation grounded in fairness.10 This argument would go as follows. When institutional arrangement. These factors (and others) often combine to make it very difficult, and often quite conflictual, to establish what counts as compliance. It is relatively easy to identify colorful and consequential examples of how these factors interact to generate conflict over the basic institutions of American politics. In the recent past, one might point to the political conflagration that emerged surrounding mid-­decade congressional redistricting in Texas following the 2000 census. For insightful journalistic accounts of this episode, see Toobin (2003; 2006). 9   Again we note that one of the distinctive features of our pragmatist account is that the conditions of causal efficacy establish the conditions of normative legitimacy. 10   See, e.g., Klosko (2004).

269

C h ap te r 9

we commit to democracy as a form of collective governance, we make a commitment to treat our fellow citizens fairly. Part of the commitment to fairness is a commitment to follow the rules of the democratic process. But there will be occasions in any democracy when some of the outcomes of the democratic process will be ones that we do not prefer. The argument from fairness claims that it would be unfair to withdraw the commitment for compliance in those circumstances. On the fairness account, the obligation follows from the commitment to fairness. While pragmatism makes a similar connection from the commitment to democracy to the obligation to comply, fairness plays no role in the pragmatist account. And with that, the sketch of the pragmatist argument for political obligation is complete. We believe that pragmatism offers a similar argument to justify coercive sanctioning on the part of legal authorities as a response to noncompliance with institutional rules. Coercion would be justified by invoking the importance of compliance in establishing or maintaining the stable expectations necessary for effective institutional performance. Again, the justification of the normative legitimacy of state authority would follow from the pragmatist commitment to tempered consequentialism. We return to this matter in the final section of the chapter. Now, having set out what we think would be the best pragmatist justification of legitimacy and obligation, we must turn to the question of how well our view satisfies the criterion set out in the literature. There are two features of our justification that raise questions about its adequacy for satisfying the existing criterion. The first relates to the contingent nature of democratic choice. The institutional choices that are the outcomes of the democratic process are always contingent on the social context and conditions under which the choices take place. They are subject to reconsideration and change whenever context and conditions change. Such flexibility is an essential feature of the process of fostering and maintaining institutional effectiveness. And the priority of democracy is, in an important way, a function of the fact that democracy’s reflexivity best facilitates a process of ongoing reconsideration of the effectiveness of social institutions. Yet this very contingency challenges any attempt to characterize the proposed pragmatist obligation in terms of universal moral principles. The second, and more important, question relates to the inexorable political dimension of democratic choice. As we have argued throughout this book, the combination of social diversity and unavoidable interdependence makes some degree of ongoing political conflict inevitable. Just as this applies to substantive policy differences, so it applies to the choice of our basic social institutions. The persistence of ongoing conflict suggests that, at the very least,

270

C o n clu sion

some of the institutional choices will be the product of political compromise, producing institutional rules that instantiate the values of some members of the group and not those of others. In such circumstances, the possibility of future political conflict exists. Efforts to derive an obligation to obey the law from such a political compromise of values, especially with its remainder of ongoing disagreement, will strike many as incompatible with a desire for strict moral justifications. For these reasons, if the criterion for an adequate justification of political obligation entails a moral reason for compliance, our pragmatist proposal will fail because of the inherently political, contingent nature of democratic institutional choice. But before rejecting the proposal out of hand, a pragmatist might reasonably insist that we ask, why must political obligation be grounded in morality? Legal and political obligations have traditionally been treated as a particular case of the more general category of obligations and duties that are understood to be moral in nature.11 The underlying logic of such treatment is that if obligations in general require a moral basis, then so do legal and political obligations. Considered from the perspective of someone who seeks to establish the normative basis of social order, we understand the desirability of having a reason for action that is grounded in morality. Such reasons are deemed more compelling, seen as deriving from a stronger, if not absolute, principled commitment. The binding force of such a moral obligation is presumed to be greater than other kinds of reasons for action. If successfully established, a properly moral obligation most effectively trumps the claims of narrow self-­interest. Nonetheless, however much we may hope to back political obligation with the presumptive force of moral reasoning, this desire runs up against the fact of social diversity. The efforts to ground normative justifications of social order in universal moral principles or values seem to deny the reality of such diversity. To us, this seems the wrong approach to the problem, a fruitless enterprise. The same inevitable conflict of values and commitments that raises questions about the adequacy of the pragmatist proposal confronts nonpragmatic efforts to normatively justify obligation and legitimate authority. There is no way to avoid the overarching significance of political conflict and compromise when it comes to issues involving collective governance and the state. In this respect, our pragmatist account hardly is distinctive; in a socially diverse world, the inevitability of political conflict represents a significant challenge for any account of legitimacy and obligation. Because of this, we would argue that questions of the legitimacy of state authority as well as those of the See, e.g., Hart (1958); Raz (1999).

11  

271

C h ap te r 9

nature of our obligations to that state are not appropriately conceived of as moral questions. They are, in fact, necessarily political. And they require answers that explicitly address the implications of politics for such justifications. Those theorists who persist in demanding justifications in terms of moral reasons must accept the burden of demonstrating how their preferred justification can be sustained without reference to the arbitrary effects of politics.12 But having acknowledged the inevitability of politics, we would insist that this pragmatist proposal might still support a normative justification of legitimate authority and obligation. The obligation, derived from the commitment to tempered consequentialism, satisfies the other two aspects of the criterion of adequacy developed in the literature. It justifies a general system of law and government action. It establishes a non-­self-­interested, binding reason for complying with the dictates of democratic governance. It is similar to the type of obligation that Gilbert defends in her critique of the morality requirement for legal and political obligation.13 On Gilbert’s account such obligations are conceptually distinct from both moral obligations and narrow self-­interest. They are entailed by our commitments to other values, principles, and collective goals. They create a binding commitment to obey the law that trumps our self-­interest but that may not trump our moral commitments. While such an obligation lacks the force of a moral claim, it is genuine in the sense that it provides a binding reason for compliance in those situations in which noncompliance would produce the greater short-­term benefits. In our pragmatist proposal, the commitment to tempered consequentialism entails a binding commitment to comply with the dictates of the democratic process, when the latter is operating effectively. This commitment holds even when it would be in our narrow self-­interest to disobey the rules. For many this will seem an insufficient basis on which to ground the normative justification of social order, but for a pragmatist this is the best that one can offer for questions about legitimate authority and political obligation. Pragmatists accept the inevitability of politics and ongoing disagreement over basic values and attachments, but they deny the inevitability of normative relativism. As in other aspects of our argument, the key to establishing political obligation is the effective operation of democratic institutions. The politics of institutional choice, when it takes place under the right conditions, We use the term “arbitrary” here advisedly. It retains the sort of purchase many normative theorists hope to derive from it only to the extent that some moral principle or commitment can survive as nonarbitrary in the face of diversity and interdependence. 13   Gilbert (2006). 12  

272

C o n clu sion

establishes a nonmoral but non-­self-­interested obligation. Pragmatists insist that when politics take place within the appropriate institutional framework (a set of democratic institutions) under the appropriate set of conditions of effectiveness (free and equal participation), the democratic process will produce good institutional choices, choices that will create their own political obligation. This justification highlights the basic point that the normative bases of social order, whether they are legitimate authority or political obligation, are neither an inherent aspect of any particular system of governance nor an essential feature of any relationship between the state and its citizens. Legitimacy and obligation are created through action, in the same way that institutions are made more effective. They must be fostered and maintained as part of an ongoing creative process. This pragmatist approach directly addresses the ways in which ongoing political conflict challenges claims for legitimate authority and political obligation. It offers a framework for assessing the divergence of the real from the ideal as well as a diagnostic for dealing with factors that undermine pragmatist justifications.14 Under the right conditions, democratic institutions can minimize the opportunities for these arbitrary factors to influence institutional decision making. This is one of the important lessons of our analysis of the effects of political argument on the phenomena of cycling and manipulation that underlie the social choice critique of democratic legitimacy. And, of equal importance, effective democratic institutions allow the people themselves to monitor and facilitate the ongoing debates. For pragmatists, this serves as the institutional framework for fostering and maintaining the normative bases of social order. If the institutional framework operates effectively, pragmatists will recognize their political obligation to obey the law and accept the legitimate authority of the state. We believe that many nonpragmatists might find this justification of democracy compelling. Consider the two stages of the argument. The first stage is the priority of democracy for the important second-­order task of choosing effective social institutions. Its appeal rests on the fundamental role of institutions in social life. A characteristic feature of any diverse society is that its members will be committed to an array of criteria, grounded in substantive principles or procedures, that would serve for them as the basis of political legitimacy. One striking feature of any of these criteria is that they will unavoidably incorporate some conception of effective institutional performance. 14   Here it is important to note that these are the same factors that other theories deem morally arbitrary.

273

C h ap te r 9

Regardless of whether claims to legitimacy are grounded in collective benefit (however measured) or are grounded in some substantive good that is considered the cornerstone of individual life plans, effective institutional arrangements will be essential to achieving that benefit or producing that good. This follows from the fact that the success of any individual’s life plan depends, in part, on her capacity to interact effectively with other members of her community. The coordination facilitated by social institutions is necessary for such interactions. Effective social institutions are a necessary, but not sufficient, condition for the realization of individual goals and commitments. Because of this connection, nonpragmatists might come to see, through the explication of our argument, that their own approaches presuppose in crucial ways the argument about institutions that we offer here. Thus, they might accept the priority of democracy on the grounds of the necessary importance of effective social institutions. The second stage is the pragmatist justification of legitimacy and obligation, grounded in the commitment to the conditions of effective institutional performance. The appeal of this second stage rests on the potential benefits that effectiveness might have for individuals’ own life plans.15 Given that democracy represents the best means of fostering and maintaining effective institutional performance, the need for effective social institutions may offer a reason for nonpragmatists to accept the argument about the important role of compliance for stable expectations and, thus, effective democracy. On this account, the more effective the social institutions, the more complete will be individuals’ enjoyment of their own goals. If one is willing to acknowledge the inevitable disagreement that exists in diverse modern societies over the reasons why democratic decision making is normatively legitimate, then our pragmatic argument might offer a good and persuasive “second-­best” reason for embracing democracy and accepting the dictates of its operation. Nonetheless, we have some doubts as to whether this reason would attain the status of a genuine obligation for nonpragmatists. For pragmatists, the obligation is entailed by the commitment to tempered consequentialism. While nonpragmatists might accept the necessity of effective social institutions, the fact that they lack the underlying commitment to tempered consequentialism may cause them to treat the reason for compliance in a different way. Rather than accepting the reason as a normative commitment, they may treat it in This aspect of the argument is similar to the argument for obligation offered by Hume (1752/1953). Hume’s argument applied to any system of laws that allowed people to pursue their interests in a cooperative way. Our argument applies only to systems of laws created democratically. 15  

274

C o n clu sion

more practical terms. That is, nonpragmatists might treat it prudentially, as a rule of thumb for strategic choice. For us, this remains an open question.

III.  Acceptance of Democracy as the Terms of Politics In our discussion of preliminaries, we highlighted a standard feature of debates in normative political theory. Participants in these normative debates commonly assume that their relevant audience has already committed to the nonviolent resolution of their disagreements and differences. In so doing, they ignore the challenge posed by those who would repudiate politics. We suggested that this raises an important question that should be explicitly addressed: “What might persuade people to eschew direct recourse to violence and coercion and to accept institutionalized ways of resolving conflict and achieving cooperation?” The lack of attention to this kind of question that characterizes the normative literature is one of its most significant weaknesses. In this section and part of the next, we address this matter directly. In constructing their normative justifications, political theorists should assess the extent to which their arguments are grounded in social and institutional mechanisms that will actually foster the behavior they idealize in their theories. On the one hand, this is an important feature of the burden of justification for any normative account. If we cannot plausibly establish how our normative vision might actually be achievable on the ground, so to speak, then there is little reason to find them compelling answers to these most important social and political questions.16 On the other hand, such an assessment will help us to clarify our basic arguments and highlight which features of the argument are really doing the normative work in any particular case. Practical assessments such as this serve as a necessary reality check for normative claims by political theorists. Just as we earlier accused many positive theorists of smuggling unstated normative values into their explanatory accounts, we might similarly question the tendency of normative theorists to incorporate unstated (and often unsubstantiated) positive assumptions into their normative justifications. At the very least, political theorists should accept an obligation to clearly identify the connection between the basic tenets of our normative arguments and the plausibility of the social mechanisms on which we rest our cases. In that spirit, we want to offer a brief assessment of the ways in which we think that our pragmatist account might answer the practical question about Here we refer readers back to the discussion in chapter 1 of “ideal” and “non-­ideal” theory.

16  

275

C h ap te r 9

the acceptance of democratic politics. Here we will be brief because we have already presented the substance of our answer, such as it is. It only remains for us to highlight the relevant points for the specific issue at hand. Unlike most normative arguments, our pragmatist account directly incorporates the inevitability of disagreement and conflict. In doing so, it provides an argument for the central role of ongoing political debate in establishing and maintaining the bases of legitimacy and obligation. And, through an analysis of the effects of democratic decision making on the collective outcomes that it produces, we make a case for the superiority of a democratic institutional framework as the forum for undertaking such debates. We argue here that when there is persistent conflict and disagreement, making collective decisions democratically is the best means of creating an institutional environmental in which both individual and collective life plans can be effectively pursued. Pragmatism holds that knowledge comes best from experience. A pragmatic answer to this question of democratic acceptance would rely, we think, on the effects of experience on the assessment of the relevant benefits of different political strategies. There are at least three types of experience that might persuade someone engaged in persistent disagreements and political conflict to accept democracy as the forum for such conflict. Two of these experiences highlight the potential effects of democratic procedures and go to the question of what effect, if any, participation might have, under the right conditions, on one’s acceptance of democratic politics. We have already discussed the first in our examination of the effects of formal institutions on the emergence of social norms in chapter 6. There we identified a potential effect of democratic participation on attitudes and beliefs about the members of other social groups. On our account, a potential by-­product of participation, under the types of conditions of institutional effectiveness that we have identified here, is the development of more positive beliefs about those with whom we are in persistent disagreement. A substantial body of social scientific literature on social contact among members of conflicting groups demonstrates that positive attitude change can occur when the right kinds of conditions hold.17 If such positive change were to occur, we believe that the improved perceptions of those with whom we are in conflict might be one factor that positively affects the assessment of the relative benefits of democratic politics. The second potential effect of democratic participation might be an improvement in attitudes and perceptions about state actors and government Forbes (1997).

17  

276

C o n clu sion

officials. What we have in mind is a process analogous to processes identified by Tyler and his collaborators in their research on social justice and legal order.18 Their research has shown that one of the most important factors in explaining why people commonly obey the law is their attitude about procedural fairness. If people believe that they have been treated fairly in their experiences with judicial, legal, and other government officials, they are more likely to obey the law even when it is not in their narrow self-­interest to do so. On their account, the experience of fair treatment through participation in the judicial and legal process enhances attitudes about law and facilitates greater compliance with the law. We suspect that participation in effective processes of democratic decision making might generate similar effects.19 The conditions of free and equal participation envisioned by our pragmatist justification, if actually realized, could positively affect the attitudes and beliefs that participants hold about democratic officials. If the experience of participants is such that they believe that they have had a real opportunity to affect the outcomes of the process, it may enhance their attitudes about that process and thus positively affect their assessment of the relative benefits of democratic politics.20 The third reason that our account provides for accepting democratic politics is substantive in focus and refers directly to the experience of the consequences of the democratic process. It is grounded in the necessary role that social institutions play in how we achieve individual and collective goals. It is, restated, the argument that we offered at the end of the last section for why we think nonpragmatists might find the pragmatist argument for political obligation appealing. Given the importance of effective social institutions for the realization of a successful and rewarding life, we think the actual experience of an effective democratic decision-­making process with its resulting institutional framework might provide political actors with a more compelling demonstration of the actual value of democracy. On this account, the actual See, e.g., Tyler (1990; 1997). For empirical support for this view, see Levi (1997). 20   We note that these potential participatory effects on attitudes and beliefs are different from the kinds of benefits touted by participatory democrats like Pateman (1970). Participatory democrats justified democratic politics, in important part, on the salutary effects of participation on an individual’s personal development. We agree with Mansbridge (1999b) that such effects are exceedingly difficult to discern. And we agree with Elster (1983) that even if we could with some confidence discern participatory effects, they are a dubious basis on which to hang a justification of democracy. That said, the effects we propose are limited to the beliefs that participants hold about others with whom they interact and thus to their beliefs about the relative benefits of democratic participation as a political strategy. 18   19  

277

C h ap te r 9

experience of good consequences in terms of institutional choice will plausibly enhance their relative assessment of the value of democratic politics as an effective political strategy. But in the end, we make this argument cautiously. We are uncertain as to how confident we should be about these proposed effects. Yet we believe that normative theorists should take this problem more seriously than they have in the past. And we present this argument as a way of demonstrating how theorists might go about offering such an answer. One of the strengths of a pragmatist approach to normative justification is its capacity to directly and explicitly address practical questions such as this.

