Dokumen - Pub - Philosophy Politics Democracy Selected Essays 9780674271579

P H I L O S O P H Y, P O L I T I C S , D E M O C R A C Y
P H I L O S O P H Y,
POLITICS,
DEMOCRACY
Selected Essays
JOSHUA COHEN
H A R VA R D U N I V E R S I T Y P R E S S
Cambridge, Massachusetts, and London, England
2009
Copyright © 2009 by the President and Fellows of Harvard College
all rights reserved
Printed in the United States of America
Library of Congress Cataloging-in-Publication Data
Cohen, Joshua, 1951–

Philosophy, politics, democracy : selected essays / Joshua Cohen.
p. cm.
Includes bibliographical references and index.
ISBN 978-0-674-03448-8 (alk. paper)
1. Social sciences—Philosophy. 2. Political science—Philosophy. 3. Democracy. I. Title.
H61.15.C64 2009
321.8—dc22 2008052224
For Ellen
CONTENTS
Acknowledgments ix
Introduction 1
1 Deliberation and Democratic Legitimacy 16
2 Moral Pluralism and Political Consensus 38

3 Associations and Democracy 61
with Joel Rogers
4 Freedom of Expression 98
5 Procedure and Substance in Deliberative Democracy 154
6 Directly Deliberative Polyarchy 181

with Charles Sabel
7 Democracy and Liberty 223
8 Money, Politics, Political Equality 268

9 Privacy, Pluralism, and Democracy 303
10 Reflections on Deliberative Democracy 326
11 Truth and Public Reason 348
Index 387
ACKNOWLEDGMENTS
T he essays collected here were written over a period of twenty-two

years, between 1986 and 2008. Their footnotes underscore the many
debts I have accumulated in writing them. I wish here to add a few ac-
knowledgments that apply to the collection itself.
Some of these essays existed in usable electronic form thanks only to
optical scanning. I am very grateful to Paul Gowder and Marilie Coetsee
(along with Christine Kim) for their generous assistance in getting the
papers in shape. Lindsay Waters, my editor at Harvard University Press,
urged me to bring this volume together. Lindsay has made a large contri-
bution—as editor and author—to fostering a more democratic society, so
I am very pleased that the book appears on his watch. Tim Scanlon’s
name appears in many footnotes, because of both the influence of his
work and his incisive comments on many of the papers. I taught at MIT
for twenty-nine years, and all the papers in this collection were written
while I was there (even the final paper was first drafted in 2005). At MIT,
I was privileged to have a wonderful group of colleagues and to teach a
remarkable bunch of undergraduates and graduate students. I single out
Archon Fung, whose imprint appears directly on the work, as research as-
sistant on essay 4, and as coauthor on some of the work that shapes essay
10. I am especially grateful to Joel Rogers and Chuck Sabel, each a coau-
thor, friend, and inspiring example.
John Rawls’s influence is evident throughout. The essays often draw di-
rectly on his views, and many ideas presented in them struck me first
when I was engaging with his work. More than that, Rawls conveyed, in
his writing, his teaching, and his bearing, a sense of the importance of
philosophical thought. I have tried to do some justice to that importance.
The essays are presented in the order in which they were written, and I
have made only very slight editorial changes: there is some overlap, and I
ix
A C K N O W L E D G M E N T S
have not tried to eliminate it. Nor have I updated the more empirical ma-
terials or the discussion of constitutional cases. Although reviewing them
has reminded me of their many flaws, I still hope that others will benefit
from reading them. The essays originally appeared in the following pub-
lications, and I thank the publishers for their permission to reprint:
Essay 1: “Deliberation and Democratic Legitimacy,” in Alan Hamlin
and Phillip Petit, eds., The Good Polity (Oxford: Blackwell, 1989),
17–34.
Essay 2: “Moral Pluralism and Political Consensus,” in David Copp,
Jean Hampton, and John Roemer, eds., The Idea of Democracy
(Cambridge: Cambridge University Press, 1993), 270–291. Copy-
right © 1993 Cambridge University Press. Reprinted with permis-
sion.
Essay 3: “Associations and Democracy” (with Joel Rogers), Social Phi-
losophy and Policy 10, 2 (Summer 1993): 282–312. Copyright © 1993
Social Philosophy and Policy Foundation. Reprinted with the per-
mission of Cambridge University Press.
Essay 4: “Freedom of Expression,” Philosophy & Public Affairs 22, 3
(Summer 1993): 207–263, by Blackwell Publishing Ltd.
Essay 5: “Procedure and Substance in Deliberative Democracy,” in
Seyla Benhabib, ed., Democracy and Difference: Changing Bound-
aries of the Political (Princeton: Princeton University Press, 1996),
95–119.
Essay 6: “Directly Deliberative Polyarchy” (with Charles Sabel), Eu-
ropean Law Journal 3, 4 (December 1997): 313–342, by Blackwell
Publishing Ltd.
Essay 7: “Democracy and Liberty,” in Jon Elster, ed., Deliberative De-
mocracy (Cambridge: Cambridge University Press, 1998), 185–231.
Copyright © 1998 Jon Elster. Reprinted with the permission of
Cambridge University Press.
Essay 8: “Money, Politics, Political Equality,” in Alex Byrne, Robert
Stalnaker, and Ralph Wedgwood, eds., Fact and Value (Cambridge,
MA: MIT Press, 2001), 47–80.
Essay 9: “Privacy, Pluralism, and Democracy,” in Joseph Keim Camp-
bell, Michael O’Rourke, and David Shier, eds., Law and Social Jus-
tice (Cambridge, MA: MIT Press, 2005), 15–40.
Essay 10: “Reflections on Deliberative Democracy,” in Thomas
x
A C K N O W L E D G M E N T S
Christiano and John Christman, eds., Contemporary Debates in Po-

litical Philosophy (Oxford: Blackwell, 2009), 247–263.
Essay 11: “Truth and Public Reason,” Philosophy & Public Affairs 37,
1 (Winter 2009): 2–42, by Blackwell Publishing Ltd.
xi
P H I L O S O P H Y, P O L I T I C S , D E M O C R A C Y
INTRODUCTION
T he essays collected here are about an ideal of democracy. It is the