IV.  Final Remarks: Is Pragmatist Democracy Utopian? Despite our preoccupation with political-­economic institutions, the realist account we offer of how they emerge, and the burdens that our account presents for those who aspire to justify any given institutional arrangement, we anticipate that our argument may be characterized as utopian. This charge might take one of at least two forms. First, it might concentrate on the broad conception of agency that we have built into our theory of democracy. It would challenge the combination of capacities and conditions that underwrites that conception of agency. Second, it might hold that, starting as we do from what we call the circumstances of politics, we sidestep from the outset the true problem of politics. That problem, of course, is how to maintain the circumstances rather than resort directly to force, compulsion, or violence. In this final section, we will anticipate objections of this sort. We do not claim to offer a definitive response. We simply sketch the contours of a reply that indicates why we consider charges of utopianism misguided and what sorts of arguments we would marshal against those tempted to level such charges. We will take each variant of the charge that our argument is utopian in turn. In the following remark, Thomas Nagel captures the position in which we find ourselves. Political theory typically has both an ideal and a persuasive function. It presents an ideal of collective life, and it tries to show people one by one that they should want to live under it. These ambitions may be universal, or they may be more local, but in either case there is a serious question of how they can be realized jointly, and whether they necessarily interfere with one another. An ideal, however attractive it may be to contemplate, is Utopian if real individuals cannot be motivated to live by it. But a

278

C o n clu sion

political system that is completely tied down to individual motives may fail to embody any ideal at all. One might try to subordinate the persuasive to the ideal function by saying that a political theory should concern itself exclusively with what is right, for if it can be shown that a certain form of social organization is the right one, that should be all the reason anyone needs to want it to be realized. But this seems excessively high-­ minded, besides ignoring the relevance of what is motivationally reasonable to what is right. If real people find it psychologically very difficult or even impossible to live as the theory requires, or to adopt the relevant institutions, that should carry some weight against the ideal. On the other hand, one has to be careful not to turn this into an excuse for giving up too easily. There is a danger that one will get into the habit of thinking that any radical departure from accustomed patterns is psychologically unrealistic.21 Some, but by no means all, of those who have read this far might consider our argument radical. We take no position on that, even though we acknowledge that the claims we advance are immodest precisely insofar as we advance a distinctive rationale for according priority to democracy in any complex political-­ economic arrangement.22 That said, it surely is germane to ask whether the position we stake out is, in Nagel’s terms “psychologically unrealistic.” This is a complaint that worried Dewey when he reflected on his own “faith” in democracy. At the time, he lamented that he had “been accused more than once and from opposed quarters of an undue, a utopian, faith” in what he referred to as “the capacity of human beings for intelligent judgment if the proper conditions are furnished.”23 In some sense, we are less concerned than Dewey on this score. We recognize, as he does, that in politics agents proceed from a complex mix of self-­regarding and (a variety of) other regarding motivations. And we place very little emphasis on the notion that those individuals who occupy places in democratic institutions will be somehow transformed, much less dramatically or necessarily for the better, by the experience. That said, our argument for the priority of democracy places considerable weight on the reflexive character of democratic practices and institutions, which in turn surely presupposes that individuals who take part in it are not Nagel (1989, 903–­4). Knight and Johnson (2007, 60). 23   Dewey (1939/1993, 242). Note that Dewey immediately goes on to insist that these conditions require institutional embodiment—­they must, he notes “be secured by effective guarantees of free inquiry, free association and free communication.” 21   22  

279

C h ap te r 9

only capable of assessing and reassessing the performance of institutions but will be motivated to do so. And it is on just that point that skeptics might find our argument utopian insofar as the individuals we have in mind must not only be intelligent in Dewey’s sense but be willing to embrace experimentalism with respect to political-­economic institutions. Differences about this very point have characterized the history of pragmatism. We already have noted that the dyspeptic Peirce held that although science afforded a more reliable method of settling belief, the “method of authority will always govern the mass of mankind.”24 One might see Richard Posner in some moods as a contemporary pragmatist who embraces something like Peirce’s position. Dewey, too, conceded persistent and widespread obeisance to tradition and authority, especially in the domain of politics. Men have got used to an experimental method in physical and technical matters. They are still afraid of it in human concerns. The fear is the more efficacious because like all deep-­lying fears it is covered up and disguised by all kinds of rationalizations. One of its commonest forms is a truly religious idealization of and reverence for established institutions: for example in our own politics, the Constitution, the Supreme Court, private property, free contract and so on. The words “sacred” and “sanctity” come readily to our lips when such things come under discussion.25 But Dewey saw this less as an inevitability to be lamented than as a problem—­ problem, to be sure, exacerbated by the daunting size and complexity of interdependencies in the modern world—­to be overcome. Contemporary pragmatists often follow him in that regard.26 We identify with this latter line of thought. Crucially, however, we insist on two caveats. First, we concede that persuading relevant constituencies to embrace pragmatist experimentalism is a daunting, ongoing problem. Where they succeed, it constitutes an achievement, but a precarious one. Arguably, the classical pragmatists underestimated how persistent the demands for certainty and foundational principles can be in practice.27 On our view of politics, it is unsurprising that there will be political actors who enthusiastically Peirce (1958, 110). Dewey (1927, 169–­70). Compare Dewey (1935/1991, 51–­52). 26   For instance, Rorty (1999, 168) holds out the hope of “getting our fellow citizens to rely less on tradition, and to be more willing to experiment with new customs and institutions.” Unger (2009, xvi) adopts a similar position, albeit in a less sanguine way. He recommends sets of political reforms—­institutional innovations—­that “would force Americans to lift the exemption from experimentalism that they have usually granted to their institutions.” 27   Bernstein (2005, 51–­52). 24   25  

280

C o n clu sion

press such demands or even resurrect them in instances where they have been overcome. Second, we do not pin our expectations regarding institutional experimentation primarily on the public-­regarding character of political agents. Although we concede that in some instances, institutional reforms might well be animated to some extent by agents pursuing some vision of what they take to be the public or common interest, we see political agents as more commonly driven by their own interests and by the difficulties they encounter in successfully advancing or protecting those interests as others seek to do the same in given circumstances. In pragmatist parlance, we see reflexivity as animated by the problems that individuals and groups encounter as they pursue their aims and projects. Because, in the circumstances of politics, conflict and disagreement abound, those problems often beset particular groups—­the excluded, the defeated, the relatively disadvantaged—­in especially pressing ways. In that sense, it is precisely “losers” who provoke the sort of democratic reflexivity we identify.28 We do not present this as an especially romantic vision. For just as the “losers” press to alter institutional arrangements that work less to their disadvantage, those already relatively advantaged will pursue strategies designed to fortify extant arrangements against such pressures. That said, there is an inherent learning process that takes place in such interactions.29 And in a democracy, ongoing political conflicts between winners and losers will, if they are to be justifiable, transpire under conditions that approximate to a greater or lesser degree the sort of freedom and equality we have discussed.30 Skeptics commonly and confidently claim that normal individual human beings simply are incapable of playing the sort of role into which democratic

“Defeat is the mother of invention. In this sense it is losers who provide a political dynamic in public life—­innovating and strategizing to become winners on the one hand, and energizing the incumbent winners to anticipate and try to deflect the losers’ maneuvers on the other” (Shepsle 2003, 310). This observation simply places a sharp rhetorical edge on the idea that those persistently disadvantaged by the rules of the game often will engage in “strategic moves” aimed at reconfiguring those rules (Schelling 1960). 29   Whether relevant actors draw salutary or even accurate lessons is an empirical question. There is no guarantee that they do so. The emergence of “antidemocratic” forces, to which we return below, is one possible outcome of losers who become disgruntled and conclude that their grievances cannot be satisfied via democratic politics. 30   It is crucial to recall that these conditions are no more idealized or more stringent than those needed to sustain the effectiveness of markets or common law or other familiar institutional forms. On the view we have presented, then, the conditions necessary to sustain the priority of democracy are not in any obvious way more susceptible to charges of utopianism than are those that any other institutional form demands. 28  

281

C h ap te r 9

theorists cast them, tacitly or explicitly.31 From our pragmatist perspective, such claims are not nearly as troubling as the skeptics presume them to be. There is no need to deny that individual members of any public are ill informed or mal-­adept at navigating various cognitive tasks. After all, Dewey acknowledged much of what skeptics of his day took to be the facts of the matter.32 But, in the first place, beginning with Peirce, pragmatists have seen knowledge as produced and held by communities of inquiry. That means we should expect no single individual to either produce or possess all the knowledge or information required to address a given social, economic or political problem. It means instead that we ought to conceive of epistemology in not just social but institutional terms.33 Like Dewey, then, contemporary pragmatists remain critical of arrangements that accord too much of a role to experts or technocrats even as they simultaneously resist naively placing their faith in the capacities of the common man except, of course, insofar as those capacities are deployed under “proper conditions.” From our pragmatist perspective, then, there is little reason to characterize our argument as “psychologically unrealistic.” And there surely are grounds for responding to such complaints. We turn now to the second mode of “utopianism” we might be accused of advancing. Here we are circling back to matters we raised in earlier sections of the chapter. For we are raising the question of how to respond to those who refuse to recognize the legitimacy of democratic arrangements or who refuse to acknowledge their obligation in the face of democratically established laws and policies. From the outset, we identify the circumstances of politics as those in which a population seeks to coordinate its ongoing interactions under conditions of diversity and interdependence and the sorts of conflict to which that endeavor will unavoidably give rise. The alternatives to politics are exit (whether individual or collective) and violence. The two often appear in combination, as when ethnic cleansing gives rise to displaced populations or when attempts to secede elicit forceful resistance. In any case, our argument takes for granted that we occupy the circumstances of politics. Critical readers might contend that by setting aside the ongoing temptation of individuals and groups to resort Claims of this sort run, minimally, from Lippmann (1927/1993) to Hardin (2009, 60–­83). That said, given that much of the evidence for widespread individual-­level ignorance regarding politics derives from studies based on standard survey techniques that themselves are designed to isolate individuals, one might raise doubts about how empirically well founded the claims of skeptics are in actuality. For a view of survey techniques that sustains such doubts, see Sanders (1999). 33   Hence Dewey (1927) both rejects technocratic remedies for the problems of the public and sketches a vision of systematic inquiry that might come to inform public debates concerning political-­economic matters. For recent attempts to elaborate this vision, see Knight and Johnson (1999), Anderson (2006), and Kitcher (2006). 31   32  

282

C o n clu sion

to exit or force, and the need a pragmatist democracy will face in confronting those who succumb to those temptations, we are being utopian. Pragmatists have confronted something like this criticism before. For instance, having portrayed the emergence of pragmatism as a reaction against fanaticism and the violence to which it gives rise, even so sympathetic a commentator as Louis Menand ends up complaining: “Pragmatism explains everything about ideas except why a person would be willing to die for one.”34 At what juncture, in other words, could political agents who truly embrace pragmatism and its constitutive commitments of fallibilism, anti-­skepticism, and consequentialism themselves resort to force or compulsion, and in the process risk their own lives, in response to opponents who threaten democratic political arrangements? Our reply to complaints of this sort complements what we already have said. In the first place, while we cannot rehearse the historical debates regarding the resources available to pragmatists in the face of hard choices, or what lately has emerged as the topic of “evil,” it is by no means obvious that those whose views Menand summarizes have properly grasped them.35 Second, as we noted early on, pragmatists are not unique in confronting the problem of how and where to draw a distinct line in defense of democratic practices and institutions. Let us elaborate on these points in a way that brings us back to the importance of the institutional scaffolding essential to a community of inquiry. Dewey, as is well known, insisted that democratic ends could only be attained and sustained by democratic means. This would seem to capture precisely the vacillation and indecisiveness for which pragmatism regularly is criticized. But here, writing in 1935, is Dewey on the topic of how, in the face of resistance, democrats might implement far-­reaching political-­economic transformation: The one exception . . . to dependence on organized intelligence as the method for directing social change is found when society through an authorized majority has entered upon the path of social experimentation leading to great social change, and a minority refuses by force to permit the method of intelligent action to go into effect. The force may be intelligently employed to subdue and disarm the recalcitrant minority.36 It is clear that Dewey is willing to draw the line precisely at that point where opposition becomes antidemocratic, where it threatens the institutions and Menand (2001, 375). Bernstein (2005) offers a spirited, historically informed, and theoretically sophisticated defense of pragmatists against the putatively “realist” critics. 36   Dewey (1935/1991, 61). 34   35  

283

C h ap te r 9

practices of systematic, experimental inquiry. And it also is clear that he insists that any resort to force and violence must be undertaken “intelligently.” This may appear simply to be a mincing of words. But it is clear that any democratic polity must address the problem raised by “antidemocratic” individuals and groups, by which we mean those committed not just to defeating this or that policy but to subverting democratic practices and institutions altogether. Once we set aside as impracticable the liberal view that the state can in some sense remain neutral between contending parties, there are three broad strategies it might adopt in the face of antidemocratic forces. These are toleration, repression, and incorporation. The first entails “respect and openness” for alternative views and practices; the second consists in “a policy aimed at weakening or dispersing a political force;” and the third involves “an active effort to include a substantial part of a political force in normal democratic political contestation.”37 From our pragmatist perspective, mere toleration is an implausible response to antidemocratic challengers. We have argued that democracy is the institutional embodiment of the basic commitments—­fallibilism, anti-­skepticism, and consequentialism—­that we identified at the outset. How then could we consistently be open and respectful of those who would subvert democratic practices and institutions? Tolerating antidemocratic activities would mean acquiescing to the subversion of the commitments we take as basic. Such antidemocratic activities demand more active, critical engagement.38 What Dewey refers to as an “intelligent” response to individuals and groups who would threaten democratic arrangements would begin from efforts to incorporate the “recalcitrant minority” into democratic politics. Democratic institutions, as we depict them, aim to facilitate such incorporation precisely to the extent that they confront the task of monitoring the conditions of their own effective operation. As we specify them, those conditions would require that winners engage the grievances of losers and that the latter have the opportunity to press their grievances, with some prospect of success, through legitimate channels. The hard constraint on such grievances, of course, would be that satisfying them could not entail the subversion of democratic institutions and practices. And that, of course, is a matter of political judgment and disputation. Efforts to incorporate antidemocratic individuals and groups will not always succeed or do so completely.39 In such cases, democratic governments can Plotke (2006, 20, 22, 27). Bernstein (2005) rightly argues that the pluralism pragmatists endorse does not reduce to a facile relativism. See also Johnson (2000). 39   It may be possible to incorporate some portions of an antidemocratic opposition but not others. 37   38  

284

C o n clu sion

legitimately resort to repression.40 They may do so, from a pragmatist vantage point, if they proceed “intelligently.” First the antidemocratic individual or group must, demonstrably, be transgressing legally protected avenues of dissent and opposition. Second, any acts taken by the state to suppress those individuals and groups must themselves proceed according to established law and be subject to legal review. Two presumptions are at work here—­the official reaction to antidemocratic forces must be proportionate and must not be based on emergency powers. These presumptions constrain the inevitable political judgments that must be made in assessing threats to democratic politics in any given instance.41 And these judgments, in turn, of necessity will be contextual and focused on the consequences of adopting one strategy or another in response to antidemocratic forces. The sketch we offer in the preceding few paragraphs admittedly is incomplete. It nevertheless gives some indication of how we would respond to the sort of skeptical view Menand articulates. It is unclear whether Menand himself embraces that view. But those, at least, for whom he purports to speak seem to presume that an individual would risk her life—­or the lives of her compatriots—­only in defense of ideas or principles or commitments that have external backing of some sort. But pragmatists from Dewey onward simply deny that premise. We have invoked Dewey already. So we will close with Rorty, who argues that “a belief can still regulate action, can still be worth dying for, among people who are quite aware that this belief is caused by nothing deeper than contingent historical circumstance.”42 To deny this is to fail to appreciate the basic claims pragmatists advance about knowledge and truth and how they are established. In this book, we present an argument for the priority of democracy and, in doing so, offer to those who may think differently reasons to recognize that 40   “It is primarily a strategy of governments. While political repression may involve police or military action, it is defined by its political aims, rather than by the direct presence of force and violence. . . . From less to more severe repression might include: explicit and stringent public criticism of anti-­democratic political forces; monitoring the activities of individuals associated with the main relevant organizations; banning specified organizations as legal entities; and pursuing legal action against individuals who play a large role in sustaining militant and violent anti-­ democratic groups and projects” (Plotke 2006, 21). 41   Plotke (2006, 21–­24). 42   Rorty (1989, 189). As he puts the point elsewhere: “What matters for pragmatists is devising ways of diminishing human suffering and increasing human equality, increasing the ability of all human children to start life with an equal chance of happiness. This goal is not written in the stars, and is no more an expression of what Kant called ‘pure practical reason’ than it is the Will of God. It is a goal worth dying for, but it does not require backup from supernatural forces” (Rorty 1998, xxix).

285

C h ap te r 9

priority. There is some chance that we are mistaken. Our pragmatist commitments press us to countenance that possibility. We nevertheless are confident in the argument we advance. It would be ironic if we were to countenance a stance that amounted to little more than capitulation in the face of antidemocratic politics of whatever stripe. We do not do so. Nothing in pragmatism suggests that we should.

286

References

Ackerman, Bruce, and Ian Ayres. 2004. Voting with Dollars: A New Paradigm for Campaign Finance. New Haven, CT: Yale Univer­sity Press. Alchian, Armed A. 1950. “Uncertainty, Evolution and Economic Theory.” Journal of Political Economy 58:211–­21. Allio, Lorene, Mariusz Mark Dobek, Nikolai Mikhailov, and David Weimer. 1997. “Post-­communist Privatization as a Test of Theories of Institutional Change.” In The Political Economy of Property Rights, ed. David Weimer. Cambridge: Cambridge Univer­sity Press. Anderson, Elizabeth. 2006. “The Epistemology of Democracy.” Episteme 3:8–­22. Arrow, Kenneth. 1951/1963. Social Choice and Individual Values. 2nd ed. New Haven, CT: Yale Univer­sity Press. —­—­—­. 1972. “Gifts and Exchanges.” Philosophy and Public Affairs 1:343–­62. —­—­—­. 1974. The Limits of Organization. New York: W. W. Norton. —­—­—­. 1994. “Methodological Individualism and Social Knowledge.” American Economic Review 82(2):1–­9. Arrow, Kenneth, and Gerard Debreu. 1954. “Existence of an Equilibrium for a Competitive Market.” Econometrica 22(3):265–­290. Austen-­Smith, David. 1992. “Strategic Models of Talk in Political Decision-­Making,” International Political Science Review 13:45–­58. —­—­—­. 1995. “Modelling Deliberative Democracy.” Unpublished manuscript, Univer­sity of Rochester. Austen-­Smith, David, and Jeffrey Banks. 1998. “Social Choice Theory, Game Theory and Positive Political Theory.” Annual Review of Political Science 1:259–­87. —­—­—­. 1999. Positive Political Theory, I. Ann Arbor: Univer­ sity of Michi-­ ­gan Press. Austen-­Smith, David, and Timothy Feddersen. 2008. “In Response to Jurg Steiner’s ‘Concept Stretching: The Case of Deliberation.’” European Political Science 7:191–­93. Bachrach, Peter, and Morton Baratz. 1970. Power and Poverty. Oxford: Oxford Univer­ sity Press. Balinski, Michael, and H. Peyton Young. 2001. Fair Representation. 2nd ed. Washington, DC: Brookings Institution. Barber, Benjamin. 1996. “Foundationalism and Democracy.” In Democracy and Difference, ed. Seyla Benhabib. Princeton, NJ: Princeton Univer­sity Press. Bardhan, Pranab. 2005. Scarcity, Conflicts and Cooperation. Cambridge, MA: MIT Press. Barry, Brian. 1991. “Is Democracy Special?” In Democracy and Power: Essays in Political Theory, I. Oxford: Oxford Univer­sity Press.

R e fere n c e s Barry, Brian. 1995. Justice as Impartiality. Oxford: Oxford Univer­sity Press. Bartels. Larry. 2008. Unequal Democracy. Princeton, NJ: Princeton Univer­sity Press. Basu, Samy. 1999. “Dialogic Ethics and the Virtue of Humor.” Journal of Political Philosophy 7:378–­403. Bates, Robert. 1989. Beyond the Miracle of the Market: The Political Economy of Agrarian Development in Kenya. Cambridge: Cambridge Univer­sity Press. Bendor, Jonathan, and Adam Meirowitz. 2004. “Spatial Models of Delegation.” American Political Science Review 98(2): 293–­310. Benhabib, Seyla. 1996. “Toward a Deliberative Model of Democratic Legitimacy.” In Democracy and Difference, ed. Seyla Benhabib. Princeton, NJ: Princeton Univer­sity Press. Bernstein, Richard. 1986. “John Dewey on Democracy.” In Philosophical Profiles. Philadelphia: Univer­sity of Pennsylvania Press. —­—­—.­ 1992. “The Resurgence of Pragmatism.” Social Research 59:813–­40. —­—­—.­ 1998. “Community in the Pragmatic Tradition.” In The Revival of Pragmatism, ed. Morris Dickstein. Durham, NC: Duke Univer­sity Press. —­—­—­. 2005. The Abuse of Evil. Cambridge, MA: Polity Press. —­—­—­. 2010. The Pragmatic Turn. Cambridge, MA: Polity Press. Bertrand, Roman, Jean-­Louis Briquet, and Peter Pels, eds. 2006. The Hidden History of the Secret Ballot. Bloomington: Indiana Univer­sity Press. Binder, Sarah, and Steven Smith. 1997. “Political Goals and Procedural Choice in the Senate.” Journal of Politics 60:398–­416. Blais, André, Louis Massicotte, and Antoine Yoshinaka. 2001. “Deciding Who Has the Right to Vote.” Electoral Studies 20:41–­62. Blinder, Alan. 1999. Central Banking in Theory and Practice. Cambridge, MA: MIT Press. Bobbio, Norberto. 1996. Left and Right: The Significance of a Distinction. Chicago: Univer­sity of Chicago Press. Bohman, James. 1996a. “Modernization and Impediments to Democracy: The Problems of Hypercomplexity and Hyperrationality.” Theoria 86: 1–­20. —­—­—­. 1996b. Public Deliberation. Cambridge, MA: MIT Press. —­—­—.­ 1997. “Deliberative Democracy and Effective Social Freedom.” In Deliberative Democracy, ed. J. Bohman and W. Rehg. Cambridge, MA: MIT Press. —­—­—.­ 1998. “The Coming of Age of Deliberative Democracy.” Journal of Political Philosophy 6:400–­445. Bolton, Patrick, and Joseph Farrell. 1990. “Decentralization, Duplication and Delay.” Journal of Political Economy 98(4): 803–­26. Brady, Henry, and Stephen Ansolabehre. 1989. “The Nature of Utility Functions in Mass Publics.” American Political Science Review 83:143–­63. Brandom, Robert. 2000. Articulating Reasons. Cambridge, MA: Harvard Univer­sity Press. Braybrooke, David. 2004. Utilitarianism: Restorations, Repairs, Renovations. Toronto: Univer­sity of Toronto Press.

288

R e fere n c e s Brennan, Geoffrey, and James Buchanan. 1985. The Reason of Rule. Cambridge: Cambridge Univer­sity Press. Brennan, Geoffrey, and Phillip Pettit. 1990. “Unveiling the Vote.” British Journal of Political Science 20:311–­33. Brighouse, Harry. 1996. “Egalitarianism and Equal Availability of Political Influence.” Journal of Political Philosophy 4:118–­41. Brint Michael, W. Weaver, and M. Garmon. 1995. “What Difference Does Anti-­ Foundationalism Make to Political Theory?” New Literary History 26: 225–­37. Brown, J. L., and E. Pollitt. 1996. “Malnutrition, Poverty and Intellectual Development.” Scientific American 1996:38–­43. Buchanan, James. 1954. “Social Choice, Democracy and Free Markets.” Journal of Political Economy 62:114–­23. Calvert, Randall. 1992. “Leadership and Its Basis in Social Coordination.” International Political Science Review 13:7–­24. . 1995a. “The Rational Choice Theory of Social Institutions.” In Mod—­—­—­ ern Political Economy, ed. J. Banks and E. Hanushek. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1995b. “Rational Actors, Equilibrium and Social Institutions.” In Explaining Social Institutions, ed. J. Knight and I. Sened. Ann Arbor: Univer­sity of Michigan Press. Calvert, Randall, and James Johnson. 1999. “Interpretation and Coordination in Constitutional Politics.” In Lessons in Democracy, ed. Ewa Hauser and Jacek Wasilewski. Krakow: Jagiellonian Univer­sity Press. Campbell, James. 1998. “Dewey’s Conception of Community.” In Reading Dewey, ed. Larry Hickman. Bloomington: Indiana Univer­sity Press. Caplan, Andrew, and Barry Nalebuff. 1988. “On 64% Majority Rule.” Econometrica 56:787–­814. Chambers, Simone. 1996. Reasonable Democracy. Ithaca, NY: Cornell Univer­sity Press. —­—­—­. 2003. “Deliberative Democratic Theory.”Annual Review of Political Science 6:307–­26. Chan, Joseph, and David Miller. 1991. “Elster on Self-­Realization in Politics.” Ethics 102:96–­102. Christiano, Thomas. 1996. The Rule of the Many. Boulder, CO: Westview Press. . 2004. “The Authority of Democracy.” Journal of Political Philosophy —­—­—­ 12:266–­90. Coase, Ronald. 1960. “The Problem of Social Cost.” Journal of Law and Economics 3:1–­44. —­—­—­. 1988. The Firm, the Market, and the Law. Chicago: Univer­sity of Chicago Press. Cohen, G. A. 2008. Rescuing Justice and Equality. Cambridge, MA: Harvard Univer­ sity Press. Cohen, Joshua. 1986a. “An Epistemic Conception of Democracy.” Ethics 97:26–­38.

289

R e fere n c e s Cohen, Joshua. 1986b. “Reflections on Rousseau.” Philosophy and Public Affairs 15:275–­97. —­—­—.­ 1989a. “Deliberation and Democratic Legitimacy.” In The Good Polity, ed. Alan Hamlin and Phillip Pettit. Oxford: Basil Blackwell, 1989. —­—­—.­ 1989b. “The Economic Basis of Deliberative Democracy.” Social Philosophy and Policy 6:25–­50. —­—­—.­ 1994a. “A More Democratic Liberalism.” Michigan Law Review 92:1503–­46. —­—­—.­ 1994b. “Pluralism and Proceduralism.” Chicago-­Kent Law Review 69:589–­618. —­—­—.­ 1995. “Sen on Equality.” Journal of Philosophy 92: 275–­88. —­—­—.­ 1996. “Procedure and Substance in Deliberative Democracy.” In Democracy and Difference, ed. Seyla Benhabib. Princeton, NJ: Princeton Univer­sity Press. —­—­—.­ 2001. “Money, Politics and Political Equality.” In Fact and Value, ed. Alex Byrne, Robert Stalnaker, and Ralph Wedgwood. Cambridge, MA: MIT Press. Cohen, Joshua, and Joel Rogers. 1995. Associations and Democracy. London: Verso. Cohen, Joshua, and Charles Sabel. 1997. “Directly Deliberative Polyarchy.” European Law Journal 3:313–­42. Coleman, Jules, and John Ferejohn. 1986. “Democracy and Social Choice.” Ethics 97:6–­25. Conover, Pamela Johnston, Donald Searing, and Ivor Crewe. 2002. “The Deliberative Potential of Political Discussion.” British Journal of Political Science 32:21–­62. Cooke, Maeve. 2000. “Five Arguments for Deliberative Democracy.” Political Studies 48:947–­69. Cotterrell, Roger. 1995. Law’s Community. Oxford: Clarendon Press. Cox, Gary. 1997. Making Votes Count. Cambridge: Cambridge Univer­sity Press. Cox, Gary, and Kenneth Shepsle. 2007. “Majority Cycling and Agenda Manipulation.” In Positive Changes in Political Science, ed. John Aldrich, James Alt, and Arthur Lupia. Ann Arbor: Univer­sity of Michigan Press. Crimp, Douglas, with Adam Rolston. 1990. Aids Demo Graphics. Seattle: Bay Press. Crimp, Douglas. 2002. Melancholia and Moralism. Cambridge, MA: MIT Press. Dagger, Richard. 2000. “Membership, Fair Play, and Political Obligation.” Political Studies 48:104–­17. Dahl, Robert. 1989. Democracy and Its Critics. New Haven, CT: Yale Univer­sity Press. —­—­—.­1996. “Democratic Theory and Democratic Experience.” In Democracy and Difference, ed. Seyla Benhabib. Princeton, NJ: Princeton Univer­sity Press. Dasgupta, Partha, and Eric Maskin. 1999. “Democracy and Other Goods.” In Democracy’s Value, ed. Ian Shapiro and Casiano Hacker-­Cordón. Cambridge: Cambridge Univer­sity Press. —­—­—.­ 2008. “On the Robustness of Majority Rule.” Journal of the European Economic Association 6:949–­73. Davidson, Donald. 1980. Actions, Reasons and Causes. Oxford: Oxford Univer­sity Press. Dekel, Eddie, Barton Lipman, and Aldo Rustichini. 1998. “Recent Developments in Modeling Unforeseen Contingencies.” European Economic Review 42:523–­42.

290

R e fere n c e s Delli Carpini, Michael, Fay Lomax Cook, and Lawrence R. Jacobs. 2004. “Public Deliberation, Discursive Participation, and Citizen Engagement: A Review of the Empirical Literature,” Annual Review of Political Science 7:315–­44. Dewey, John. 1911. “The Problem of Truth.” In The Essential Dewey, ed. L. Hickman. Bloomington: Indiana Univer­sity Press. —­—­—­. 1916. Democracy and Education. New York: Free Press. . 1919/1993. “Philosophy and Democracy.” In The Political Writings, ed. —­—­—­ I. Shapiro and D. Morris. Indianapolis: Hackett. —­—­—­. 1920/1948. Reconstruction in Philosophy. Boston: Beacon Press. —­—­—­. 1927. The Public and Its Problems. Chicago: Swallow Press. —­—­—­. 1929/1958. Experience and Nature. Mineola, NY: Dover Press. —­—­—­. 1935/1991. The Later Works, 1925–­1953. Volume 11: 1935–­1937, ed. Jo Ann Boydston. Carbondale: Southern Illinois Univer­sity Press. —­—­—­. 1938/1980. Logic: The Theory of Inquiry. The Later Works, 1925–­1953. Volume 12: 1938, ed. Jo Ann Boydston. Carbondale: Southern Illinois Univer­sity Press. —­—­—­. 1939/1993. “Creative Democracy—­The Task Before Us.” In The Political Writings, ed. I. Shapiro and D. Morris. Indianapolis: Hackett. —­—­—­. 1980. The Middle Works, 1899–­1924. Volume 10: 1916–­1917, ed. Jo Ann Boydston. Carbondale: Southern Illinois Univer­sity Press. Dickstein, Morris. 1998. The Revival of Pragmatism: New Essays on Social Thought, Law, and Culture. Durham, NC: Duke Univer­sity Press. Dixit, Avinash. 2009. “Governance Institutions and Economic Activity.” American Economic Review 99:5–­24. Doron, Gideon, and Itai Sened. 2001. Political Bargaining. Thousand Oaks, CA: Sage Publications. Dryzek, John. 1996. Democracy in Capitalist Times. Oxford: Oxford Univer­sity Press. —­—­—­. 2000. Deliberative Democracy and Beyond. Oxford: Oxford Univer­sity Press. Dryzek, John, and Christian List. 2003. “Social Choice Theory and Deliberative Democracy: A Reconciliation.” British Journal of Political Science 33:1–­28. Dummett, Michael. 1984. Voting Procedures. Oxford: Oxford Univer­sity Press. —­—­—­. 1997. Principles of Electoral Reform. Oxford: Oxford Univer­sity Press. Dworkin, Ronald. 1987. “What Is Equality, Part 4: Political Equality.” Univer­sity of San Francisco Law Review 22:1–­30. —­—­—­. 1988. Law’s Empire. Cambridge, MA: Harvard Univer­sity Press. Eggertsson, Thrain. 1990. Economic Behavior and Institutions. Cambridge: Cambridge Univer­sity Press. Elster, Jon. 1979. Ulysses and the Sirens. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1983. Sour Grapes. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1986. “The Market and the Forum.” In The Foundations of Social Choice Theory, ed. Jon Elster and A. Hylland. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1995. “Strategic Uses of Argument.” In Barriers to Conflict Resolution, ed. Kenneth Arrow, Lawrence Susskind, and Richard Zeckhause. New York: W. W. Norton.

291

R e fere n c e s Enelow, James. 1997. “Cycling and Majority Rule.” In Perspectives on Public Choice, ed. Dennis Mueller. Cambridge: Cambridge Univer­sity Press. Ensminger, Jean. 1992. Making a Market: The Institutional Transformation of an African Society. Cambridge: Cambridge Univer­sity Press. Epstein, Lee, and Jack Knight. 1998. The Choices Justices Make. Washington, DC: Congressional Quarterly Press. Epstein, Lee, Jack Knight, and Andrew Martin. 2003. “The Norm of Prior Judicial Experience (and Its Consequences for the U.S. Supreme Court).” California Law Review 91(4): 903–­66. Epstein, Steven. 1997. Impure Science: Aids, Activism and the Politics of Knowledge. Berkeley: Univer­sity of California Press. —­—­—.­ 2000. “Democracy, Expertise and AIDS Treatment Activism.” In Science, Technology and Democracy, ed. Daniel Kleinman. Albany: State Univer­sity of New York Press. Estlund, David. 1993. “Making Truth Safe for Democracy.” In The Idea of Democracy, ed. David Copp, Jean Hampton, and John Roemer. Cambridge: Cambridge Univer­ sity Press. Estlund, David, Jeremy Waldron, Bernard Grofman, and Scott Feld. 1989. “Democratic Theory and the Public Interest: Condorcet and Rousseau Revisited.” American Political Science Review 83:1317–­40. Farrar, Cynthia, James S. Fishkin, Donald P. Green, Christian List, Robert C. Luskin, and Elizabeth Levy Paluck. 2010. “Disaggregating Deliberation’s Effects: An Experiment within a Deliberative Poll.” British Journal of Political Science 40:333–­47. Farrell, Joseph. 1987. “Information and the Coase Theorem.” Journal of Economic Perspectives 1:113–­29. Farrell, Joseph, and Matthew Rabin. 1996. “Cheap Talk.” Journal of Economic Perspectives 10:103–­18. Fearon, James. 1998. “Deliberation as Discussion.” In Deliberative Democracy, ed. Jon Elster. Cambridge: Cambridge Univer­sity Press. Femia, Joseph. 1996. “Complexity and Deliberative Democracy.” Inquiry 39:359–­98. Ferejohn, John. 1993. “Must Preferences Be Respected in a Democracy?” In The Idea of Democracy, ed. David Copp, Jean Hampton, and John Roemer. Cambridge: Cambridge Univer­sity Press. —­—­—.­ 2000. “Instituting Deliberative Democracy.” In Designing Democratic Institutions: Nomos XLII, ed. Ian Shapiro and Stephen Macedo. New York: NYU Press. Festenstein, Matthew. 1997. Pragmatism and Political Theory. Chicago: Univer­sity of Chicago Press. —­—­—.­ 2003. “Pragmatism’s Boundaries.” Millennium 31:549–­71. Forbes, H. D. 1997. Ethnic Conflict: Commerce, Culture, and the Contact Hypothesis. New Haven, CT: Yale Univer­sity Press. Fraser, Nancy. 1992. “Rethinking the Public Sphere.” In Habermas and the Public Sphere, ed. Craig Calhoun. Cambridge, MA: MIT Press.

292

R e fere n c e s Freeman, Samuel. 2000. “Deliberative Democracy: A Sympathetic Comment.” Philosophy and Public Affairs 29:371–­417. Fuller, Lon. 1969. The Morality of Law, 2nd ed. New Haven, CT: Yale Univer­sity Press. —­—­—­. 1981. The Problem of Social Order. Durham, NC: Duke Univer­sity Press. Fung, Archon. 2001. “Accountable Autonomy: Toward Empowered Deliberation in Chicago Schools and Policing.” Politics and Society 29:73–­104. Geuss, Raymond. 2001. History and Illusion in Politics. Cambridge: Cambridge Univer­sity Press. —­—­—­. 2008. Philosophy and Real Politics. Princeton, NJ: Princeton Univer­sity Press. Gibbard, Alan. 1973. “Manipulation of Voting Schemes: A General Result.” Econometrica 41:587–­602. Gilbert, Margaret. 2006. A Theory of Political Obligation. Oxford: Oxford Univer­sity Press. Goodin, Robert. 1986 “Laundering Preferences.” In Foundations of Social Choice Theory, ed. Jon Elster and Aanund Hylland. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1995. Utilitarianism as a Public Philosophy. Cambridge: Cambridge Univer­ sity Press. . 2000. “Democratic Deliberation Within.” Philosophy and Public Affairs —­—­—­ 29:81–­109. —­—­—­. 2008. “Deliberative Lies.” European Political Science 7:194–­98. Goodin, Robert, and Christian List. 2006. “A Conditional Defense of Plurality Rule.” American Journal of Political Science 50:940–­49. Greif, Avner. 1997. “Microtheory and Recent Developments in the Study of Economic Institutions through Economic History.” In Advances in Economics and Econometrics, vol. 2, ed. D. Kreps and K. Walks. Cambridge: Cambridge Univer­sity Press. Greif, Avner, Paul Milgrom, and Barry Weingast. 1994. “Coordination, Commitment, and Enforcement.” Journal of Political Economy 102:745–­76. Guinier, Lani. 1995. The Tyranny of the Majority. New York: Free Press. Gunn, Giles. 1994. “Response to Robert Westbrook.” American Quarterly 46:297–­303. Gutmann, Amy. 1993. “The Challenge of Multiculturalism in Political Ethics.” Philosophy and Public Affairs 22:171–­206. —­—­—­. 1996. “Democracy, Philosophy and Justification.” In Democracy and Difference, ed. Seyla Benhabib. Princeton, NJ: Princeton Univer­sity Press. —­—­—­. 1998. Freedom of Association. Princeton, NJ: Princeton Univer­sity Press. Gutmann, Amy, and Dennis Thompson. 1990. “Moral Conflict and Political Consensus.” Ethics 101:64–­88. —­—­—­. 1996. Democracy and Disagreement. Cambridge, MA: Harvard Univer­sity Press. Habermas, Jürgen. 1985. “Civil Disobedience: Litmus Test for the Democratic Constitutional State.” Berkeley Journal of Sociology 30:95–­116. —­—­—­. 1996. Between Facts and Norms. Cambridge, MA: MIT Press.