demanding ideal that we, as equals and on a footing of mutual re-
spect, should guide the conduct of our common affairs through the use
of our shared reason.
Although focused on democracy, these essays are animated by a more
general conception of the relationship between philosophical reflection
and social-political life. I begin this Introduction with some comments
about that conception. I then make some semi-autobiographical remarks
about the intellectual background that led me to the animating concep-
tion of democracy. Next, I describe the substance of the essays, how they
are connected, and their lines of evolution. I conclude with some brief
comments on whether the essays express an objectionably Panglossian
outlook.
1
We all have ideas that belong to political morality. We have some mix of
thoughts about rights and responsibilities, about the evil of slavery, cru-
elty, destitution, humiliation, and intolerance, about the fairness of social
inequality and the virtues of open discussion, about religious tolerance
and personal privacy, about whether the state is like a large family or
something fundamentally different, about the importance of a rule of law
and democracy, about when it is acceptable to fight a war and what we
owe to people in other places, and about whether justice is somehow
fixed by the nature of things or is a human construction.
Moreover, those thoughts, both more abstract and more concrete, play
a role in political life. That’s because politics is not only—though it is as-
suredly in (perhaps large) part—a struggle for power and a strategic pur-
suit of personal and group interests. Particularly in a democracy, politics
1
I N T R O D U C T I O N
is also a matter of expressing convictions that belong to political morality

and of arguing with others about how we should act in ways that keep
faith with our separate and shared convictions.
Consider the response of Alberto Mora, the U.S. Navy’s general coun-
sel, to his discoveries about the cruel treatment of prisoners sanctioned
by the Bush administration in conducting its “war on terror.” Mora
thought—and acted on the thought—that “cruelty” and torture, though
different, are “equally pernicious.” “If cruelty is no longer declared un-
lawful . . . it alters the fundamental relationship of man to government. It
destroys the whole notion of individual rights. The Constitution recog-
nizes that man has an inherent right, not bestowed by the state or laws, to
personal dignity, including the right to be free of cruelty. It applies to all
human beings, not just in America—even those designated as ‘unlawful
enemy combatants.’ If you make this exception, the whole Constitution
crumbles.”1 Here we have an argument—concise and clear, whether cor-
rect or incorrect—about the equivalence in political morality of cruelty
and torture, concluding that torture and cruelty are both to be con-
demned as violations of a right to personal dignity. Thoughts such as
this, which prompted Mora to concerted action, belong as much to our
political life as the pressures that led to the abuses he sought to halt (and
as the defenses of torture and cruelty as warranted under special circum-
stances).
Cynics will say that appeals to normative ideas in political argument
are a high-minded sham. They say that we should not, for example, take
Lincoln seriously when he says that the country was born in an idea and
dedicated to a proposition. They say that the Universal Declaration of
Human Rights is a strategy of political control by Western powers or by
an emerging class of globe-circling lawyers and bureaucrats. The cynics
say that normative conviction and argument provide a smokescreen for
the interests of those who have power and a distraction for those who
don’t. Political struggle, they remind us, is not discussion in the seminar
room: it is not an argument about who is right, or a proof of a theorem, or
an experiment to test a hypothesis, but a conflict about who will win.
Cynicism (tendentiously self-styled as realism, sometimes more simply as
the grown-up view) strikes me as an implausible position, though its tire-
less reiteration threatens to turn it into a self-fulfilling prophecy. But I am
1. See Jane Mayer, The Dark Side (New York: Doubleday, 2008), 219.
2
not here arguing for a thesis, only expressing an outlook, the outlook ex-
pressed in these essays, which all are premised on a rejection of cynicism
about normative political discourse.
Normative political discourse is one thing, however; political philoso-
phy is something else. What do these observations about the place of nor-
mative thought in politics have to do with philosophy?
In his Prison Notebooks, Antonio Gramsci draws attention to the role
of normative ideas in our individual and collective conduct.2 Formu-
lating the point in striking terms, he says, “Everyone is a philosopher.”
That is because a “specific conception of the world” underlies our intel-
lectual activity. That conception, Gramsci says, is typically implicit, often
adapted from the “external environment,” passively internalized rather
than worked out “consciously and critically.” Gramsci probably over-
states the extent to which we each embrace a “specific conception of the
world,” a comprehensive doctrine with a significant degree of coherence
and stability, as distinct from a somewhat more inchoate family of convic-
tions. Still, his comment points to the role of a more or less connected set
of ideas in our ordinary understanding and conduct, a conception of the
world that comprises normative ideas, even if it is not confined to them.
More to the point, Gramsci’s invocation of philosophy suggests conti-
nuity between endorsing the conception of the world that shapes our
conduct—what he calls a spontaneous philosophy—and pursuing the
activity that we call political philosophy.3 Think of political philosophy as
a more reflective engagement with the political morality that shapes our
conduct and that we offer in argument to others, an engagement that
aims to bring greater coherence, comprehensiveness, and—most impor-
tantly—reasoned consideration to that morality. We suppose that our
convictions withstand scrutiny, and we think they are worthy of our alle-
giance—and suitable for political argument—because they do. Political
philosophy puts this thought to the test, aiming to judge, for example,
whether cruelty and torture are indeed morally on a par, each violating a
right to dignity; or whether treating people differently because of race or
gender is on a par with treating them differently because of differences in
2. Antonio Gramsci, Prison Notebooks, ed. and trans. Quintin Hoare and Geoffrey Nowell
Smith (New York: International, 1971), 323.
3. Gramsci distinguishes the possession of a conception of the world from a “second level,
which is that of awareness and criticism,” and he emphasizes the importance—sometimes the
necessity—of ordering “in a systematic, coherent and critical fashion one’s own intuitions of
life and the world.” Ibid., 323, 327.
3
sexual orientation; or whether the rights that many of us suppose people