293

R e fere n c e s Hacking, Ian. 2007. “On Not Being a Pragmatist.” In New Pragmatists, ed. Cheryl Misak. Oxford: Oxford Univer­sity Press. Hall, Peter, and Rosemary Taylor. 1996. “Political Science and the Three New Institutionalisms.” Political Studies 44 (5): 952–­73. Hardin, Russell. 1982. Collective Action. Baltimore, MD: Johns Hopkins Univer­sity Press. —­—­—­. 1988. Morality within the Limits of Reason. Chicago: Univer­ sity of Chicago Press. —­—­—­. 1995. One for All: The Logic of Group Conflict. Princeton, NJ: Princeton Univer­sity Press. —­—­—­. 2009. How Do You Know? The Economics of Ordinary Knowledge. Princeton, NJ: Princeton Univer­sity Press. Hart, H.L.A. 1958. “Legal and Moral Obligation.” In Essays in Moral Philosophy, ed. A. I. Melden. Seattle: Univer­sity of Washington Press. —­—­—­. 1994. The Concept of Law. Oxford: Clarendon Press. Hayek, F. A. 1944. The Road to Serfdom. Chicago: Univer­sity of Chicago Press. —­—­—­. 1948. Individualism and Economic Order. Chicago: Univer­sity of Chicago Press. —­—­—­. 1976. Law, Legislation and Liberty. Vol. 2, The Mirage of Social Justice. Chicago: Univer­sity of Chicago Press. —­—­—.­ 1984. “Competition as a Discovery Procedure.” In The Essence of Hayek, ed. Chiaki Nishiyama and Kurt R. Leube. Stanford, CA: Hoover Institution Press. Heath, Joseph. 2001. Communicative Action and Rational Choice. Cambridge, MA: MIT Press. Hechter, Michael, and Karl-­Dieter Opp. 2001. Social Norms. New York: Russell Sage Foundation. Heckathorn, Douglas, and Steven Maser. 1987. “Bargaining and Constitutional Contracts.” American Journal of Political Science 31:142–­68. Herron, Michael. 2000. “Estimating the Economic Impact of Political Party Competition in the 1992 British Election.” American Journal of Political Science 44:326–­37. Hirschman, Albert. 1991. The Rhetoric of Reaction. Cambridge, MA: Harvard Univer­ sity Press. —­—­—.­ 1993. “The Rhetoric of Reaction—­Two Years Later.” Government and Opposition 28:292–­314. Hochschild, Jennifer, and Nathan Scorvonick. 2003. The American Dream and the Public Schools. Oxford: Oxford Univer­sity Press. Holmes, Oliver Wendell. 1881. The Common Law. Boston: Little, Brown. Holmes, Stephen. 1988. “Gag Rules or the Politics of Omission.” In Constitutionalism and Democracy, ed. Jon Elster and Rune Slagstad. Cambridge: Cambridge Univer­ sity Press. Honneth, Axel. 1998. “Democracy as Reflexive Cooperation.” Political Theory 26:763–­83.

294

R e fere n c e s Hume, David. 1752/1953. “Of the Original Contract.” In David Hume’s Political Essays, ed. C. W. Hendel. Indianapolis: Bobbs-­Merrill. Hurwicz, Leonid 2008. “But Who Will Guard the Guardians?” American Economic Review 98:577–­85. Isaac, Jeffrey. 1998. Democracy in Dark Times. Ithaca, NY: Cornell Univer­sity Press. —­—­—­. 1999. “Is the Revival of Pragmatism Practical, or What Are the Consequences of Pragmatism?” Constellations 6:561–­87. —­—­—­. 2003. The Poverty of Progressivism. Lanham, MD: Rowman and Littlefield. Johnson, James. 1991. “Habermas on Strategic and Communicative Action.” Political Theory 19:181–­201. —­—­—­. 1993. “Is Talk Really Cheap?: Prompting Conversation Between Critical Theory and Rational Choice.” American Political Science Review 87:74–­86. —­—­—­. 1997. “Communication, Criticism, and the Postmodern Consensus.” Political Theory 25:559–­83. —­—­—­. 1998. “Arguing for Deliberation: Some Skeptical Considerations.” In Deliberative Democracy, ed. Jon Elster. Cambridge: Cambridge Univer­sity Press. —­—­—­. 2000. Why Respect Culture?” American Journal of Political Science 44:405–­18. —­—­—­. 2001. “Inventing Constitutional Traditions.” In Constitutional Culture and Democratic Rule, ed. J. Ferejohn, J. Rakove and J. Riley. Cambridge: Cambridge Univer­sity Press. —­—­—­. 2002. “Liberalism and the Politics of Cultural Authenticity,” Politics, Philosophy and Economics 1:213–­36. —­—­—­. 2005. “What the Politics of Enfranchisement Can Tell Us About How Rational Choice Theorists Study Institutions.” In Preferences and Situations, ed. Ira Katznelson and Barry Weingast. New York: Russell Sage Foundation. —­—­—­. 2006. “Consequences of Positivism: A Pragmatist Assessment.” Comparative Political Studies 39:224–­52. —­—­—­. 2010. “What Rationality Assumption? Why ‘Positive Political Theory’ Rests on a Mistake.” Political Studies 58:282–­99. Johnson, Paul. 1998. Social Choice: Theory and Research. Quantitative Applications in the Social Sciences no. 123. Thousand Oaks, CA: Sage Publications. Keech, William. 1995. Economic Politics: The Costs of Democracy. Cambridge: Cambridge Univer­sity Press. Kitcher, Phillip. 2006. “Public Knowledge and the Difficulties of Democracy.” Social Research 73:1205–­24. Kloppenberg, James. 1998. “Pragmatism: An Old Name for Some New Ways of Thinking?” In The Revival of Pragmatism, ed. Morris Dickstein. Durham, NC: Duke Univer­sity Press. Klosko, George. 2004. The Principle of Fairness and Political Obligation, 2nd ed. Lanham, MD: Rowman and Littlefield. Knight, Jack. 1992. Institutions and Social Conflict. Cambridge: Cambridge Univer­sity Press.

295

R e fere n c e s Knight, Jack. 1993. “Public Goods.” In The Oxford Companion to Politics of the World, ed. Joel Krieger. Oxford: Oxford Univer­sity Press. —­—­—.­ 1995. “Models, Interpretations and Theories: Constructing Explanations of Institutional Emergence and Change.” In Explaining Social Institutions, ed. J. Knight and I. Sened. Ann Arbor: Univer­sity of Michigan Press. —­—­—.­ 2001. “A Pragmatist Approach to the Proper Scope of Government.” Journal of Institutional and Theoretical Economics 157:28–­48. —­—­—.­ 2009. “Comment: Causal Mechanisms and Generalizations.” In Philosophy of the Social Sciences, ed. C. Mantzavinos. Cambridge: Cambridge Univer­sity Press. Knight, Jack, and Lee Epstein. 1996. “On the Struggle for Judicial Supremacy.” Law and Society Review 30 (1): 87–­120. Knight, Jack, and James Johnson. 1993. “Public Choice and the Rule of Law: Rational Choice Theories of Statutory Interpretation.” In Nomos XXXVI: The Rule of Law, ed. Ian Shapiro. New York: NYU Press. —­—­—.­ 1994. Aggregation and Deliberation.” Political Theory 22:277–­96. —­—­—.­ 1996. “Political Consequences of Pragmatism.” Political Theory 24:68–­96. —­—­—.­ 1997. “What Kind of Equality Does Democratic Deliberation Require?” In Deliberative Democracy, ed. J. Bohman and W. Rehg. Cambridge, MA: MIT Press. —­—­—.­ 1999. “Inquiry into Democracy.” American Journal of Political Science 43:566–­89. —­—­—.­ 2007. “On the Priority of Democracy: A Pragmatist Approach to Political-­ Economic Institutions and the Burden of Justification.” American Political Science Review 101:47–­62. Knight, Jack, and Gary Miller. 2007. “Democracy, Bureaucracy and Credible Commitment.” Working paper, Washington Univer­sity in St. Louis. Knight, Jack, and Douglass North. 1997. “Explaining the Complexity of Institutional Change.” In The Political Economy of Property Rights, ed. David Weimer. Cambridge: Cambridge Univer­sity Press. Kreps, David. 1990a. “Corporate Culture and Economic Theory.” In Perspectives on Positive Political Economy, ed. James Alt and Kenneth Shepsle. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1990b. A Course in Microeconomic Theory. Princeton Univer­sity Press. —­—­—­. 1990c. Game Theory and Economic Modeling. Oxford: Oxford Univer­sity Press. Kuran, Timur. 1990. “Private and Public Preferences.” Economics and Philosophy 6:1–­26. —­—­—.­ 1993. “The Unthinkable and the Unthought.” Rationality and Society 5:473–­505. Kymlicka, Will. 1995. Multicultural Citizenship: A Liberal Theory of Minority Rights. Oxford: Oxford Univer­sity Press. Landa, Dimitri, and Adam Meirowitz. 2009. “Game Theory, Information and Deliberative Democracy,” American Journal of Political Science 53: 427–­44. Laudan, Larry. 1996. Beyond Positivism and Relativism. Boulder, CO: Westview Press.

296

R e fere n c e s Levi, Margaret. 1997. Consent, Dissent and Patriotism. Cambridge: Cambridge Univer­sity Press. Levin, Jonathan, and Barry Nalebuff. 1995. “An Introduction to Vote Counting Schemes.” Journal of Economic Perspectives 9:3–­26. Lewis, David. 1969. Convention. Cambridge, MA: Harvard Univer­sity Press. Lindblom, Charles. 1980. Politics and Markets: The World’s Political Economic Systems. New York: Basic Books. —­—­—­. 2001. The Market System. New Haven, CT: Yale Univer­sity Press. Lippmann, Walter. 1927/1993. The Phantom Public. Piscataway, NJ: Transaction Publishers. Lipset, Seymour Martin. 1960. Political Man. Baltimore: Johns Hopkins Univer­sity Press. List, Christian. 2002a. “Intradimensional Single-­Peakedness and the Multidimensional Arrow Problem.” Theory and Decision 52:287–­301. —­—­—­. 2002b. “Two Concepts of Agreement.” Good Society 11:72–­79. List, Christian, and Robert Goodin. 2001. “Epistemic Democracy: Generalizing the Condorcet Jury Theorem.” Journal of Political Philosophy 9:277–­306. List, Christian, Robert C. Luskin, James S. Fishkin, and Iain McLean. 2009. “Deliberation, Single-­Peakedness, and the Possibility of Meaningful Democracy: Evidence from Deliberative Polls.” Unpublished manuscript, London School of Economics and Political Science. List, Christian, and Phillip Pettit. 2002. “Aggregating Sets of Judgements: An Impossibility Result.” Economics and Philosophy 18:89–­110. Loury, Glenn. 1994. “Self-­Censorship in Public Discourse.” Rationality and Society 6:428–­61. Lukes, Stephen. 2005. Power: A Radical View, 2nd ed. London: Palgrave. Lummis, Douglas. 1996. Radical Democracy. Ithaca, NY: Cornell Univer­sity Press. Luskin, Robert, James Fishkin, and Roger Jowell. 2002. “Considered Opinions: Deliberative Polling in Britain.” British Journal of Political Science 32:455–­88. Mackie, Gerry. 2003. Democracy Defended. Cambridge: Cambridge Univer­sity Press. —­—­—­. 2006. “Does Deliberation Change Minds?” Politics, Philosophy and Economics 5: 279–­303. —­—­—­. 2009. “Schumpeter’s Leadership Democracy” Political Theory 37:128–­53. Manin, Bernard. 1987. “On Legitimacy and Political Deliberation.” Political Theory 15:338–­68. —­—­—­1994. “The Metamorphoses of Representative Government.” Economy and Society 23:133–­171. —­—­—­. 1997. Principles of Representative Government. Cambridge: Cambridge Univer­sity Press. Mansbridge, Jane. 1996. “Using Power/Fighting Power: The Polity.” In Democracy and Difference, ed. Seyla Benhabib. Princeton, NJ: Princeton Univer­sity Press. —­—­—­. 1999a. “Everyday Talk in the Deliberative System.” In Deliberative Politics, ed. S. Macedo. Oxford: Oxford Univer­sity Press.

297

R e fere n c e s Mansbridge, Jane.1999b. “On the Idea that Participation Makes Better Citizens.” In Citizen Competence and Democratic Institutions, ed. S. Elkin and K. Soltan. Univer­sity Park, PA: Penn State Univer­sity Press. Mansbridge, Jane, James Bohman, Simone Chambers, David Estlund, Andreas Føllesdal, Archon Fung, Cristina Lafont, Bernard Manin, and José Luis Martí. 2010. “The Place of Self-­Interest and the Role of Power in Deliberative Democracy.” Journal of Political Philosophy 18:64–­100. Maskin, Eric. 2008. “Mechanism Design: How to Implement Social Goals.” American Economic Review 98: 567–­76. Massicotte, Louis, Andre Blais, and Antoine Yoshinaka. 2004. Establishing the Rules of the Game. Toronto: Univer­sity of Toronto Press. Matthews, Steven, and Andrew Postlewaite. 1995. “On Modeling Cheap Talk in Bayesian Games.” In The Economics of Informational Decentralization, ed. John Ledyard. Dordrecht, Netherlands: Kluwer Academic. May, Kenneth. 1952. “A Set of Independent Necessary and Sufficient Conditions for Simple Majority Decision.” Econometrica 20:680–­84. McCubbins, Mathew, Roger Noll, and Barry Weingast. 1987. “Administrative Procedures as Instruments of Political Control.” Journal of Law, Economics, and Organization 3 (2): 243–­77. McKelvey, Richard. 1976. “Intransitivities in Multidimensional Voting Models and Some Implications for Agenda Control.” Journal of Economic Theory 12:472–­82. McKelvey, Richard, and Talbot Page. 1999. “Taking the Coase Theorem Seriously.” Economics and Philosophy 15:235–­47. —­—­—.­ 2000. “An Experimental Study of the Effect of Private Information in the Coase Theorem.” Experimental Economics 3:187–­213. McLean, Iain, and Arnold Urkin, eds. 1997. Classics of Social Choice. Ann Arbor: Univer­sity of Michigan Press. McMillan, John. 2002. Reinventing the Bazaar: A Natural History of Markets. New York: W. W. Norton. Menand, Louis. 2001. The Metaphysical Club. New York: Farrar, Straus and Giroux. Mendez, Michelle, and Linda Adair. 1999. “Severity and Timing of Stunting in the First Two Years of Life Affect Performance on Cognitive Tests in Late Childhood.” Journal of Nutrition 129:1555–­62. Michnik, Adam. 1998. “Postface—­Gray Is Beautiful.” In Letters from Freedom. Berkeley: Univer­sity of California Press. Milgrom, Paul, Douglass North, and Barry Weingast. 1990. “The Role of Institutions in the Revival of Trade.” Politics and Economics 2:1–­23. Mill, John Stuart. 1861/1991. Considerations on Representative Government. Amherst, NY: Prometheus Books. Miller, David. 1992. “Deliberative Democracy and Social Choice.” Political Studies 40:54–­67. Miller, Gary. 1992. Managerial Dilemmas: The Political Economy of Hierarchy. Cambridge: Cambridge Univer­sity Press.

298

R e fere n c e s —­—­—­. 2005. “The Political Evolution of Principal-­Agent Models.” Annual Review of Political Science 8:2003–­25. Miller, Gary, and Thomas Hammond. 1994. “Why Politics Is More Fundamental than Economics: Incentive-­Compatible Mechanisms Are Not Credible.” Journal of Theoretical Politics 6:5–­26. Miller, Nicholas. 1983. “Pluralism and Social Choice.” American Political Science Review 77:134–­47. Misak, Cheryl. 2000. Truth, Politics, Morality. London: Routledge. Mishel, Lawrence, Jared Bernstein, and Heidi Shierholz. 2009. The State of Working America, 2008–­2009. Ithaca, NY: Cornell Univer­sity Press. Montesquieu, Charles de. 1989. The Spirit of the Laws. Cambridge: Cambridge Univer­sity Press. Mouffe, Chantal. 1999. “Deliberative Democracy or Agonistic Pluralism?” Social Research 66:745–­58. Mueller, Dennis. 1989. Public Choice II. Cambridge: Cambridge Univer­sity Press. Murphy, Walter, C. Herman Pritchett, Lee Epstein, and Jack Knight. 2005. Courts, Judges and Politics, 6th ed. New York: McGraw-­Hill. Mutz, Diana. 2008. “Is Deliberative Democracy a Falsifiable Theory?” Annual Review of Political Science 11:521–­38. Myerson, Roger. 1999. “Theoretical Comparisons of Electoral Systems.” European Economic Review 43:671–­97. —­—­—­. 2008. “Perspectives on Mechanism Design in Economic Theory.” American Economic Review 98:586–­603 Myles, Gareth. 1995. Public Economics. Cambridge: Cambridge Univer­sity Press. Nagel, Thomas. 1989. “What Makes a Political Theory Utopian?” Social Research 56:903–­20. Niemi, Richard. 1969. “Majority Decision-­Making with Partial Unidimensionality.” American Political Science Review 63:488–­97. Niemi, Richard, and Steven Wright. 1987. “Voting Cycles and the Structure of Individual Preferences.” Social Choice and Welfare 4:173–­83. North, Douglass. 1990a. Institutions, Institutional Change, and Economic Performance. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1990b. “A Transaction Cost Theory of Politics.” Journal of Theoretical Politics 2 (4): 355–­67. North, Douglass, and Barry Weingast. 1989. “Constitutions and Commitment.” Journal of Economic History 49:803–­32. Nozick, Robert. 1993. The Nature of Rationality. Princeton, NJ: Princeton Univer­sity Press. Ober, Josiah. 2008. Democracy and Knowledge: Innovation and Knowledge in Classical Athens. Princeton, NJ: Princeton Univer­sity Press. O’Donnell, Guillermo. 2001. “Democracy, Law and Comparative Politics.” Studies in Comparative International Development 36:7–­36. Office of Judges Programs. 2003. Understanding the Federal Courts. Washington, DC: Office of the Judges Programs, Administrative Office of the U.S. Courts.

299

R e fere n c e s Ordeshook, Peter. 1997. “The Spatial Analysis of Elections and Committees.” In Perspectives on Public Choice, ed. Dennis Mueller. Cambridge: Cambridge Univer­sity Press. Ostrom, Elinor. 1990. Governing the Commons. Cambridge: Cambridge Univer­sity Press. —­—­—.­ 1999. “Coping with Tragedies of the Commons.” Annual Review of Political Science 2:493–­535. —­—­—­. 2005. Understanding Institutional Diversity. Princeton, NJ: Princeton Univer­ sity Press. Page, Scott. 2007. The Difference. Princeton, NJ: Princeton Univer­sity Press. Pateman, Carole. 1970. Participation and Democratic Theory. Cambridge: Cambridge Univer­sity Press. Peirce, Charles. 1958. Selected Writings. Mineola, NY: Dover Press. Persson, Torsten, and Guido Tabellini. 2000. Political Economics. Cambridge, MA: MIT Press. Pettit, Philip. 1997. Republicanism: A Theory of Freedom and Government. Oxford: Oxford Univer­sity Press. —­—­—­. 2001. A Theory of Freedom: From the Psychology to Politics of Agency. Oxford: Oxford Univer­sity Press. Phillips, Anne. 1995. The Politics of Presence. Oxford: Oxford Univer­sity Press. Pigou, Arthur C. 1912. Wealth and Welfare. London: Macmillan. Piketty, Thomas, and Emmanuel Saez. 2003. “Income Inequality in the United States, 1913–­1998.” Quarterly Journal of Economics 118:1–­39. Pitkin, Hannah. 1967. The Concept of Representation. Berkeley: Univer­sity of California Press. Plotke, David. 1997. “Representation Is Democracy.” Constellations 4:19–­34. —­—­—.­ 2006. “Democratic Polities and Anti-­democratic Politics.” Theoria 53:6–­44. Plott, Charles. 1967. “A Notion of Equilibrium and Its Possibility under Majority Rule.” American Economic Review 57: 787–­806. Polletta, Francesca. 2004. Freedom Is an Endless Meeting: Democracy in American Social Movements. Chicago: Univer­sity of Chicago Press. Posner, Richard. 2003. Law, Pragmatism and Democracy. Cambridge, MA: Harvard Univer­sity Press. Przeworski, Adam. 1998. “Deliberation and Ideological Domination.” In Deliberative Democracy, ed. Jon Elster. Cambridge: Cambridge Univer­sity Press. —­—­—.­ 1999. “A Minimalist Conception of Democracy.” In Democracy’s Value, ed. Ian Shapiro and Casiano Hacker-­Cordón. Cambridge: Cambridge Univer­sity Press. —­—­—­. 2003. States and Markets. Cambridge: Cambridge Univer­sity Press. Przeworski, Adam, Michael Alvarez, Jose Antonio Cheibub, and Fernando Limongi. 2000. Democracy and Development. Cambridge: Cambridge Univer­sity Press. Przeworski, Adam, and Michael Wallerstein. 1988. “Structural Dependence of the State on Capital.” American Political Science Review 82:11–­29. Putnam, Hilary. 1991. “A Reconsideration of Deweyian Democracy.” In Pragmatism in Law and Society, ed. Michael Brint and William Weaver. Boulder, CO: Westview Press.

300

R e fere n c e s —­—­—­. 1994. Words and Life. Cambridge, MA: Harvard Univer­sity Press. —­—­—­. 1995a. Pragmatism. London: Blackwell. —­—­—­. 1995b. “Pragmatism.” Proceedings of the Aristotelian Society 95:291–­306. —­—­—­. 2002. The Collapse of the Fact/Value Dichotomy and Other Essays. Cambridge, MA: Harvard Univer­sity Press. —­—­—­. 2004. Ethics without Onthology. Cambridge, MA: Harvard Univer­sity Press. Radcliff, Benjamin. 1993. “The Structure of Voter Preferences.” Journal of Politics 55:714–­19. Radin, Margaret. 1994–­95. “A Deweyian Perspective on the Economic Theory of Democracy.” Constitutional Commentary 11:539–­56. Rae, Douglas, and Eric Schickler. 1997. “Majority Rule.” In Perspectives on Public Choice, ed. Dennis Mueller. Cambridge: Cambridge Univer­sity Press. Rawls, John. 1989. “The Domain of the Political and Overlapping Consensus.” New York Univer­sity Law Review 64 (2): 233–­55. —­—­—­. 1993. Political Liberalism. New York: Columbia Univer­sity Press. —­—­—­. 2001. Justice as Fairness. Cambridge, MA: Harvard Univer­sity Press. Raz, Joseph. 1999. “The Obligation to Obey: Revision and Tradition.” In The Duty to Obey the Law, ed. W. A. Edmundson. Lanham, MD: Rowman and Littlefield. Reed, Adolph. 1986. The Jesse Jackson Phenomenon. New Haven, CT: Yale Univer­sity Press. Remer, Gary. 1999. “Political Oratory and Conversation—­Cicero versus Deliberative Democracy.” Political Theory 27:39–­64. —­—­—­. 2000. “Two Models of Deliberation: Oratory and Conversation in Ratifying the Constitution.” Journal of Political Philosophy 8:68–­90. Riker, William. 1982. Liberalism against Populism. Prospect Heights, IL: Waveland Press. Riley, Jonathan. 1990. “Utilitarian Ethics and Democratic Governance.” Ethics 1000:335–­48. Ripstein, Arthur. 1987. “Foundationalism in Political Theory.” Philosophy and Public Affairs 16:115–­37. Risse, Mathias. 2004. “Arguing for Majority Rule.” Journal of Political Philosophy 12:41–­64. —­—­—­. 2009a. “On the Philosophy of Group Decision Making I: The Non-­Obviousness of Majority Rule.” Philosophical Compass 4:793–­802. —­—­—­. 2009b. “On the Philosophy of Group Decision Making II: Alternatives to Majority Rule.” Philosophical Compass 4:803–­12. Rodrik, Dani. 2007. One Economics, Many Recipes. Princeton, NJ: Princeton Univer­ sity Press. Roemer, John. 1993. “A Pragmatic Theory of Responsibility for the Egalitarian Planner.” Philosophy and Public Affairs 22 (2):146–­66. —­—­—­. 1995. “Equality and Responsibility.” Boston Review, April/May. —­—­—­. 1998. Equality of Opportunity. Cambridge: Cambridge Univer­sity Press. Rorty, Amelie. 1985. “Varieties of Rationality, Varieties of Emotions.” Social Science Information 24:343–­53.

301

R e fere n c e s Rorty, Richard. 1989. Contingency, Irony and Solidarity. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1991a. Essays on Heidegger and Others: Philosophical Papers II. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1991b. Objectivity, Relativism and Truth: Philosophical Papers I. Cambridge: Cambridge Univer­sity Press. —­—­—.­ 1996. “Idealizations, Foundations and Social Practices.” In Democracy and Difference, ed. Seyla Benhabib. Princeton, NJ: Princeton Univer­sity Press. —­—­—­. 1998. Truth and Progress. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1999. Philosophy and Social Hope. New York: Penguin. Rosenblum, Nancy. 1998. Membership and Morals. Princeton, NJ: Princeton Univer­ sity Press. Rosenblum, Nancy, and Robert Post. 2002. Civil Society and Government. Princeton, NJ: Princeton Univer­sity Press. Rosenthal, Howard. 1990. “The Setter Model.” In Advances in the Spatial Theory of Voting, ed. James Enelow and Melvin Hinich. Cambridge: Cambridge Univer­sity Press. Rubinstein, Ariel. 1991. Comments on the Interpretation of Game Theory.” Econometrica 59:909–­24. —­—­—­. 2000. Economics and Language. Cambridge: Cambridge Univer­sity Press. Ryan, Alan. 1995. John Dewey and the High Tide of American Liberalism. New York: W. W. Norton. Ryfe, David. 2005. “Does Democratic Deliberation Work?” Annual Review of Political Science 8:49–­71. Sabel, Charles. 1992. “Studied Trust: Building New Forms of Cooperation in a Volatile Economy.” In Industrial Districts and Local Economic Regeneration, ed. F. Pyke and W. Sengenberger. Geneva: International Institute for Labour Studies. Sally, David. 1995. “Conversation and Cooperation in Social Dilemmas.” Rationality and Society 7:58–­91. Sandel, Michael. 1996. “Dewey Rides Again.” New York Review of Books, June 1, 35–­38. Sanders, Lynn. 1997. “Against Deliberation.” Political Theory 25:347–­76. —­—­—.­ 1999. “Democratic Politics and Survey Research.” Philosophy of the Social Sciences 29:248–­80. Satterthwaite, Mark. 1975. “Strategy-­Proofness and Arrow’s Conditions.” Journal of Economic Theory 10:187–­218. Satz, Debra. 1992. “Markets in Women’s Reproductive Labor.” Philosophy and Public Affairs 21:107–­31. —­—­—.­ 1995. “Markets in Women’s Sexual Labor.” Ethics 106:63–­85. —­—­—­. 2010. Why Some Things Should Not Be for Sale. Oxford: Oxford Univer­sity Press. Saward, Michael. 2000. “Less than Meets the Eye: Democratic Legitimacy and Deliberative Theory.” In Democratic Innovation, ed. Michael Saward. London: Routledge.

302

R e fere n c e s Schelling, Thomas. 1960. The Strategy of Conflict. Cambridge, MA: Harvard Univer­ sity Press. —­—­—­. 1978. Micromotives and Macrobehavior. New York: W. W. Norton. Schneider, Gerald. 2008. “De Deliberatibus Disputandum est: A Response to Jürg Steiner,” European Political Science 7:199–­206. Schofield, Norman. 1978. “Instability of Simple Dynamic Games.” Review of Economic Studies 45:575–­94. —­—­—­ . 1985. “Anarchy, Altruism and Cooperation.” Social Choice and Welfare 2:207–­219. Schotter, Andrew. 1981. The Economic Theory of Social Institutions. Cambridge, MA: Cambridge Univer­sity Press. Sen, Amartya. 1966. “A Possibility Theorem on Majority Decisions.” Econometrica 34:491–­99. —­—­—­. 1987. “Social Choice.” In The New Palgrave: A Dictionary of Economics, ed. John Eatwell, Murray Milgate, and Peter Newman. Hampshire, England: Palgrave Macmillan. —­—­—­. 1992. Inequality Reexamined. Cambridge, MA: Harvard Univer­sity Press. —­—­—­. 1995. “Rationality and Social Choice.” American Economic Review 85:1–­24. —­—­—­. 1999. Development as Freedom. New York: Knopf. —­—­—­. 2003. Rationality and Freedom. Cambridge, MA: Harvard Univer­sity Press. —­—­—­. 2009. The Idea of Justice. Cambridge, MA: Harvard Univer­sity Press. Shapiro, Ian. 2002a. “Contemporary Democratic Theory.” In Political Science: The State of the Discipline, Centenial Edition, ed. Ira Katznelson and Helen Milner. New York: W. W. Norton. —­—­—­. 2002b. “Optimal Deliberation,” Journal of Political Philosophy 10:196–­211. —­—­—­. 2003. The State of Democratic Theory. Princeton, NJ: Princeton Univer­sity Press. Shepsle, Kenneth. 1979. “Institutional Arrangements and Equilibrium in Multidimensional Voting Models.” American Journal of Political Science 23:27–­60. —­—­—­. 1989. “Studying Institutions.” Journal of Theoretical Politics 1:131–­49. —­—­—­. 2003. “Losers in Politics (and How They Sometimes Become Winners): William Riker’s Heresthetic.” Perspectives on Politics 1: 307–­15. Shiffman, Gary. 2002. “Construing Disagreement—­Consensus and Invective in ‘Constitutional’ Debate.” Political Theory 30:175–­203. Simmons, A. John. 2010. “Ideal and Non-­Ideal Theory.” Philosophy and Public Affairs 38:5–­36. Singer, Peter. 1972. “Altruism and Commerce.” Philosophy and Public Affairs 2:312–­20. Smiley, Marion. 1999. “Democratic Citizenship: A Question of Competence?” In Citizen Competence and Democratic Institutions, ed. S. Elkin and K. Soltan. Univer­sity Park, PA: Penn State Univer­sity Press. Spitz, Elaine. 1984. Majority Rule. Chatham, NJ: Chatham House.

303

R e fere n c e s Squires, Judith. 2002. “Deliberation and Decision-­making.” In Democracy as Public Deliberation, ed. Maurizio Passerin D’Entreves. Manchester, England: Manchester Univer­sity Press. Steiner, Jürg. 2008. “Concept Stretching: The Case of Deliberation.” European Political Science 7:186–­90. Stiglitz, Joseph. 1998. “Central Banking in a Democratic Society.” De Economist 146:199–­226. —­—­—.­ 2002. “Information and the Change of Paradigm in Economics,” American Economic Review 92:460–­501. Stiglitz, Joseph, Amartya Sen, and Jean-­Paul Fitoussi. 2010. Mis-­Measuring Our Lives: Why GDP Doesn’t Add Up. New York: New Press. Stinchcombe, Arthur. 2001. When Formality Works. Chicago: Univer­sity of Chicago Press. Stone-­Sweet, Alec. 2000. Governing with Judges: Constitutional Politics in Europe. Oxford: Oxford Univer­sity Press. Sugden, Robert. 1986. The Economics of Rights, Cooperation and Welfare. London: Basil Blackwell. Sunstein, Cass. 1992. “Neutrality in Constitutional Law (with Special Reference to Pornography, Abortion and Surrogacy).” Columbia Law Review 92:1–­52. ­ ­. 1993. ‘Democracy and Shifting Preferences.” In The Idea of Democracy, ed. David —­—— Copp, Jean Hampton, and John Roemer. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1995. Democracy and the Problem of Free Speech. New York: Free Press. —­—­—.­ 2000. “Deliberative Trouble? Why Groups Go to Extremes.” Yale Law Journal 110:71–­119. —­—­—­. 2001. Republic.com. Princeton, NJ: Princeton Univer­sity Press. —­—­—.­ 2002a. “The Law of Group Polarization.” Journal of Political Philosophy 10:175–­95. —­—­—.­ 2002b. “On a Danger of Deliberation,” Dædalus 131:120–­24. Sutton, John. 1986. “Noncooperative Bargaining Theory: An Introduction.” Review of Economic Studies 53 (5): 709–­24. Swedberg, Richard. 1994. “Markets as Social Structures.” In The Handbook of Economic Sociology, ed. N. Smelser and R. Swedberg. Princeton, NJ: Russell Sage Foundation/Princeton Univer­sity Press. “Symposium: Economics of Voting.” 1995. Journal of Economic Perspectives 9:3–­98. Talisse, Robert. 2007. A Pragmatist Philosophy of Democracy. London: Routledge. —­—­—.­ 2011. “A Farewell to Deweyian Democracy.” Political Studies (59:2–18). Taylor, Michael. 1982. Community, Anarchy and Liberty. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1987. The Possibility of Cooperation. Cambridge: Cambridge Univer­sity Press. Thompson, Dennis. 2008. “Deliberative Democratic Theory and Empirical Political Science.” Annual Review of Political Science 11: 497–­520. Titmus, Richard. 1971/1997. The Gift Relationship. New York: New Press. Toobin, Jeffrey. 2003. “The Great Election Grab: When does gerrymandering become a threat to democracy?” New Yorker, December 8.

304

R e fere n c e s —­—­—­. 2006. “Drawing the Line: Will Tom DeLay’s redistricting in Texas cost him his seat?” New Yorker, March 6. Tyler, Tom. 1990. Why People Obey the Law. New Haven, CT: Yale Univer­sity Press. —­—­—­. 1997. Social Justice in a Diverse Society. Boulder, CO: Westview Press. Ullman-­Margalit, Edna. 1978. The Emergence of Norms. Oxford: Oxford Univer­sity Press. Unger, Roberto Mangabeira. 2005. What Should the Left Propose? London: Verso. —­—­—­. 2007. Free Trade Reimagined. Princeton, NJ: Princeton Univer­sity Press. —­—­—­. 2009. The Self Awakened: Pragmatism Unbound. Cambridge, MA: Harvard Univer­sity Press. Van Mill, David. 1996. “The Possibility of Rational Outcomes from Democratic Discourse and Procedures.” Journal of Politics 58:734–­52. Vanberg, Georg. 1998. “Abstract Judicial Review, Legislative Bargaining and Policy Compromise.” Journal of Theoretical Politics 10:299–­26. Vermeule, Adrian. 2007. Mechanisms of Democracy. Oxford: Oxford Univer­sity Press. Waldron, Jeremy. 1993. Liberal Rights. Cambridge: Cambridge Univer­sity Press. —­—­—­. 1998. “Participation: The Right of Rights.” Proceedings of the Aristotelian Society, n.s., 98:307–­38. —­—­—­. 1999. Law and Disagreement. Oxford: Oxford Univer­sity Press. Walzer, Michael. 1982. Spheres of Justice. New York: Basic Books. —­—­—­. 2004. Politics and Passion. New Haven, CT: Yale Univer­sity Press. Warren, Mark. 1992. “Democratic Theory and Self-­Transformation.” American Political Science Review 86:8–­23. —­—­—­. 1996. “What Should We Expect from More Democracy?” Political Theory 24:241–­70. Weale, Albert. 1999. Democracy. New York: St. Martin’s Press. Weber, Max. 1978. Economy and Society. Berkeley: Univer­sity of California Press. Weingast, Barry. 1984. “The Congressional-­Bureaucratic System: A Principal Agent Perspective (with applications to the SEC).” Public Choice 44 (1): 147–­91. —­—­—­. 1995. “The Economic Role of Political Institutions: Market-­Preserving Federalism and Economic Development.” Journal of Law, Economics and Organization 11 (1): 1–­31. Weingast, Barry, and Mark J. Moran. 1983. “Bureaucratic Discretion or Congressional Control? Regulatory Policymaking by the Federal Trade.” Journal of Political Economy 91 (5): 765–­800. West, Cornel. 1989. The American Evasion of Philosophy. Madison: Univer­sity of Wisconsin Press. Westbrook, Robert. 1991. John Dewey and American Democracy. Ithaca, NY: Cornell Univer­sity Press. —­—­—­. 1993. “Democratic Evasions.” Praxis International 13:1–­13. —­—­—­. 1998. “Pragmatism and Democracy: Reconstructing the Logic of John Dewey’s Faith.” In The Revival of Pragmatism, ed. Morris Dickstein. Durham, NC: Duke Univer­sity Press.

305

R e fere n c e s Westbrook, Robert. 2005. Democratic Hope. Ithaca, NY: Cornell Univer­sity Press. Williams, Melissa. 2000. “The Uneasy Relation of Group Representation and Deliberative Democracy.” In Citizenship in Diverse Societies, ed. Will Kymlicka and Wayne Norman. Oxford: Oxford Univer­sity Press. Williamson, Oliver. 1985. The Economic Institutions of Capitalism. New York: Free Press. Wolff, Edward. 2001. Top Heavy: The Increasing Inequality of Wealth in America and What Can Be Done about It. New York: New Press. Wolin, Sheldon. 2004. Politics and Vision, expanded ed. Princeton, NJ: Princeton Univer­sity Press. Young, Iris Marion. 1990. Justice and Politics of Difference. Princeton, NJ: Princeton Univer­sity Press. —­—­—­. 2000. Inclusion and Democracy. Oxford: Oxford Univer­sity Press. Young, Oran. 1999. Governance in World Affairs. Ithaca, NY: Cornell Univer­sity Press. Young, Peyton. 1988. “Condorcet’s Theory of Voting.” American Political Science Review 82:1221–­44. —­—­—­. 2004. Equity. Princeton, NJ: Princeton Univer­sity Press.