to be entitled to are really anything more than the conditions for a
smoothly functioning liberal, market society; or whether our convictions
about the moral significance of political boundaries express an insight
about justice or an unreflective prejudice; or whether there really is a
morally salient distinction between forcing people to work and taxing
their income. Political philosophy, thus understood, is animated by con-
cerns about whether the normative views we are drawn to—views we take
to be reasonable and correct and offer in argument to others—are as
compelling on reflection and worthy of allegiance as we take them to be.
Because normative thought is part of politics, and political philosophy
is continuous with normatively inflected public discussion, political phi-
losophy is continuous with politics. And that continuity is particularly
clear in pluralistic societies, where political ideals compete and argu-
mentative challenges to those ideas are settled parts of the political cul-
ture. Political philosophy does not, then, aim to substitute reflective rea-
son for politics, as if political life itself shared nothing with philosophical
argument, as if it were all about power and interests. Nor is political phi-
losophy about judging the political world. Instead, the point of political
philosophy is to contribute reflectively to the public reasoning about
what we ought to do that always already forms one (sometimes strikingly
small) part of political life.
The idea that political philosophy is continuous with politics informs
the essays collected here in at least two ways. First, many of these essays
grew out of and were engaged with a political controversy. The philo-
sophical argument locates the problem in a larger setting, with the hope
of getting some practical guidance in how best to address it. “Freedom
of Expression” (essay 4), for example, was prompted by debates in the
United States in the 1980s and early 1990s about regulations of pornogra-
phy and hate speech, campus speech codes, and campaign finance.
Reflection on these issues and the conflicting pulls of competing ideas
about how to address them—that words are not just vehicles for argu-
ment but also instruments of human injury, or that silencing speech is in-
sulting to audiences and destructive to social advance—led to larger
questions about the rationale for freedom of expression, all inspired by
the hope that that rationale might guide judgment on contested ques-
tions. Essay 8, on campaign finance and political equality, started as a
presentation to a group working on campaign finance reform, in particu-
4
lar on efforts to secure public financing, which led to considerations

about the nature of democracy. Essay 9, on privacy, was part of a debate
that resulted from the late 1990s public preoccupation with Bill Clinton’s
sex life. And several of the essays (3, 6, 10) grew out of public political de-
bates of the past three decades about how to achieve constructive regula-
tion for the common good in democratic societies, given concerns about
the capacity of the state to achieve such regulation.
Second, if the concerns in the essays have typically grown out of politi-
cal argument, they also return to it. Political philosophy is philosophy,
but it is also political-practical. Accordingly, I have tried to be attentive to
concerns about how the norms and ideals explored in reflection on nor-
mative political ideas might be realized in the world, and that put these
essays in closer touch with theoretical and empirical materials from the
social sciences than is common in much political philosophy. Philoso-
phy does not settle the question of how we ought to live together, but sen-
sible normative political ideas should be workable in the political world
as it might be, and concerns about such workability are part of a re-
flective assessment of their reasonableness and correctness. Political phi-
losophy has always been practical and political as well as philosophi-
cal, and these essays reflect an effort to be faithful to the inevitable
complexities that result from this attention to both principles and practi-
cal possibilities, to the circumstances to which normative reflection is ad-
dressed, and to the possibility of realizing ideals from where we are, given
who we are.
2
I attended graduate school in the mid-1970s, a period of great creativity in
political philosophy. John Rawls’s Theory of Justice had appeared in 1971,
followed soon by Robert Nozick’s Anarchy, State, and Utopia (1974).
This was also a period of political rethinking for people who, like me,
identified themselves as democrats and socialists and thought of their so-
cialism as principally a matter of political values—egalitarian, participa-
tory, and antiauthoritarian—rather than more specifically institutional
convictions about public ownership or a planned economy.
In 1975–76, I tried to write a PhD dissertation on socialism. The guid-
ing idea was that the socialism worth being concerned about—anyway,
the socialism that interested me normatively—had something to do with
democracy. The point was to create a more democratic society in many
5
senses of that term, including a more democratic economy: to extend

democratic ideas outside the state. And here my political concerns came
into contact with my philosophical interests. While I found Rawls’s views
deeply attractive, I also found it striking that Rawls (not to mention
Nozick and others) had very little to say about democracy and was fo-
cused instead on individual rights and opportunities, and a fair distribu-
tion of income and wealth.4
I did not make much headway on the dissertation topic, but I returned
to the concern about democracy in a paper I wrote in 1980. The paper
sketched a “principle of democratic legitimacy” and tried to show how it
could serve as a fundamental political idea, a kind of axiom that would
organize other political principles, about liberties, economic democracy,
and distributive fairness, which all could be understood as features of a
democratic political society. I thought of democracy as a compelling nor-
mative idea, requiring that people be treated as equals in the processes of
collective decision-making. It also struck me as a demanding one: treat-
ing people as equals had implications for outcomes as well as processes,
so democracy was both a procedural and a substantive ideal, with a “for
the people” as well as a “by the people” dimension.
That paper grew (in much improved form) into the last chapter of On
Democracy, which Joel Rogers and I wrote in the summer of 1982 and
published in 1983. Rogers and I had organized a meeting with a group of
academics and other intellectuals, all of whom shared our convictions
about the misdirection of American politics under Reagan. In the meet-
ing, Rogers and I kept referring to the shared ideas and convictions of the
group. Someone suggested that we draft a statement expressing our sense
of what was shared. Responding to the challenge, we drafted the book.
Soon after On Democracy appeared, I felt dissatisfied with several parts
of the view. Three dissatisfactions are germane here. First, the concep-
tion of democracy—with its emphasis on treating people as equals—
seemed too thin: not wrong, but too limited to serve the central norma-
tive role I had been assigning to it. Second, and not unrelated to the first
point, I thought the brief remarks about personal liberties and democracy
4. I eventually came to see that Rawls’s views were far more deeply shaped by a concep-
tion of democracy than I then understood. See Joshua Cohen, “For a Democratic Society,” in
Samuel Freeman, ed., Cambridge Companion to Rawls (Cambridge: Cambridge University
Press, 2002), 86–138.
6
were insufficient and perhaps misguided. Third, I was troubled by the

“dualism” in our account of the relationship between the book’s norma-
tive conception of democracy as an ideal and its structural theory of cap-
italist democracy. In presenting the account of democracy as an ideal, we
appealed to what we understood to be a widely shared set of political val-
ues. But those values played no role in our account of politics in a cap-
italist democracy, which emphasized structural constraints.
Over the next few years, I wrote on Locke5 and Rousseau6 with these
concerns close at hand, and also explored the idea of an “epistemic” con-
ception of democracy. The results of this exploration appeared in a 1986
essay, “An Epistemic Conception of Democracy.”7 The epistemic con-
ception says, in essence, that democracy is about combining judgments
about what we ought to do—in particular, opinions about the common
good—not aggregating interests or preferences, even the preferences of
people under equal conditions. The critical part of the essay—arguing
that William Riker’s Liberalism Against Populism was too quick to draw
large conclusions about democracy from social choice theory—seems
right, but I was not convinced (even as I was writing the paper) by the
idea that democracy is about offering opinions regarding the right answer
to a reasonably well-defined question and then pooling those answers.
That conception seemed to be missing something about democracy as a
fundamental political value.
The angle of approach that I was eventually drawn to centered on the
idea of deliberative democracy. The deliberative democrat emphasizes
that democracy is not simply about treating people as equals in a process
of collective decision-making, or about fair bargaining among groups,
but also about reasoning together as equals on matters of common con-
cern (or it could be described as treating people as equals by relying on
our common reason as a basis for justification). Here, I was moved in
part by Habermas’s ideal of decisions arrived at through a process that re-
flects no force other than the force of the better argument (though
Habermas had not presented this ideal in connection with a conception
5. “Structure, Choice, and Legitimacy: Locke’s Theory of the State,” Philosophy and Public
Affairs 15, 4 (Fall 1986): 301–324.
6. “Reflections on Rousseau: Autonomy and Democracy,” Philosophy and Public Affairs 15,
3 (Summer 1986): 275–297.
7. “An Epistemic Conception of Democracy,” Ethics 97, 1 (October 1986): 26–38.
7
of democracy). And I was working with the intuitive idea that democracy
might be thought of not only as a fair process or as an instrument for
achieving just ends but as a way to realize in actual political life an ideal
of justification through public reason-giving.
“Deliberation and Democratic Legitimacy” (essay 1) was the result.
Originally written in 1986, it located deliberation—understood as a kind
of mutual reason-giving—as central to the democratic ideal in its most
attractive form. By making such reason-giving central, the deliberative
conception of democracy also showed how democracy was connected,
so I argued, with ideas of autonomy, equality, and the common good: de-
mocracy, understood deliberatively, was both a procedural and a substan-
tive ideal. And also, I thought, a specifically political ideal, not tied to
an encompassing moral outlook. The idea was not that people ought to
reason about everything, or guide their personal choices by reflective, au-
tonomous judgment, or that the unexamined life is not worth living, but
that the legitimacy that emerges from democratic collective choice re-
flects the role of reason-giving—of a kind of mutual justification—in the
process.
3
“Deliberation and Democratic Legitimacy” proposed, then, that the de-
liberative conception, with its idea of reason-giving among equals, pro-
vides the most compelling account of settled elements of the democratic
ideal. And this suggested two lines of further evolution.
First, I wanted to see whether and how the idea of deliberative democ-
racy could serve as a unifying normative political ideal. Addressing that
issue required filling out its normative content, showing how it provided
argumentative traction in addressing disputed issues about liberty, equal-
ity, and the common good. I wrote several essays along these lines, which
aimed to show how deliberation, understood as a kind of reasoning, was
different from simply discussing and how the idea of deliberative democ-
racy could be used to organize an account of other political values. An es-
sential part of the argument, which I developed in “Moral Pluralism and
Political Consensus” (essay 2, originally written in 1990), was the idea of
reasonable pluralism.
Beginning in the mid-1980s, Rawls emphasized that societies that pro-
tect liberties are marked by what he called a “fact of pluralism,” a plural-
ism of distinct and incompatible comprehensive doctrines, religious and
8
philosophical.8 A conception of justice suited to such societies needed,