306

Index

adjudication, 5, 25, 36–37, 42, 164, 171 aggregation, 11; argument and, 128, 143, 146–49, 153–60; decentralization and, 80, 82; justification and, 95, 101–4, 107–18, 123; liberal guarantees and, 211, 216–17, 220; pragmatism and, 53, 241 Alchian, Armed A., 66 American Pragmatism, 25n2, 27n14 Anderson, Elizabeth, 151n64 anonymity, 113, 116n74, 157n84, 211–12, 217 antidemocratic forces, 156, 281n29, 283–86 anti-skepticism, 21; consquentialism and, 27–28, 283–84; fallibilism and, 27–28, 283–84; holism and, 27–28; justification and, 96; pragmatism and, 27–28, 31–32, 39, 283–84 Arendt, Hannah, 39 argument, 11; aggregation and, 128, 143, 146–49, 153–60; agreement and, 128–29, 134–37, 142–46, 150–51, 154, 157, 161–65; alternative possibilities and, 136–42; bargaining and, 119–21; better decisions and, 151–61; cheap talk and, 130–31, 220n66; civilizing force of hypocrisy and, 138–40, 145; cognitive effects on dimensionality and, 144–51; communication and, 129–36; community and, 156n79; concept of, 119; Condorcet jury theorem and, 112–13, 116, 153–58, 220; conflict and disagreement and, 128, 130nn6,7, 137, 144, 149–50, 158n86, 161–63; consensus and, 20, 29–32, 42, 50, 94, 96n5, 128, 135,

137, 144, 150–51, 157, 163–64, 167, 185, 214, 247, 255n90, 256, 260, 263, 265n4; conversation and, 132; decentralization and, 53, 80, 82; decision making and, 142–44, 150–58, 161–66; deliberative reason and, 134n18; democratic deliberation and, 95, 103, 118–29, 133–45, 150n60, 151n63, 152, 209–13, 216–18, 221, 237–47; Dewey and, 129–30, 147, 156; discursive democracy and, 133–34; discussion, debate and persuasion and, 114n69, 123, 127, 136; diversity and, 129, 151, 158–63; economic issues and, 154–59; effective participation and, 225–28, 231, 234, 237, 240–42, 243n59, 247n74, 248–49, 253; endogenous mechanisms and, 141–42; equilibrium and, 130n6, 141n36, 147–48; ethics and, 134n18; excluded groups and, 120; force of better, 133; freedom and, 142–43, 160, 164; game theory and, 129–30, 132n12, 134, 146, 165n95; group polarization and, 124– 25; Habermas on, 131–35; identifying mechanisms and, 129–31; information and, 128, 130nn6,7, 136–37, 141–42, 145–46, 151–52, 160, 165n95; institutional arrangements and, 151, 161–66, 169–71, 180–82, 190; institutional performance and, 160, 165–66; justification and, 94–97, 101–9, 114–27, 139–40; language and, 129–34; legal issues and, 134–35, 159, 161–63; liberal guarantees and, 211, 216–17, 220; majority rule and, 115–17; Nash equilibrium and, 129; normative

In d e x argument (cont’d) theory and, 128–35, 138–47, 157, 162, 165–66; opinion and, 133–34, 137nn25,26, 141, 148n54; philosophy and, 132–33, 146n47; pragmatism and, 143, 151–61, 165, 241; preference and, 128, 136–52, 159; realist approach and, 143; reflexivity and, 161–67, 172–75; religion and, 135, 154; robustness and, 148, 164–65; second-order priority and, 12, 19–21, 23, 42, 56, 95, 163–71, 176, 180–83, 188, 194, 252–53, 260, 273; singlepeakedness and, 149; social choice theory and, 143, 146–49, 162, 260; social norms and, 131, 136, 182; strategic action and, 129, 131–35, 138–43, 146; structured disagreement and, 146; supermajority and, 147–48; voting and, 128, 142–58, 162–63, 194; war and, 150n61; will-information and, 133–34 Arrow, Kenneth: decentralization and, 61n34; justification and, 109–10, 111n58, 113, 118; liberal guarantees and, 211; reflexivity and, 168; role of political argument and, 146–47, 148n54, 149n55, 157 Athens, 10–11, 249n77 Austen-Smith, David, 109n50, 220nn65,66 Banks, Jeffrey, 109n50 bargaining: argument and, 119–21; Bolton on, 78; Coasian, 21, 66–80, 176, 180, 189; decentralization and, 66–80; externalities and, 71–75, 78, 80; fairness and, 72–73; Farrell on, 77–79; justification and, 119–21; reflexivity and, 176, 180, 189 Barry, Brian, 184 Basu, Samy, 133n15 Bendor, Jonathan, 178

Benhabib, Seyla, 103n32, 137n25 Berlin, Isaiah, 201 Bernstein, Richard, 33–35, 382n35 Blinder, Alan, 156n80 Bohman, James, 216n51, 237, 240, 240–42, 247n74 Bolton, Patrick, 78 Borda count, 112 Brennan, Goeffrey, 7n14 Brighouse, Harry, 218n55, 235n30 brute induction, 68n57 Buchanan, James, 241n53 burdens of judgment, 214–15 bureaucracy, 175–81 Calvert, Randall, 61–63, 83 cheap talk, 130–31, 220n66 Christiano, Thomas, 94n3, 152n67, 210n28, 213n37, 235n30, 244n64, 246n72, 253n82 Coasian bargaining, 21; Bolton on, 78; decentralization and, 66–80; externalities and, 71–75, 78, 80; fairness and, 72–73; Farrell on, 77–79; reflexivity and, 176, 180, 189 coercion, 3–4; argument and, 134; decentralization and, 51, 87n99; justification of, 270; legitimacy and, 264, 270, 275; liberal guarantees and, 197, 202, 206–8; as limiting power; 254; promises and, 227, 235 Cohen, Joshua: pragmatism and, 37n48, 229n17, 239n44, 241n53, 242, 247; justification and, 103n32, 120nn88,90 Coleman, Jules, 107n41 collective choice, 20, 44, 149 common pool resources, 9–10 communication, 11; argument and, 129–35; decentralization and, 62; justification and, 123, 125n97; liberal guarantees and, 207, 209, 220n66; pragmatism and, 36, 39, 46, 279n23; reflexivity and, 170, 182, 186–89

308

In d e x communicative action, 131–36 community, 1n1, 21–22; boundary establishment and, 40–41; burden of justification and, 104; bureaucracy and, 175–81; community-based cooperation and, 79–83; compliance and, 268; decentralization and, 56n40, 65, 71, 79–88; incentive-compatible rules and, 83–88; legal issues and, 185–86 (see also legal issues); pragmatism and, 37–42, 232, 241, 268, 274, 283; prisoner’s dilemma (PD) and, 62, 80–83, 87, 131n8; reflexivity and, 39, 171, 174, 181, 184–85, 189–90; role of political argument and, 156n79; social norms and, 181–90 (see also social norms); spillover effect and, 185, 188 competition, 3–4; Coasian bargaining and, 21, 66–80, 176, 180, 189; conflict and, 1n1 (see also conflict and disagreement); decentralization and, 53–72, 76, 79, 258; differing environments for, 23; elections and, 95, 102–3, 107–10, 118–19, 143, 153, 177–79, 217n52, 222, 250; hypercompetitive environment and, 12; market, 6, 12n33, 56–57, 60, 69, 72, 168, 254; military, 11; normative features of, 13–14; organization of, 167–71; pragmatism and, 32, 50; reflexivity and, 167–71; regulation of political, 228–29; second-order priority and, 163–66 compliance, 104n37; argument and, 134; community and, 268; pragmatism and, 231–32, 264, 267–74, 277; reflexivity and, 184 Condorcet method, 112–13, 116, 153–58, 220 conflict and disagreement, 1–5, 8, 12, 18, 20; argument and, 128, 130nn6,7, 137, 144, 149–50, 158n86, 161–63; decentralization and, 56, 67, 71, 74,

88–89; distributional, 41, 89, 259; inevitability of, 271–72, 276; justification and, 100–1, 119–22, 126n99; legitimacy and, 270–71; liberal guarantees and, 195n1, 198, 207, 215n47, 219, 220n66, 221; political obligation and, 270–71; pragmatism and, 39–41, 229–33, 244, 248, 253–55, 259–63, 268n8, 270–71, 273–76, 281–82; reconciliation and, 15n41, 16n42, 88, 215n47, 229, 232–33, 253, 258–59; reflexivity and, 169, 186–88 Conover, Pamela Johnston, 119 consensus, 20; community and, 42 (see also community); democratic deliberation and, 247; institutional choice and, 256; justification and, 94, 96n5; normative criteria and, 50, 263; overlapping, 214, 265n4; persistent disagreement and, 260; Posner-Rorty, 29–32; reflexivity and, 167, 185; role of political argument and, 128, 135, 137, 144, 150–51, 157, 163–64; Sen and, 255n90 consequentialism: anti-skepticism and, 27–28, 283–84; fallibilism and, 27–28, 283–84; institutional design and, 27–28, 31–32, 39, 44–47; justification and, 94–98, 104, 196n4; tempered, 18, 21, 44–45, 94, 104, 190, 194, 251, 257, 262, 266–74 Considerations on Representative Government (Mill), 217n52 Constant, Benjamin, 201 constitutional provisions, 8–9, 61, 280; effective participation and, 222, 233, 242–43, 250; liberal guarantees and, 196, 200, 202–6, 210–11, 214; reflexivity and, 173–74, 177; tempered, 262 Cotterrell, Roger, 184 courts, 171–75 Cox, Gary, 111n55

309

In d e x creation science, 215–16 “Creative Democracy--The Task Before Us” (Dewey), 35–36 Dahl, Robert, 84, 210n28 Dasgupta, Partha, 113, 115–17 Davidson, Donald, 138n27 Debreu, Gerard, 168 decentralization, 1, 22, 38, 93; Calvert and, 61–63, 83; Coasian bargaining and, 21, 66–80, 176, 180, 189; communication and, 62; community and, 56n40, 65, 71, 79–88; competition and, 258; conflict and disagreement and, 56, 67, 71, 74, 88–89; decision making and, 71, 89; demands for, 51; distributed knowledge and, 53–55, 260–61; economic issues and, 51–63, 66, 69–79, 80n86, 83n93; equilibrium and, 56–67, 80–83, 87; Hayek and, 52–55, 59n24, 79; incentive-compatible rules and, 83–88; information and, 53–54, 59, 77–79, 85–88; institutional arrangements and, 51–55, 60, 64–65, 69–70, 75–76, 80n86, 83–89; institutional choice and, 51–55, 70, 73–76, 79; institutional performance and, 52, 55, 72; intuitive appeal of, 51–52; justification and, 52–55, 102–3; legal issues and, 74; moral issues and, 61n34, 64n47; normative theory and, 58–61, 68–73, 76, 79, 89; prisoner’s dilemma (PD) and, 62, 80–83, 87, 131n8; privileging markets and, 55–71; robustness and, 68–71, 76, 83, 87, 89, 258–59; self-interest and, 53, 67–68, 75, 81n88; social norms and, 56n10, 62, 64, 82–83, 185; technocracy and, 65, 83–88 decision making: better, 151–61; burden of justification and, 93–95, 101, 104–10, 115–18, 122, 126n98;

310

Coasian bargaining and, 21, 66, 71–80, 176, 180, 189; coercion and, 3–4, 51, 87n99, 134, 197, 202, 206–8, 237, 254, 270, 275; collective choice and, 20, 44, 149; community-based cooperation and, 79–83; Condorcet jury theorem and, 112–13, 116, 153, 153–58, 220; decentralization and, 71, 89; discharge rule and, 9; distributedknowledge problem and, 11, 53, 55, 144, 151, 159, 161, 167, 180, 260–61 (see also information); effectiveness and, 261–62; equal opportunity for political influence and, 20, 197–98, 210, 217n53, 225–28, 232–38, 247, 255, 262–63; freedom and, 224 (see also freedom); House Rule XI and, 9; institutional pluralism and, 5–6, 9; Journal Clause and, 9; justification and, 115– 16 (see also justification); liberal guarantees and, 194–211, 212n33, 215n47, 218–19, 220n65; liberty and, 140n34, 199, 201–2, 217n52, 255; majority rule and, 9, 101, 112–17, 147–48, 241n52, 248n76; pragmatism and, 32, 43–45, 48, 51–54, 225–27, 230–32, 235, 239, 241, 246, 249, 252, 255n90, 259–63, 266–69, 273–77; prisoner’s dilemma (PD) and, 62, 80–83, 87, 131n8 (see also game theory); property rights and, 9–10, 60–61, 64n47, 65, 72–75, 77n78, 78, 80n86; reflexivity and, 167–76, 179–85, 188–90; role of political argument and, 142, 144, 150–58, 161–66; Rule of Four and, 9; second-order priority and, 12, 19–21, 23, 42, 56, 95, 163–66, 163–71, 176, 180–83, 188, 194, 252–53, 260, 273; Senate Rule XXII and, 9; Seven Member Rule and, 9; social choice and, 18, 22, 82, 95, 102n31, 106–11, 115, 118, 123, 143, 146–49, 162, 167, 211, 216,

In d e x 220, 241n53, 260, 273; social norms and, 181–85, 188–90; voting and, 115–17 (see also voting) defunding, 8 deliberative reason, 134n18 democracy: acceptance of as terms of politics, 275–78; actually existing, 100–1; aggregative-deliberative dichotomy of, 118–19; alternative institutional forms and, 19; antidemocratic forces and, 156, 281n29, 283–86; assessing, 93; associational, 15n41, 103–4; as close to nonnegotiable, 93; competition and, 1 (see also competition); conditions of effective, 222–55; consensus and, 20, 29–32, 42, 50, 94, 96n5, 128, 135, 137, 144, 150–51, 157, 163–64, 167, 185, 214, 247, 255n90, 256, 260, 263, 265n4; decentralization and, 51–89; decision making and, 4–5, 9, 19–23, 32, 52, 89 (see also decision making); defining, 4, 18–19; deliberative concept of, 118; Dewey on, ix-xi, 4, 97–98 (see also Dewey, John); discursive, 133–34; epistemic, 37, 43–44, 152–58; equality and, 20, 23, 37, 45, 47, 54n7, 69, 118–23, 139, 143, 160, 188, 195–99, 205–11, 217n53, 218, 221–29, 233– 55, 259–66, 281, 285n42; institutional embodiment and, 33–36; interference and, 34, 196–98, 201–8, 217, 222–27, 230–33, 249–50, 254, 278; justification of, 52–55, 93–127; legitimacy and, 4–5, 263–75 (see also legitimacy); liberal guarantees and, 193–201; liberal theory of, 108, 112n61, 140n33, 143n42, 193–222, 235, 249, 284; liberty and, 140n34, 199, 201–2, 217n52, 255; minimalist view of, 101n24, 108, 143, 209n21, 215–16; nondictatorship and, 107n43, 110, 113n67, 146–47,

211; pragmatism and, ix-x, 20–22 (see also pragmatism); priority of, 12, 18–19, 22–24, 51–55, 95, 106, 128, 164–67, 169, 193–95, 222, 248–49, 256, 260–61, 267, 270, 273–74, 279, 281n30, 285; reflexivity and, 183–90 (see also reflexivity); representative government and, 102, 156n79, 217n52, 218; second-best alternatives and, 94–95; second-order priority and, 12, 19–21, 23, 42, 56, 95, 163–71, 176, 180–83, 188, 194, 252–53, 260, 273; as social idea, 33–34, 46, 99–100; state power and, 248–51; unrestricted domain and, 107n43, 110, 113n67, 114, 146, 148–49, 211–12, 216n51; utopian, 15, 16n42, 24, 37, 143n42, 256, 278–86 Democracy and Knowledge (Ober), 10 democratic deliberation, 95, 103; better decisions and, 151–61; civilizing force of hypocrisy and, 138–40, 145; cognitive effects on dimensionality and, 144–51; justification and, 118–27; pragmatism and, 209–13, 216–18, 221, 237–47; reflexivity and, 128 (see also reflexivity); role of political argument and, 128–29, 133–45, 150n60, 151n63, 152; standard case for, 128–29, 133–45 Dewey, John, ix-xi, 5; burden of justification and, 96–102, 104, 108, 115; discussion, debate and persuasion and, 123, 127, 136; intelligent response and, 284; pragmatism and, 25–42, 45–46, 49, 266, 279–85; representative government and, 102; role of political argument and, 129–30, 147, 156; rule of law and, 4; violence and, 3 disagreement. See conflict and disagreement discharge rule, 9

311

In d e x discounting, 45, 81n89, 82 diversity: benefits of, 260–61; common pool resources and, 9–10; community-based cooperation and, 79–83; conflict and, 1–5, 8, 12, 15, 18, 20 (see also conflict and disagreement); constitutional politics and, 8–9; decision making and, 9 (see also decision making); economic exchange and, 7–8; franchise distribution and, 8; institutions and, 1, 7, 8n15, 10, 18, 23, 36–37, 40, 42–45, 49, 257–58; irreducible, 129, 151, 163, 256, 261; liberal guarantees and, 194, 198, 206, 208; pragmatism and, 36–37, 40, 42–45, 49, 230, 234, 251, 256, 260–67, 270–72, 282; property rights and, 9–10; reflexivity and, 167, 180; role of political argument and, 129, 151, 158–63; as social fact, 18 Dixit, Avinash, 165n95 Dryzek, John, 102n31, 122n90, 133n16, 137n25 Dummett, Michael, 111n56, 112n63 Dworkin, Ronald, 120n85, 196n4 economic issues, ix, 103, 260; Coasian bargaining and, 21, 66–80, 176, 180, 189; decentralization and, 51–89; defunding and, 8; discounting and, 45, 81n89, 82; free riders and, 80n86; game theory and, 2n2, 18, 61–63, 64n47, 77n78, 80–83, 86–87, 110, 111n55, 129–30, 132n12, 134, 146, 165n95, 178, 219, 220n66, 281n28; institutional design and, 28, 31, 36, 38, 42–43, 44n64; institutional pluralism and, 5–8; justification and, 98, 104–6, 118, 120n85, 121–23; liberal guarantees and, 193, 199, 203, 207, 217; macroeconomic analysis and, 80n86, 154–58; market competition and, 6, 11, 12n33, 52–57, 60,

69, 72, 168, 254; microeconomic analysis and, 56–61, 69, 118; Pareto efficiency and, 56, 59n27, 65, 74, 110, 113, 116n74, 146–47; politics and, 12, 19–21; pragmatism and, 28, 31, 36, 38, 42–43, 44n64, 223–29, 236, 243–48, 254, 255n90, 278–83; privileging markets and, 55–71; property rights and, 9–10, 60–61, 64n47, 65, 72–75, 77n78, 78, 80n86; reflexivity and, 168, 170n5, 177, 182; role of political argument and, 154–59; secret ballot and, 7–8, 217; transaction costs and, 12n33, 59, 66, 75–78 effectiveness, 95; acceptance of democracy as terms of politics, 275–78; compliance and, 231–32, 264, 267–74, 277; concepts of effective participation and, 223–26; decision making and, 261–62; democratic, 222–23, 252–55; equality and, 226–34, 262–63; freedom and, 222–44, 248–55, 262–63; goal of, 224–25; group-maximization problem and, 263; Hayek and, 253– 55; horizontal relationships and, 226– 29; individual capacities and, 234–48; inevitability of politics and, 252–55; institutional performance and, 6–7, 19, 21, 52, 55, 72, 96 (see also institutional performance); interventions and, 248–51; polycentric systems and, 251; power distribution and, 225–27, 230–31, 235–38, 245–55; pragmatism and, 222–55; robustness and, 252–53; social factors and, 226–34; state power and, 248–51 Eggertsson, Thrain, 65–66 elections: effective participation and, 222, 250; justification and, 95, 102–3, 107–10, 118–19; liberal guarantees and, 217n52; role of political argument and, 143, 153; reflexivity and, 177–79