he argued, to be able to win support from those doctrines, “each from its
own point of view,”9 and this would be aided by confining the conception
of justice and the case for it to specifically political matters. I thought that
the concern about the fact of pluralism was misplaced. Because justice is
not fixed by some sort of de facto consensus, it is not a problem for a con-
ception of justice—a problem that might prompt us to reconsider its
correctness—that it cannot win support from the doctrines that people
happen to be drawn to, whatever the content of those doctrines.10 The is-
sue about pluralism, I proposed, was really about reasonable disagree-
ment, about the fact that there are many reasonable views with differ-
ent fundamental elements, say religious and secular, that people are
drawn to. Political argument should, then, take place on a terrain that
could be endorsed by people with different reasonable views. Though
“Moral Pluralism and Political Consensus” is not specifically about de-
mocracy, a notion of reasonable pluralism and reasonable disagreement
subsequently played an essential role in my account of deliberative de-
mocracy, because I thought that the kinds of reasons that could be used
to justify collective decisions needed to be attentive to the pluralism that
lies at the heart of democratic societies.
In “Procedure and Substance in Deliberative Democracy” (essay 5), I
modified the conception of deliberative democracy to take account of
the fact of reasonable pluralism. Democratic deliberation is about rea-
son-giving to others as equals. Reasonable disagreement, then, restricted
the range of reasons that could be advanced in political justification. It
also led to an account of democracy that was not simply procedural, but
had substantive elements as well, with protections of liberties and a
concern for the common good emerging as elements of democracy. The
paper also explored the implications of the conception of deliberative de-
mocracy for some contested issues of democratic politics: how associa-
8. See the papers on “The Idea of an Overlapping Consensus,” “The Priority of Right and
Ideas of the Good,” and “The Domain of the Political and Overlapping Consensus,” all of
which emphasized the importance of the “fact of pluralism.” See John Rawls: Collected Papers,
ed. Samuel Freeman (Cambridge, MA: Harvard University Press, 1999), chaps. 20, 21, 22.
9. Ibid., 479.
10. The paper presents a criticism of Rawls, but the point was to make a case about what he
really ought to have been saying in his account of political liberalism and overlapping consen-
sus. Rawls did shift focus to the fact of reasonable pluralism, though he also observes that both
facts are relevant. See Political Liberalism (New York: Columbia University Press, 1996), 36–
37, 58–66.
9
tions could play a larger regulatory role in advancing the common good
without ruining deliberative politics, and electoral finance (issues dis-
cussed in much greater detail in essays 3 and 8).
Because deliberative democracy has strong roots in a radical demo-
cratic tradition, emphasizing ideas of political autonomy, I was especially
concerned to show that and how it provided a way to think about the “lib-
erties of the moderns,” including religious and moral liberty. This, as I
mentioned earlier, was a gap in On Democracy. In “Deliberation and
Democratic Legitimacy,” I had tentatively suggested (in footnote 22) that
those liberties provided protections that bolster the confidence required
for independent participation. That thesis had the smell of the lamp.
Maybe confidence born of those protections emboldens people. Maybe
it privatizes. Who knows? And anyway, the argument assigned a kind
of unwanted privilege to political engagement. How could it be that
protections of liberties of conscience and of the person are simply a
shadow cast by the preconditions of participation? But wasn’t that conclu-
sion the natural consequence of an unnatural beginning? If you treat de-
mocracy—or, more abstractly still, politically autonomous lawgiving—as
basic, and are moved by convictions about the liberties of the moderns,
how can you avoid the suspicious result?
To address the issue, I focused initially on freedom of expression.
Though my concerns about freedom of expression grew partly out of an
interest in exploring the implications and plausibility of the conception
of deliberative democracy, I was also independently interested, as I men-
tioned earlier, in addressing a set of public debates about campus speech
codes, pornography, hate speech, and campaign finance. I was troubled
by the hate speech, could understand the concerns about pornography,
and saw that many people who shared an egalitarian sensibility thought
that concerns about regulation required a kind of abstract allegiance to
rights or liberty and that their egalitarianism put them on the side of a
less expression-friendly position.
After reading lots of constitutional cases and thinking about the differ-
ences between American and Western European protections of free ex-
pression, I was persuaded that no simple theory founded on a single
value or interest (autonomy, for example) would suffice to capture the
terrain. I proposed a view that had (to my mind) an attractive complexity,
founded on three basic interests—expressive, informational, and deliber-
ative—and a range of plausible judgments about the costs of regulation.
10
In “Freedom of Expression” (essay 4), I did not present the theory as an il-
lustration of my account of deliberative democracy but as an indepen-
dently plausible view, which needed subsequently to be fit into the ac-
count of democracy, if it could be.11
“Procedure and Substance” pointed in the right direction, but I drew
out the connections most fully in “Democracy and Liberty” (essay 7),
which argues that religious, expressive, and moral liberties are essential
parts of democracy, not constraints upon it, that a political society is less
democratic to the extent that it fails to protect these liberties, and that
those liberties are not destructive of community, but integrative. Once
more, the argument turns on the reasons that can be used to defend regu-
lations in a society of equals, assuming the fact of reasonable pluralism.
Given the kinds of reasons that are expressed in religious conviction, for
example, and the need to justify to others on the basis of reasons that they
can reasonably accept, it will generally not be possible to justify regula-
tions of religious exercise. An essential part of the argument was to tie the
idea of finding such reasons to democracy itself, to the deliberative idea
that democracy is an arrangement of collective decision-making that
treats people as equals in part by offering reasons of a suitable kind. The
reason-giving among equals essential to deliberative democracy, along
with reasonable pluralism, brought the liberties inside the conception of
democracy. Deliberation and pluralism shape the content of what, in the
paper on “Privacy, Pluralism, and Democracy” (essay 9), I call “democ-
racy’s public reason.”
I also had been thinking for some time about political equality, which I
discussed briefly in “Procedure and Substance.” As a way to provide fo-
cus, I concentrated my attention on the troubled American system of
financing elections (essay 8). Once more—in part prompted by practical
engagement with the issue—I sketched an account of political equality
and campaign finance largely independent of the general conception of
deliberative democracy, drawing on arguments in American constitu-
tional law and empirical research on the role of money in elections. The
line of thinking adopted by the U.S. Supreme Court in Buckley v. Valeo
11. I also wrote a companion piece on pornography, “Freedom, Equality, and Pornography,”
in Austin Sarat and Thomas R. Kearns, eds., Justice and Injustice in Law and Legal Theory
(Ann Arbor: University of Michigan Press, 1996), 99–137, and in Jessica Spector, ed., Prostitu-
tion and Pornography: Philosophical Debate About the Sex Industry in the U.S. (Stanford: Stan-
ford University Press, 2006).
11
(1976), and preserved more or less consistently thereafter, reflected, I ar-

gued, an elite conception of democracy in the tradition of democratic
theory associated with Joseph Schumpeter’s Capitalism, Socialism, and
Democracy. The elite conception is defined by a focus on electorally or-
ganized competition for office, as the distinctively democratic method of
deciding which elites would rule. Correspondingly, the conception is
marked by an inattention to the value of fair chances for citizens to exer-
cise various forms of political influence, including influence on public
argument. The challenge, I observed, is to find a way to accommodate
both the fundamental importance of free political speech, unrestricted
by content and viewpoint, and fair chances for equal citizens to influence
public discussion and collective decision, not simply to subordinate the
latter to the former.
The privacy paper (essay 9) aims to clarify further the place of personal
liberties in the deliberative conception of democracy, while underscor-
ing the limits of privacy. I argue that an idea of independent judgment
has an important role in democracy’s public reason, and that this respect
for independent judgment has important implications for authoritative
regulation in areas of life, death, and sex. This case for “privacy rights,” I
argue, does not dependent on philosophical liberalism but is a natu-
ral consequence of the reason-giving among equals and reasonable plu-
ralism associated with deliberative democracy. But—and here are the
limits—the kinds of restrictions on justification that are appropriate to
public, political justification are not to be extended to informal public
discussion in civil society, which can be much more open-ended, more
agonistic, less confined by standards of civility and norms of reticence.
The aim here is to free the ideal of deliberative democracy, with its con-
cerns about justification on common ground, from a cultural conven-
tionalism with which it might be associated.
A second line of thought has moved along a less philosophical track.
Two of the essays (4, 6) focus on how to achieve a more deliberative de-
mocracy rather than on the ideal itself. In 1988 Joel Rogers and I wrote
a long article on “associative democracy.” The central intuitive idea
was that the nature of associative life—the strength and role of associa-
tions intermediate between state and market—made a morally conse-
quential difference to the operations of a capitalist democracy. In part we
were concerned about the decline of unions and other kinds of second-
ary associations in the United States and the implications of that decline
12
for representing interests. But we were also concerned, perhaps more