312

In d e x Elster, Jon, 120n84, 138n28, 140n32, 169n3, 277n20 Enelow, James, 111n55 epistemic democrats, 37, 43–44, 152–58 Epstein, Lee, 180n25 equality, 23, 281, 285n42; argument and, 139, 143, 160; decentralization and, 54n7, 69; effective freedom and, 226–48; effective participation and, 222–55, 262–63; horizontal relationships and, 226–29; importance of, 208–10; individual capacities and, 234–48; institutional design and, 37, 45, 47; institutionalizing effective, 226–48; institutionalizing formal, 211–21; justification and, 118–23; legally binding, 208–9; liberal guarantees and, 195–99, 198, 205–11, 217n53, 218, 221; minimizing effects and, 198; neutrality and, 198, 211–12; opportunity for political influence and, 20, 197–98, 210, 217n53, 225–28, 232–38, 247, 255, 262–63; Rawls on, 235–37; reflexivity and, 188; Sen on, 237–39; social factors and, 226–34; unrestricted domain and, 107n43, 110, 113n67, 114, 146, 148–49, 211–12, 216n51 equilibrium: Calvert and, 61–63, 83; Coasian bargaining and, 21, 66–80, 176, 180, 189; decentralization and, 56–67, 80–83, 87; game theory and, 2n2, 18, 61–63, 64n47, 77n78, 80–83, 86–87, 110, 111n55, 129–30, 132n12, 134, 146, 165n95, 178, 219, 220n66, 281n28; institutionalizing formal, 211–21; liberal guarantees and, 219; Nash, 82, 87, 129, 130n6; new contracts and, 65; reflexivity and, 168; role of political argument and, 130n6, 141n36, 147–48; structure-induced, 147–48 Estlund, David, 37n48

ethics, 1; hypocrisy and, 138–40, 145; justification and, 94n4, 121n89; nondictatorship and, 107n43, 110, 113n67, 146–47, 211; pragmatism and, 37, 265, 288; role of political argument and, 134n18; vote-counting schemes and, 106–9 ethnic issues, 3, 146, 181, 202, 230, 232, 282 exit strategy, 3, 162, 282–83 experimentalism, 7, 144; burden of justification and, 98; institutional choice and, 6, 48, 50; institutional design and, 45–49; liberal guarantees and, 193–94; pragmatism and, 28, 37, 43–50, 250, 257–60, 263, 280–81, 283; reflexivity and, 169, 174, 182–88; role of political argument and, 151, 159–60 fallibilism, 21, 257; anti-skepticism and, 27–28, 283–84; argument and, 159; consequentialism and, 27–28, 283–84: indeterminacy and, 2, 13, 26–27, 43, 82, 107, 116, 137, 146–47; institutional design and, 26–28, 31–32, 39–40, 45; justification and, 96 Farrell, Joseph, 72n64, 77–79, 130n6, 220n66 Ferejohn, John, 107n41 Fishkin, James S., 141n37 “Fixation of Belief, The” (Peirce), 39–40 Fraser, Nancy, 212n33, 247n74 freedom: argument and, 142–43, 160, 164; capacity and, 234–48; of choice, 47, 55; cognitive abilities and, 238–39; cultural resources and, 238; dimensions of, 199–205; effective participation and, 222–55, 262–63; of empowerment, 47; equal opportunity for political influence and, 20, 197–98, 210, 217n53, 225–28, 232–38, 247, 255, 262–63 (see also equality);

313

In d e x freedom (cont’d) experimentalism and, 37; horizontal relationships and, 207–8, 210, 225–30, 233, 249; individual, 55, 69, 234–48; initial conditions of, 45, 123; institutionalizing effective, 226–48; institutionalizing formal, 205–8; interference of, 34, 196–98, 201–8, 217, 222–27, 230–33, 249–50, 254, 278; justice and, 236; justification and, 118–19, 122–23; liberal guarantees and, 195–208, 218; liberty and, 140n34, 199, 201–2, 217n52, 255; as nondomination, 202–4; as noninterference, 201–4, 254; as nonlimitation, 204–5, 231; pragmatism and, 37, 45; preference and, 224, 238–41, 246, 260; Sen on, 223–24, 239–40; social factors and, 226–34; vertical relationship and, 207–8 free riders, 80n86 Fuller, Lon, 184, 186 Fung, Archon, 104n36 game theory, 2n2, 18, 281n28; argument and, 129–30, 132n12, 134, 146, 165n95; cheap talk and, 130–31, 220n66; Coasian bargaining and, 21, 66–80, 176, 180, 189; decentralization and, 61–63, 64n47, 77n78, 80–83, 86–87; incentive-compatible rules and, 83–88; justification and, 110, 111n55; liberal guarantees and, 219, 220n66; Nash equilibrium and, 82, 87, 129, 130n6; prisoner’s dilemma (PD) and, 62, 80–83, 87, 131n8; reflexivity and, 178 Geuss, Raymond, 1n1 Gibbard, Alan, 107n43 Gilbert, Margaret, 272 Goodin, Robert, 116n74 group-maximization problem, 263

Guinier, Lani, 123n92 Habermas, Jürgen, 39, 131–35, 196n4, 213n40, 218 Hacking, Ian, 29n20 Hammond, Thomas, 86n98, 87n99 Hardin, Russell, 101n24 Hart, H.L.A., 171–73 Hayek, F. A., 52–55, 59n24, 79, 176, 180, 253–55 Heath, Joseph, 132n13 Heidegger, Martin, 29–31 Hirschman, Albert, 120n84, 139n29 holism, 27–28 Holmes, Oliver Wendell, 185 Holmes, Stephen, 117n75, 142n40, 149n59, 156n81 horizontal relationships: effective participation and, 225–30, 233, 249; freedom and, 226–29; liberal guarantees and, 207–8, 210; self-censorship and, 226 House Rule XI, 9 Hume, David, 274n15 hybrids, 181–83 hypocrisy, 138–40, 145 ideal theory, 15–17, 94n4, 275n16 impeachment, 8 indeterminacy, 2, 13, 26–27, 43, 82, 107, 116, 137, 146–47 individual capacities: adequacy of, 235; asymmetries in, 234–38, 241–47; causal effect and, 235; cognitive abilities and, 238–39; cultural resources and, 238; effective participation and, 234–48; institutionalizing freedom/ equality and, 234–48; lack of, 234–35, 238, 241–42, 246; measurement problem of, 239–43; mechanisms for enhancing, 243–48; Rawls on, 235–37, 242; relevance of, 234–39; resource

314

In d e x distribution and, 247–48; Sen on, 239–44; transparent cases of, 242 information, 6; burden of justification and, 112–13, 116–19, 120n88, 125n97; decentralization and, 53–54, 59, 77–79, 85–88; distributedknowledge problem and, 11, 53, 55, 144, 151, 159, 161, 167, 180, 260–61; Habermas on, 133–34;liberal guarantees and, 194; market exchange and, 59; pragmatism and, 234, 238–39, 242, 282; reflexivity and, 167–70, 177, 180–81; role of political argument and, 128, 130nn6,7, 136–37, 141–42, 145–46, 151–52, 160, 165n95; Seven Member Rule and, 9; social norms and, 181–90 institutional arrangements, 3, 12, 21, 23; decentralization and, 51–55, 60, 64–65, 69–70, 75–76, 80n86, 83–89; determining best, 5–6; Dewey and, 4; effective, 19; franchise distribution and, 8; identification of, 7; indeterminate, 2; justification and, 20, 94n3, 95, 98–111, 119; lack of neutral, 1; liberal guarantees and, 196, 221; political theory and, 13–14; pragmatism and, 33, 35, 41–43, 227, 229, 244, 252–53, 256–63, 268n8, 274, 278, 281; Rawls and, 16–17; role of political argument and, 151, 161–66, 169–71, 180–82, 190; various combinations and, 5 institutional choice: basis of, 47; Coasian bargaining and, 73–76; constraints and, 73; decentralization and, 51–55, 70, 73–76, 79; diversity and, 49–50; effective, 20; experimentalism and, 6, 48, 50; justification and, 111n58; liberal guarantees and, 193–94, 198; pragmatism and, 46–50, 252–58, 261–63, 270–73, 278; property rights and, 74; reflexivity and, 168–69,

173–74, 179–82, 189; problem of, 18, 46, 54, 73; social norms and, 182–83 institutional design: diversity and, 42–45; economic issues and, 28, 31, 36, 38, 42–43, 44n64; embodiment and, 33–36; experimentalism and, 45–49; legal issues and, 34, 42; philosophy and, 25–33; pragmatism and, 25–50 institutional effectiveness, 51; liberal guarantees and, 193–96; pragmatism and, 226, 234–38, 257, 261, 267–70, 276; reflexivity and, 174, 183. See also effectiveness institutional performance: argument and, 160, 165–66; assessing, 16, 72, 173, 194; causal claims on, 93; changing social conditions and, 166; conditions necessary for effective, 6–7, 19, 21, 52, 55, 72, 96, 160, 169–73, 188, 194, 197, 200–1, 207–10, 222–24, 227, 230–34, 243, 249, 254–55, 258, 267–70, 273–4; decentralization and, 52, 55, 72; justification and, 93, 96; liberal guarantees and, 194–97, 200–1, 206–10; pragmatism and, 222–24, 227, 230–34, 243, 249, 254–55, 258, 267–70, 273–74; reflexivity and, 169–75, 188; undermining of, 230–31 institutions: bureaucracy and, 175–81; Coasian bargaining and, 21, 66–80, 176, 180, 189; common pool resources and, 9–10; competition and, 1 (see also competition); conception of formal, 184; constitutional politics and, 8–9; decentralization and, 51–89; decision making and, 9 (see also decision making); defined, 2; democratic commitments and, 33–36; diversity and, 1, 7, 8n15, 10, 18, 23, 36–37, 40, 42–45, 49; economic exchange and, 7–8 (see also economic issues); effective freedom and, 226–48;

315

In d e x institutions (cont’d) embodiment and, 33–36; emergence of, 62–71, 95; experimentalism and, 6–7, 37, 41, 45–50, 144, 169, 174n16, 182–88, 257–60, 280–81; formal equality and, 211–21; formal freedom and, 205–8; franchise distribution and, 8; individual capacities and, 234–48; judicial, 171–75; justification and, 93–111, 119n82, 121n89, 122n90; liberal guarantees and, 193–201; pluralism and, 5–7, 10, 45, 229; pragmatism and, 25–50 (see also pragmatism); property rights and, 9–10, 60–61, 64n47, 65, 72–75, 77n78, 78, 80n86; rational-choice models and, 17–18, 59n28, 63, 68, 81n88, 99, 130–31; reflexivity and, 167–70, 177, 180–81; rules and, 2, 7, 46, 56, 64–65, 68, 147, 182, 229, 267–71; self-enforcement and, 2; social factors and, 36–42, 226–34; state-centered context and, 103; systemic quality of, 2 interference: effective participation and, 222–27, 230–33, 249–50, 254; freedom from, 34, 196–98, 201–8, 217, 222–27, 230–33, 249–50, 254, 278; liberal guarantees and, 196–98, 201–8, 217 Johnson, James: argument and, 132nn11,12, 133n14, 151n64; Habermas and, 133n14; institutional design and, 29n17; justification and, 99n18, 105n38, 127n101; liberal guarantees and, 220n66; universal suffrage and, 105n38 Johnson, Paul, 109n50, 111n55, 147n50, 148n52, 216n51 Journal Clause, 9 Jowell, Roger, 141n37 judicial review, 8–9, 175n17

justification, 22–23, 259–60; acceptance of democracy as terms of politics, 275–78; aggregate values and, 106–15; ambiguity and, 111–15; anti-skepticism and, 96; argument and, 94–97, 101–9, 114–27, 139–40; bureaucracy and, 178; Coasian bargaining and, 72–73; coercion and, 270; communication and, 123, 125n97; community and, 104; Condorcet method and, 112–13, 116; conflict and disagreement and, 100–1, 119–22, 126n99; consequentialism and, 94–98, 104; decentralization and, 102–3; decision making and, 93–95, 101, 104–10, 115–18, 122, 126n98; Dewey and, 96–102, 104, 108, 115; discussion, debate and persuasion and, 123, 127, 136; economic issues and, 98, 104–6, 118, 120n85, 121–23; ethics and, 94n4, 121n89; excluded groups and, 120; fallibilism and, 96; freedom and, 118–19, 122–23; heavy burden of, 104–6; heterogeneous constituency and, 105–6, 113; information and, 112–13, 116–19, 120n88, 125n97; instability and, 109–11; institutional arrangements and, 20, 94n3, 95, 98–111, 119; institutional choice and, 111n58; institutional performance and, 93, 96; legal issues and, 103; legitimacy and, 94, 108n47, 111; liberal guarantees and, 193–201, 204–8, 212n33, 213–19; majority rule and, 101, 112–17; May’s theorem and, 116; moral issues and, 94, 96, 100, 105, 108n47, 109, 120n86, 121n89; nondictatorship and, 107n43, 110, 113n67; normative theory and, 93–97, 102–15, 118, 121–27; philosophy and, 96–100, 109; political obligation and, 263–75; political theory and,

316

In d e x 98–99; populism and, 108, 112n61; pragmatism and, 93–100, 104–5, 115–17, 263–75; principal-agent model and, 178; procedural, 94; realist approach and, 95, 105–8; religion and, 99, 103n35; robustness and, 95, 114; second-best alternatives and, 94; self-interest and, 120; social norms and, 119n82; state power and, 248–51; technology and, 98, 125n97; violence and, 101, 124; voting and, 95–96, 101–19, 123, 127 Kant, Immanuel, 285n42 Keech, William, 154n73, 155nn74,76,77, 158n86 Kitcher, Phillip, 160n89 Knight, Jack: argument and, 151n64; brute induction and, 68n57; decentralization and, 67, 80n86; free riders and, 80n86; institutional design and, 29n17; judicial review and, 9n18; justification and, 99n18, 127n101; reflexivity and, 181n27; resource asymmetries and, 67 Kreps, David, 57, 58nn21,22,23, 59n27, 120n85 Landa, Dimitri, 220n65 Laudan, Larry, 216n50 legal issues, ix, 264; adjudication and, 5, 25, 36–37, 42, 164, 171; argument and, 134–35, 161–63; constitutional provisions and, 8–9, 61, 173–74, 177, 196, 200–6, 210–11, 214, 222, 233, 242–43, 250, 262, 280; courts and, 171–75; decentralization and, 74; equality and, 208–9; Hart on, 171–73; institutional design and, 34, 42; judicial review and, 8–9, 175n17; justification and, 103; liberal guarantees and, 196, 202, 207, 212n33;

McLean v. Arkansas, 216n50; political argument and, 134–35, 159, 162; political obligation and, 269–72; pragmatism and, 34, 42, 250n78, 277, 285; reflexivity and, 161–63, 171–75, 179, 181, 184–90; rules of standing and, 172; U.S. Supreme Court and, 9, 175n17, 280 legitimacy, 5, 21–24; actions of state and, 264; argument and, 133, 152; coercion and, 264, 270, 275; conflict and disagreement and, 270–71; justification and, 94, 108n47, 111; liberal guarantees and, 193, 195, 198–200, 205, 209; as moral question, 265; normative, 4, 22, 167, 195, 199–200, 207, 222, 248, 255, 262–63, 269–75; political obligation and, 263–75; pragmatism and, 233, 238, 245, 250, 256, 260, 263–75, 269n9, 270–76, 282; reflexivity and, 167, 171, 178, 184; self-interest and, 264–73 Levin, Jonathan, 148n53 liberal guarantees, 23–24; burdens of judgment and, 214–15; coercion and, 197, 202, 206–8; communication and, 207, 209, 220n66; conflict and disagreement and, 195n1, 198, 207, 215n47, 219, 220n66, 221; decision making and, 194–211, 212n33, 215n47, 218–19, 220n65; dimensions of freedom and, 199–205; diversity and, 194, 198, 206, 208; economic issues and, 193, 199, 203, 207, 217; equal opportunity for political influence and, 197–98, 210, 217n53; equilibrium and, 219; freedom and, 195–208, 218; horizontal relationships and, 207–8, 210, 225–26, 230, 233, 249; importance of equality and, 208–10; information and, 194; institutional arrangements and, 196,

317

In d e x liberal guarantees (cont’d) 221; institutional choice and, 193–94, 198; institutionalizing formal equality and, 211–21; institutionalizing formal freedom and, 205–8; institutional performance and, 194–97, 200–1, 206–10; legal issues and, 196, 202, 207, 212n33; legitimacy and, 193, 195, 198–200, 205, 209; moral issues and, 215n47, 221; neutrality and, 211–12; nondomination and, 202–4; noninterference and, 201–4, 254; nonlimitation and, 204–5, 231; normative theory and, 193–95, 198–201, 205–7, 210–15, 220; overlapping consensus and, 214; philosophy and, 216n50; pragmatism and, 193–98; reasonable pluralism and, 213–16; religion and, 217; selfinterest and, 212–13, 219–20; social norms and, 198, 204–5, 208n20; unrestricted domain and, 211–12, 216n51; vertical relationship and, 207–8; voting and, 194, 197, 199, 202, 208–9, 211, 217n52, 220; war and, 202, 212 liberalism: effective participation and, 222, 235, 249; justification and, 108, 112n61, 140n33, 143n42, 193–222, 235, 249, 284 liberty, 140n34, 199, 201–2, 217n52, 255 Lippmann, Walter, 36n42, 101n24 List, Christian, 116n74, 149n56, 157n84 Luskin, Robert C., 141n37 Mackie, Gerry, 114n69, 143n42, 145n46, 146n47, 149n56 McKelvey, Richard, 72n64, 77n78, 111n55 McLean v. Arkanses, 216n50 macroeconomic analysis, 80n86, 154–58 majority rule, 9, 241n52, 248n76; argument for, 115–17; fairness and, 117; general will and, 114–15; justification and, 101, 112–17; May’s theorem and,

116; Pareto efficiency and, 113; reasonableness of, 113–14; supermajority and, 147–48 Manin, Bernard, 141n36 Mansbridge, Jane, 118n78, 119, 277n20 markets: Coasian bargaining and, 21, 66–80, 176, 180, 189; decentralization and, 51–89; institutional emergence and, 61–71; institutional interaction and, 60–61; macroeconomic analysis and, 80n86, 154–58; microeconomic analysis and, 56–61, 69, 118; privileging, 59–71; textbook, 55–61; trading environment and, 61. See also economic issues Maskin, Eric, 113, 115, 116n74, 117 May’s theorem, 116 Meirowitz, Adam, 178, 220n65 metainstitution, 12 method of science, 28, 36n42, 40 microeconomic analysis, 56–61, 69, 118 military issues, 11, 285n40; competition and, 12n33; war and, ix, 4, 150n61, 202, 212 Mill, John Stuart, 140n34, 156n79, 159n88, 201, 217n52 Miller, Gary, 86n98, 87n99 minimalist approach, 101n24, 108, 143, 209n21, 215–16 moral issues, ix, 16, 21; decentralization and, 61n34, 64n47; ethics and, 1, 37, 94n4, 121n89, 134n18, 265, 288; justification and, 94, 96, 100, 105, 108n47, 109, 120n86, 121n89; legitimacy and, 265; liberal guarantees and, 215n47, 221; political argument and, 134n18, 144n43, 157; pragmatism and, 29, 32–37, 236, 265–66, 270–73; reflexivity and, 177, 179; second-best alternatives and, 94 Nagel, Thomas, 278–79 Nalebuff, Barry, 148n53

318

In d e x Nash equilibrium, 82, 87, 129, 130n6 Nazis, 29–32, 202 neutrality: justification and, 113, 116; technocratic, 71; voter equality and, 113, 211–12, 216, 220–21 nondictatorship, 107n43, 110, 113n67, 146–47, 211 nondomination, 202–4 nonfoundationalism, 44n64 nonimposition, 211n31 noninterference, 201–4, 254 nonlimitation, 204–5, 231 normative legitimacy, 4, 22; liberal guarantees and, 195, 199–200, 207; pragmatism and, 222, 248, 255, 262–63, 269n9, 270, 274; reflexivity and, 167 normative theory, 19–24; acceptance of democracy as terms of politics, 275– 78; burden of justification and, 93–97, 102–15, 118, 121–27; causal mechanisms and, 14; decentralization and, 58–61, 68–73, 76, 79, 89; Dewey and, 4–5 (see also Dewey, John); horizontal relationships and, 226–29; legitimacy and, 4, 22, 167, 195, 200, 262–66, 269–75; liberal guarantees and, 193–95, 198–201, 205–7, 210–15, 220; method and, 17–18; pragmatism and, 37–38, 42–45, 49–50, 222, 226, 234–35, 248, 252–66, 269–78; Rawls and, 17; reflexivity and, 167, 169, 174, 177–81, 189–90; role of political argument and, 128–35, 138–47, 157, 162, 165–66; task of analysis and, 13–14, 16; violence and, 3 voting and, 123 North, Douglass, 189 Nozick, Robert, 120n85 Oakeshott, Michael, 39 Ober, Josiah, 10–13, 249 O’Donnell, Guillermo, 143n42 One Economics, Many Recipes (Rodrik), 10

Ostrom, Elinor, 10n26, 170n5, 251 Overton, William, 216n50 Page, Scott, 72n64, 77n78, 151n64 Pareto efficiency, 56, 59n27, 65, 74, 110, 113, 116n74, 146–47 Pateman, Carole, 277n20 Peirce, Charles, 27–28, 36n42, 39–40, 280, 282 Persson, Torsten, 147n48 Pettit, Philip, 7n14, 157n84, 201n8, 203n13 Phillips, Anne, 247n74 philosophy, 285n42; anti-skepticism and, 21, 27–28, 31–32, 39, 96, 283–84; consequentialism and, 18, 21, 27–28, 31–32, 39, 44–47, 94–98, 104, 190, 194, 196n4, 257, 262, 266–74, 283–84; democratic commitments and, 33–34; fallibilism and, 21, 26–28, 31–32, 39–40, 45, 96, 159, 257, 283–84; indeterminacy and, 2, 13, 26–27, 43, 82, 107, 116, 137, 146–47; justification and, 96–100, 109; liberal guarantees and, 216n50; nonfoundationalism and, 44n64; pragmatism and, 25–33, 96, 264; role of political argument and, 132–33, 146n47; speech acts and, 132–33; utilitarianism and, 44n64, 47, 94, 195n2 Plotke, David, 195n1, 285n40 Plott, Charles, 148n52 political argument. See argument political obligation: compliance and, 267–74; conflict and, 270–71; democratic choice and, 270–71; ethics and, 265–66; legal issues and, 269–72; legitimate authority and, 263–75; moral issues and, 265–66; self-interest and, 264–73 politics: acceptance of democracy as terms of, 275–78; African-American, 103; analytical task of, 13–14;

319

In d e x politics (cont’d) bureaucracy and, 175–81; campaign finance and, 228; circumstances of, 3; coercion and, 3–4, 51, 87n99, 134, 202, 206–8, 237, 254, 270, 275; community-based cooperation and, 79–83; conflict and, 1 (see also conflict and disagreement); consensus and, 20, 29–32, 42, 50 (see also consensus); constitutional, 8 (see also constitutional provisions); decentralization and, 51–89; defunding and, 8; diversity and, 1, 7, 8n15, 10, 18, 23; elections and, 95, 102–3, 107–10, 118–19, 143, 153, 177–79, 217n52, 222, 250; equality and, 20, 23, 37, 45, 47, 54n7, 69, 118–23 (see also equality); equal opportunity for influence and, 20, 197–98, 210, 217n53, 225–28, 232–38, 247, 255, 262–63; excluded groups and, 120; exit strategy and, 3, 162, 282–83; game theory and, 2n2, 18, 61–63, 64n47, 77n78, 80–83, 86–87, 110, 111n55, 129–30, 132n12, 134, 146, 165n95, 178, 219, 220n66, 281n28; ideal vs. nonideal theories and, 14–15; impeachment and, 8; incentive-compatible rules and, 83–88; inevitability of, 252–55, 271–72; institutional experimentalism and, 45–50; interference and, 34, 196–98, 201–8, 217, 222–27, 230–33, 249–50, 254, 278; legitimacy and, 263–75 (see also legitimacy); liberty and, 140n34, 199, 201–2, 217n52, 255; majority rule and, 9, 101, 112–17, 147–48, 241n52, 248n76; minimalist approach to, 101n24, 108, 143, 209n21, 215–16; nondictatorship and, 107n43, 110, 113n67, 146–47, 211; normative theory and, 3–5 (see also normative theory); obligation and, 263–75; polycentric systems and,

251; pragmatism and, 28–30 (see also pragmatism); prisoner’s dilemma (PD) and, 62, 80–83, 87, 131n8; property rights and, 9–10, 60–61, 64n47, 65, 72–75, 77n78, 78, 80n86; reconciliation and, 15n41, 16n42, 88, 215n47, 229, 232–33, 253, 258–59; representative government and, 102, 156n79, 217n52, 218; second-best alternatives and, 94; state-centered context and, 103; state power and, 248–51; unrestricted domain and, 107n43, 110, 113n67, 114, 146, 148–49, 211–12, 216n51; violence and, 3–5, 24, 101, 124, 256, 275, 278, 282–85 populism, 108, 112n61 Posner, Richard, 29–32, 37, 101n24 pragmatism, 20, 23; acceptance of democracy as terms of politics, 275–78; American, 25n2, 27n14; anti-skepticism and, 21, 27–28, 31–32, 39, 96, 283–84; better decisions and, 151–61; burden of justification and, 93–100, 104–5, 115–17, 263–75; communication and, 36, 39, 46; community and, 37–42, 232, 241, 268, 274, 283; competition and, 32, 50; compliance and, 231–32, 264, 267–74, 277; conflict and disagreement and, 39–41, 229–33, 244, 248, 253–55, 259–63, 268n8, 270–71, 273–76, 281–82; consequentialism and, 18, 21, 27–28, 31–32, 39, 44–47, 94–98, 104, 190, 194, 196n4, 257, 262, 266–74, 283–84; contemporary depictions of, 97; cultural criticism and, 97; decision making and, 32, 43–45, 48, 51–54, 225–27, 230–32, 235, 239, 241, 246, 249, 252, 255n90; defining, 25–26; democratic commitments and, 33–36; Dewey and, ix-x, 25–42, 45–46, 49, 266, 279– 85; difference from traditional, 257; diversity and, 36–37, 40, 42–45, 49, 230,

320

In d e x 234, 251, 282; economic issues and, 28, 31, 36, 38, 42–43, 44n64, 223–29, 236, 243–48, 254, 255n90, 278–83; effectiveness and, 222–55 (see also effectiveness); ethics and, 37, 265, 288; experimentation and, 28, 43–50, 250, 257–58, 263, 281–83; fallibilism and, 21, 26–28, 31–32, 39–40, 45, 96, 159, 257, 283–84; Heidegger and, 29–32; holism and, 27–28; implications of, 36–42; indeterminacy and, 2, 13, 26–27, 43, 82, 107, 116, 137, 146–47; information and, 40, 234, 238–39, 242, 282; institutional arrangements and, 33, 35, 41–43; institutional choice and, 46–50, 252–58, 261–63, 270–73, 278; institutional design and, 25–50; institutional performance and, 169, 222–24, 227, 230–34, 243, 249, 254–55, 258, 267–70, 273–74; legal issues and, 34, 42, 250n78, 277, 285; legitimacy and, 222, 233, 238, 245, 248, 250–78, 282; liberal guarantees and, 193–98; moral issues and, 29, 32–37, 236; nonfoundationalism and, 44n64; normative theory and, 37–38, 42–45, 49–50, 222, 226, 234–35, 248, 252–66, 269–78; philosophical view of, 25–33, 264; political consequences and, 25–33; Posner and, 29–32, 37; realist approach and, 278–79, 282, 283n35; reconstructive characterization of, 26; reflexivity and, 40–41, 169n4, 170n5, 176, 187–90; religion and, 230, 232; robustness and, 252– 53, 258; role of political argument and, 143, 151–61, 165; Rorty and, 29–32, 48n72; scientific inquiry and, 28, 33–34, 36n42, 40–41, 46, 246, 280; self-interest and, 224, 264–73, 277; social norms and, 42, 225, 229–33, 276; state power and, 248–51; technology and, 255; utopian, 15,

16n42, 24, 37, 143n42, 256, 278–86; violence and, 256, 275, 278, 282–85; voting and, 225, 241n52, 247, 260 principal-agent models, 155n74, 176–80 prisoner’s dilemma (PD), 62, 80–83, 87, 131n8 promises, 119, 134, 227, 235 property rights, 9–10, 60–61, 64n47, 65, 72–75, 77n78, 78, 80n86 Przeworsky, Adam, 101n24, 141n36 Public and Its Problems, The (Dewey), ix, 35, 40, 96–98 Putnam, Hilary, 37n48 Rabin, Matthew, 130n6, 220n66 racial issues, 181, 202, 230, 232 Radin, Margaret, 100 rational-choice models: advocates of, 18; critics of, 18; institutions and, 17–18, 59n28, 63, 68, 81n88, 99, 130–31; pragmatism and, 18 Rawls, John, John, 15–17, 265n4; effective participation and, 228n15, 235–37, 242; liberal guarantees and, 196n4, 212–16 realist approach, 12, 15; decentralization and, 57, 58n21; justification and, 95, 105–8; pragmatism and, 36n42, 278–79, 282, 283n35; reflexivity and, 182–83, 188; role of political argument and, 143 reasonable pluralism, 213–16 reconciliation, 15n41, 16n42, 88, 215n47, 229, 232–33, 253, 258–59 reductionism, 10n26 Reed, Adolph, 103n35 reflexivity, 19, 23, 261; argument and, 161–67, 172–75; benefits of, 170; bureaucracy and, 175–81; choice and, 167–70, 173–74, 179–85, 189; clarifying role of, 168; Coasian bargaining and, 176, 180, 189; communication and, 170, 182, 186–89; community

321

In d e x reflexivity (cont’d) and, 39, 171, 174, 181, 184–85, 189–90; courts and, 171–75; decision making and, 167–76, 179–85, 188–90; diversity and, 167, 180; economic issues and, 168, 170n5, 177, 182; equilibrium and, 168; hybrids and, 181–83; information and, 167–70, 177, 180–81; institutional choice and, 168–69, 173–74, 179–82, 189; institutional performance and, 169–75, 188; legal issues and, 161–63, 171–75, 179, 181, 184–90; legitimacy and, 167, 171, 178, 184; moral issues and, 177, 179; normative theory and, 167, 169, 174, 177–81, 189–90; organization of competition and, 167–71; pragmatism and, 40–41, 169n4, 170n5, 176, 187–90; principal-agent models and, 176–80; realist approach and, 182–83, 188; science and, 178; second-order priority and, 163–71, 176, 180–83, 188; self-interest and, 178; social norms and, 181–90; technology and, 178; voting and, 167–68 religion, ix, 280; justification and, 99, 103n35; liberal guarantees and, 217; political argument and, 135, 154; pragmatism and, 230, 232; social norms and, 232 Riker, William, 108nn47,48, 112nn61,63, 114n69, 117n75, 147n50, 150n61 Risse, Mathias, 115–17 robustness: argument and, 148, 164–65; decentralization and, 68–71, 76, 83, 87, 89, 258–59; effectiveness and, 252–53; justification and, 95, 114; pragmatism and, 252–53, 258; significance of, 258–59 Rodrik, Dani, 10–13, 43n62, 169n4, 170n5, 249 Roemer, John, 244–45

Rogers, Joel, 120n88, 247 Rorty, Richard, 28n16, 29–32, 48n72, 186–87, 280n26, 285n42 Rosenthal, Howard, 147n48 Rousseau, Jean-Jacques, 108–9, 152, 156 Rubinstein, Ariel, 130n7 Rule of Four, 9 rules: community, 38; compliance and, 104n37, 134, 184, 231–32, 264, 267–74, 277; Condorcet jury theorem and, 112–13, 116, 153–58, 220; decentralization and, 83–88; incentivecompatible, 83–88; institutional, 2, 7, 46, 56, 64–65, 68, 147, 182, 229, 267–71; legal issues and, 172 (see also legal issues); majority, 9, 101, 112–17, 147–48, 241n52, 248n76; pragmatism and, 227–29, 233, 241n52, 247–54, 257, 267–72, 275; privileging markets and, 55–71 Sabel, Charles, 187 Sally, David, 131n8 Satterthwaite, Mark, 107n43 Satz, Debra, 231n23 Schelling, Thomas, 7n14, 54n6, 64, 120n85, 178–79, 281n28 Schofield, Norman, 111n55 Schotter, Andrew, 64nn47,49 Schumpeter, Joseph, 108, 143n42, 159n88, 209n21 science: creation, 215–16; intellectual development and, 246; political, 130n7, 159; pragmatism and, 28, 33–34, 36n42, 40–41, 46, 246, 280; reflexivity and, 178; social, 98, 119n79, 230, 276 second-order priority: decision making and, 12, 19–21, 23, 42, 56, 95, 163– 71, 176, 180–83, 188, 194, 252–53, 260, 273; reflexivity and, 163–71, 176, 180–83, 188 secret ballot, 7–8, 217

322

In d e x self-interest: decentralization and, 53, 67–68, 75, 81n88; excluded groups and, 120; justification and, 120; legitimacy and, 264–73; liberal guarantees and, 212–13, 219–20; political obligation and, 264–73; pragmatism and, 224, 264–73, 277; reflexivity and, 178; role of political argument and, 155–56; unrestricted domain and, 213 Sen, Amartya: argument and, 149n56; effective participation and, 223–24, 230, 237–40, 242–44, 255n90; justification and, 94n3, 109n50; liberal guarantees and, 200n7, 205–6 Senate rule XXII, 9 Seven Member Rule, 9 Shapiro, Ian, 126n98, 127n101, 259n1 Shepsle, Kenneth, 111n55, 147n48, 281n28 Shiffman, Gary, 133n15 social choice theory, 18, 22, 260, 273; argument and, 143, 146–49, 162, 260; decentralization and, 82; effective participation and, 241n53; justification and, 95, 102n31, 106–11, 115, 118, 123; liberal guarantees and, 211, 216, 220; reflexivity and, 167 social norms, 5, 23; alternative definitions of, 230; argument and, 131, 136, 182; customs and, 229–34; decentralization and, 56n10, 62, 64, 82–83, 185; decision making and, 181–85, 188–90; democracy and, 183–90; effective participation and, 229–34; emergence of, 182; evolution of, 185; freedom and, 205, 229–34 (see also freedom); hybrids and, 181–83; institutional choice and, 182–83; institutional experimentation and, 187; justification and, 119n82; liberal guarantees and, 198, 204–5, 208n20; pragmatism and, 42, 225, 229–33, 276; reflexivity and, 181–90; relevancy

of, 182; religion and, 232; spillover effect and, 185, 188 spillover effect, 185, 188 Spitz, Elaine, 101n25 Stiglitz, Joseph, 156n80 Stone-Sweet, Alec, 174n16 Sunstein, Cass, 125n97 supermajority, 147–48 Tabellini, Guido, 147n48 Taylor, Michael, 81n88 technology, ix; decentralization and, 65, 83–88; incentive-compatible rules and, 83–88; justification and, 98, 125n97; pragmatism and, 255; reflexivity and, 178 transaction costs, 12n33, 59, 66, 75–78 Unger, Roberto Mangabeira, 48n72, 70n60 unrestricted domain: argument and, 146, 148–49; liberal guarantees and, 211–12, 216n51; justification and, 107n43, 110, 113n67, 114 U.S. Constitution, 177, 280 U.S. Supreme Court, 9, 175n17, 280 utilitarianism, 44n64, 47, 94, 195n2 utopianism: Dewey, and, 37; pragmatist democracy and, 24, 256, 278–86; Rawls on, 15, 16n42; realistic, 15; Schumpeter and, 143n42 Vermeule, Adrian, 117n75, 259n1 vertical relationships, 207–8 violence: Dewey and, 3; justification and, 101, 124; politics and, 3–5, 24, 101, 124, 256, 275, 278, 282–85; pragmatism and, 256, 275, 278, 282–85 voting, 7n14, 19, 22; aggregate values and, 106–15; alternative policies and, 109; ambiguity of, 107–8, 111–15; anonymity and, 113, 116n74, 157n84, 211–12, 217; approval and, 112;

323

In d e x voting (cont’d) burden of justification and, 95–96, 101–19, 123, 127; Condorcet jury theorem and, 112–13, 116, 153–58, 220; counting schemes and, 106–9; decision making and, 9 (see also decision making); elections and, 95, 102–3, 107–10, 118–19, 143, 153, 177–79, 217n52, 222, 250; epistemic interpretation of, 152–58; franchise distribution and, 8; instability of, 107–11, 260; institutionalizing formal equality and, 211; liberal guarantees and, 194, 197, 199, 202, 208–9, 211, 217n52, 220; majority rule and, 9, 101, 112–17, 147–48, 241n52, 248n76; neutrality and, 113, 211–12, 216, 220–21; normative theory and,

123; options and, 109, 112–13; populist theories and, 108; positional methods and, 112; pragmatism and, 225, 241n52, 247, 260; reflexivity and, 167–68; role of political argument and, 128, 142–58, 162–63 Waldron, Jeremy, 156n79, 173 Walzer, Michael, 1n1, 227n14 war, ix, 4, 110, 150n61, 202, 212 Weale, Albert, 156n79 Weber, Max, 184 Westbrook Robert, 29n20, 96 World War II era, 110, 202 Young, Iris Marion, 133n16, 134n18, 137n26, 144n43 Young, Oran, 104n37

324