fundamentally, with limits on state capacity and thus focused on the po-
tential for community organizations, environmental groups, NGOs, and
worker organizations to play a larger and more constructive role in
achieving important regulatory aims. The ideas were connected to West-
ern European and Latin American debates about a successor to cor-
poratism as a scheme of macroeconomic management; in the United
States, these ideas were part of a discussion about how to do anything
constructive given significantly reduced confidence in state capacity.
The full version appeared (with crucial encouragement from Erik Olin
Wright) as “Associations in Democratic Governance,” in Politics and So-
ciety.12 I have included a shorter version here.
The general line of thought—about achieving a reasonably decent,
effective, and egalitarian democracy without relying so heavily on the
state—still seems promising. But the associative view struck me as lim-
ited, both because it paid insufficient attention to deliberation and be-
cause it focused too much on the role of associations, as a specific cate-
gory. The essay on directly deliberative polyarchy (essay 6), jointly
authored with Chuck Sabel, sought to remedy those deficiencies. The es-
say describes a very general architecture of deliberatively democratic
problem-solving. The architecture has application to a range of condi-
tions, even without the state’s shadow looming, and Sabel and I draw in
that architecture in subsequent work on the European Union and on
global democracy.13 It describes a scheme for the provision of important
public goods that is disciplined by reason-giving and the need for com-
parisons but that does not impose the kind of uniformity that is conven-
tionally associated with the state, law, and public administration. It shares
with the associative democracy view a willingness to rely less on the state
(without endorsing any kind of hostility to politics) but does not depend
on enlisting associations as the preferred vehicle for constructive public
participation.
12. Politics and Society 20, 4 (December 1992): 393–472. The essay was accompanied by re-
plies from Ellen Immergut, Andrew Levine, Jane Mansbridge, Philippe Schmitter, Wolfgang
Streeck, Andrew Szasz, and Iris Young.
13. “Sovereignty and Solidarity in the EU” (with Charles Sabel), in Jonathan Zeitlin and
David Trubek, eds., Governing Work and Welfare in a New Economy: European and American
Experiments (Oxford: Oxford University Press, 2003), 345–375; “Global Democracy?” (with
Charles Sabel), New York University Journal of International Law and Policy 37, 4 (2006): 763–
797.
13
The penultimate essay (essay 10) aims at synthesis of some of the more
and the less philosophical arguments. I restate the idea of deliberative de-
mocracy, emphasizing the central role of reason-giving, explain why that
conception is not inattentive to the facts about political power, and dis-
cuss a range of considerations that might lead us to endorse the delibera-
tive view. I then explore some potential tensions between the deliberative
conception of democracy and the proposals to secure a more participa-
tory democracy. Deliberation and participation are arguably both impor-
tant elements of the democratic ideal, but it is easy to see how we might
be led to sacrifice the one for the other. The essay then concludes with a
discussion of some ways to remedy the tensions.
The concluding essay (“Truth and Public Reason”) is the most recent
and the most philosophical. Normative discourse, I said at the outset, is
part of politics. Moreover, it should be taken at face value, as concerned
with what we ought to do, not simply as a mask of attitude and interest. In
addition, it is continuous with the reflective, critical engagement with po-
litical values characteristic of political philosophy. In this essay, I argue
that the concept of truth has a role in public political discourse and pres-
ent a political conception of truth suited to that role. The kind of com-
mon ground that we might aim to occupy in political argument—a
ground of public reason—cannot make room for the whole truth. But be-
cause it is a form of reasoning about what we ought to do, it cannot dis-
avow a concern for truth. Preserving this concern does not turn practice
into theory but affirms the place of reflective normative thought in life.
4
I mentioned that essay 10 is concerned with accommodating the impor-
tance of deliberation and participation, which stand in uneasy tension.
This concern about reconciling tensions underscores a theme that runs
throughout the essays and is an especially important presence in the es-
says on political equality, procedure and substance, and associative de-
mocracy. Much work in political theory points in a very different direc-
tion: it draws our attention to the need to make hard choices. In the spirit
of Max Weber, Isaiah Berlin, and Bernard Williams, it reminds us that
all good things in life do not go together, and reminds us, too, that we
sometimes, in a cheerfully Panglossian or romantic spirit, gloss over these
tensions.
Such reminders are important. But the essays here are animated by a
14
very different idea. My own experience is not that we are constantly yield-
ing to Panglossian temptations. Quite to the contrary, we are often too
quick to suppose that important values cannot be jointly realized, that
political life is filled with tragic conflicts between and among important
values. So we respond to political ideals with a knowing irony of the in-
tellect and a lassitude of the heart.
In his Gettysburg Address, Lincoln drew attention to the idea that the
United States was conceived in liberty and dedicated to a kind of equal-
ity. The question whether a country “so conceived and so dedicated”
could endure was very much open and not to be resolved through philo-
sophical argument. Still, Lincoln urged in effect that we not be too quick
to assume an intractable conflict. Perhaps conscientious and sustained
effort at improvement is not doomed to fail; perhaps we live in a perfect-
ible union. And here philosophy, too, has a role. It must not, of course,
deny the possibility of conflict between important values. But inspired by
hope, guided by reason, and focused on the political world as it might be,
philosophy can help us to see the possibility of more fully realizing our
largest political ideals.
15
1
D E L I B E R AT I O N A N D D E M O C R AT I C L E G I T I M A C Y
I n this essay I explore the ideal of a “deliberative democracy.”1 By a de-

liberative democracy I shall mean, roughly, an association whose af-
fairs are governed by the public deliberation of its members. I propose an
account of the value of such an association that treats democracy itself as
a fundamental political ideal and not simply as a derivative ideal that can
be explained in terms of the values of fairness or equality of respect.
The essay is in three sections. In section 1, I focus on Rawls’s discussion
of democracy and use that discussion both to introduce certain features
of a deliberative democracy, and to raise some doubts about whether
their importance is naturally explained in terms of the notion of a fair sys-
tem of social cooperation. In section 2, I develop an account of delibera-
tive democracy in terms of the notion of an ideal deliberative procedure.
The characterization of that procedure provides an abstract model of de-
liberation, which links the intuitive ideal of democratic association to a
more substantive view of deliberative democracy. Three features of the
ideal deliberative procedure figure prominently in the essay. First, it
helps to account for some familiar judgments about collective decision-
making, in particular about the ways that collective decision-making
ought to be different from bargaining, contracting, and other market-type
I have had countless discussions of the subject matter of this essay with Joel Rogers, and
wish to thank him for his unfailingly sound and generous advice. For our joint treatment of the
issues that I discuss here, see Joshua Cohen and Joel Rogers, On Democracy (Harmondsworth:
Penguin, 1983), ch. 6. The main differences between the treatment of issues here and the treat-
ment in the book lie in the explicit account of the ideal deliberative procedure, the fuller treat-
ment of the notions of autonomy and the common good, and the account of the connection of
those notions with the ideal procedure. An earlier draft of this essay was presented to the
Pacific Division Meetings of the American Philosophical Association. I would like to thank
Loren Lomasky, Alan Hamlin, and Philip Pettit for helpful comments on that draft.
1. I originally came across the term “deliberative democracy” in Cass Sunstein, “Interest
Groups in American Public Law,” Stanford Law Review 38 (1985): 29–87. He cites (n. 26) an ar-
ticle by Bessette, which I have not consulted.
16
D E L I B E R A T I O N A N D D E M O C R A T I C L E G I T I M A C Y
interactions, both in its explicit attention to considerations of the com-

mon advantage and in the ways that that attention helps to form the aims
of the participants. Second, it accounts for the common view that the no-
tion of democratic association is tied to notions of autonomy and the
common good. Third, the ideal deliberative procedure provides a dis-
tinctive structure for addressing institutional questions. And in section 3
of the essay I rely on that distinctive structure in responding to four objec-
tions to the account of deliberative democracy.
1
The idea of deliberative democracy is a familiar ideal. Aspects of it have
been highlighted in recent discussion of the role of republican concep-
tions of self-government in shaping the American constitutional tradition
and contemporary public law.2 It is represented as well in radical demo-
cratic and socialist criticisms of the politics of advanced industrial socie-
ties.3 And some of its central features are highlighted in Rawls’s account
of democratic politics in a just society, particularly in those parts of his
account that seek to incorporate the “liberty of the ancients” and to re-
spond to radical democrats and socialists who argue that “the basic liber-
ties may prove to be merely formal.” In the discussion that follows I shall
first say something about Rawls’s remarks on three such features, and
then consider his explanation of them.4
First, in a well-ordered democracy, political debate is organized
around alternative conceptions of the public good. So an ideal pluralist
2. For some representative examples, see Cass Sunstein, “Naked Preferences and the Con-
stitution,” Columbia Law Review 84 (1984): 1689–1732; idem, “Interest Groups in American
Public Law”; idem, “Legal Interference with Private Preferences,” University of Chicago Law
Review 53 (1986): 1129–1184. Frank Michelman, “The Supreme Court, 1985 Term—Foreword:
Traces of Self-government,” Harvard Law Review 100 (1986): 4–77. Bruce Ackerman, “The
Storrs Lectures: Discovering the Constitution,” Yale Law Journal 93 (1984): 1013–1072.
3. I have in mind, in particular, criticisms that focus on the ways in which material inequali-
ties and weak political parties restrict democracy by constraining public political debate or un-
dermining the equality of the participants in that debate. For discussion of these criticisms, and
of their connections with the ideal of democratic order, see Cohen and Rogers, On Democ-
racy, chs. 3, 6; Roberto Unger, False Necessity (Cambridge: Cambridge University Press, 1987),
ch. 5.
4. In the discussion that follows, I draw on John Rawls, A Theory of Justice (Cambridge, MA:
Harvard University Press, 1971), esp. sections 36, 37, 43, 54; John Rawls, “The Basic Liberties
and Their Priority,” Tanner Lectures on Human Values (Salt Lake City: University of Utah
Press, 1982). [The Tanner Lectures were reprinted as Lecture 8 in John Rawls, Political Liberal-
ism (New York: Columbia University Press, 1996). I have kept the references to the original
here.]
17
P H I L O S O P H Y , P O L I T I C S , D E M O C R A C Y
scheme, in which democratic politics consists of fair bargaining among

groups each of which pursues its particular or sectional interest, is un-
suited to a just society.5 Citizens and parties operating in the political
arena ought not to “take a narrow or group-interested standpoint.”6 And
parties should only be responsive to demands that are “argued for openly
by reference to a conception of the public good.”7 Public explanations
and justifications of laws and policies are to be cast in terms of concep-
tions of the common good (conceptions that, on Rawls’s view, must be
consistent with the two principles of justice), and public deliberation
should aim to work out the details of such conceptions and to apply them
to particular issues of public policy.8
Second, the ideal of democratic order has egalitarian implications that
must be satisfied in ways that are manifest to citizens. The reason is that
in a just society political opportunities and powers must be independent
of economic or social position—the political liberties must have a fair
value9—and the fact that they are independent must be more or less evi-
dent to citizens. Ensuring this manifestly fair value might, for example,
require public funding of political parties and restrictions on private po-
litical spending, as well as progressive tax measures that serve to limit
inequalities of wealth and to ensure that the political agenda is not
controlled by the interests of economically and socially dominant
groups.10 In principle, these distributional requirements might be more
stringently egalitarian than those fixed by the difference principle.11 This
is so in part because the main point of these measures is not simply to en-
sure that democratic politics proceeds under fair conditions, nor only to
encourage just legislation, but also to ensure that the equality of citizens
5. Rawls, Theory of Justice, 360–361. [The page references to A Theory of Justice in essays 1–7
are to the 1971 edition. For a conversion table to page numbers in the revised, 1999 edition, see
John Rawls, A Theory of Justice, rev. ed. (Cambridge, MA: Harvard University Press, 1999), 517–
519.] This rejection is not particularly idiosyncratic. Sunstein, for example, argues that ideal
pluralism has never been embraced as a political ideal in American public law. See his “Naked
Preferences” and “Interest Groups.”
6. Rawls, Theory, 360.
7. Ibid., 226, 472.
8. Ibid., 362.
9. Officially, the requirement of fair value is that “everyone has a fair opportunity to hold
public office and to influence the outcome of political decisions.” Rawls, “Basic Liberties,” 42.
10. Rawls, Theory, 225–226, 277–278; “Basic Liberties,” 42–43.
11. “Basic Liberties,” 43. Whatever their stringency, these distributional requirements take
priority over the difference principle, since the requirement of fair value is part of the principle
of liberty; that is, the first principle of justice. See ibid., 41–42.
18
is manifest and to declare a commitment to that equality “as the public

intention.”12
Third, democratic politics should be ordered in ways that provide a ba-
sis for self-respect, that encourage the development of a sense of political
competence, and that contribute to the formation of a sense of justice;13
it should fix “the foundations for civic friendship and [shape] the ethos of
political culture.”14 Thus the importance of democratic order is not con-
fined to its role in obstructing the class legislation that can be expected
from systems in which groups are effectively excluded from the channels
of political representation and bargaining. In addition, democratic poli-
tics should also shape the ways in which the members of the society un-
derstand themselves and their own legitimate interests.
When properly conducted, then, democratic politics involves public
deliberation focused on the common good, requires some form of manifest
equality among citizens, and shapes the identity and interests of citizens
in ways that contribute to the formation of a public conception of com-
mon good. How does the ideal of a fair system of social cooperation pro-
vide a way to account for the attractiveness and importance of these three
features of the deliberative democratic ideal? Rawls suggests a formal and
an informal line of argument. The formal argument is that parties in the
original position would choose the principle of participation15 with the
proviso that the political liberties have their fair value. The three condi-
tions are important because they must be satisfied if constitutional ar-
rangements are to ensure participation rights, guarantee a fair value to
those rights, and plausibly produce legislation that encourages a fair dis-
tribution according to the difference principle.
Rawls also suggests an informal argument for the ordering of political
institutions, and I shall focus on this informal argument here: “Justice as
fairness begins with the idea that where common principles are necessary
and to everyone’s advantage, they are to be worked out from the view-
point of a suitably defined initial situation of equality in which each per-
son is fairly represented. The principle of participation transfers this no-

13. The importance of democratic politics in the account of the acquisition of the sense of
justice is underscored in ibid., 473–474.
14. Ibid., 234.
15. The principle of participation states that “all citizens are to have an equal right to take
part in, and to determine the outcome of, the constitutional process that establishes the laws
with which they are to comply.” Ibid., 221.
19
tion from the original position to the constitution . . . [thus] preserv[ing]

the equal representation of the original position to the degree that is fea-
sible.”16 Or, as he puts it elsewhere: “The idea [of the fair value of politi-
cal liberty] is to incorporate into the basic structure of society an effective
political procedure which mirrors in that structure the fair representation
of persons achieved by the original position.”17 The suggestion is that,
since we accept the intuitive ideal of a fair system of cooperation, we
should want our political institutions themselves to conform, insofar as it
is feasible, to the requirement that terms of association be worked out un-
der fair conditions. And so we arrive directly at the requirement of equal
liberties with fair value, rather than arriving at it indirectly, through a hy-
pothetical choice of that requirement under fair conditions. In this infor-
mal argument, the original position serves as an abstract model of what
fair conditions are, and of what we should strive to mirror in our political
institutions, rather than as an initial-choice situation in which regulative
principles for those institutions are selected.
I think that Rawls is right in wanting to accommodate the three condi-
tions. What I find less plausible is that the three conditions are natural
consequences of the ideal of fairness. Taking the notion of fairness as fun-
damental, and aiming (as in the informal argument) to model political
arrangements on the original position, it is not clear why, for example,
political debate ought to be focused on the common good, or why the
manifest equality of citizens is an important feature of a democratic asso-
ciation. The pluralist conception of democratic politics as a system of
bargaining with fair representation for all groups seems an equally good
mirror of the ideal of fairness.
The response to this objection is clear enough: the connection be-
tween the ideal of fairness and the three features of democratic politics
depends on psychological and sociological assumptions. Those features
do not follow directly from the ideal of a fair system of cooperation, or
from that ideal as it is modeled in the original position. Rather, we arrive
at them when we consider what is required to preserve fair arrange-
ments and to achieve fair outcomes. For example, public political debate
should be conducted in terms of considerations of the common good be-
cause we cannot expect outcomes that advance the common good unless
16. Ibid., 221–222. I assume that the principle of participation should be understood here to
include the requirement of the fair value of political liberty.
17. Rawls, “Basic Liberties,” 45; emphasis added.
20
people are looking for them. Even an ideal pluralist scheme, with equal
bargaining power and no barriers to entry, cannot reasonably be expected
to advance the common good as defined by the difference principle.18
But this is, I think, too indirect and instrumental an argument for the
three conditions. Like utilitarian defenses of liberty, it rests on a series of
highly speculative sociological and psychological judgments. I want to
suggest that the reason why the three are attractive is not that an order
with, for example, no explicit deliberation about the common good and
no manifest equality would be unfair (though of course it might be). In-
stead it is that they comprise elements of an independent and expressly
political ideal that is focused in the first instance19 on the appropriate
conduct of public affairs—on, that is, the appropriate ways of arriving at
collective decisions. And to understand that ideal we ought not to pro-
ceed by seeking to “mirror” ideal fairness in the fairness of political ar-
rangements, but instead to proceed by seeking to mirror a system of ideal
deliberation in social and political institutions. I want now to turn to this
alternative.
2
The notion of a deliberative democracy is rooted in the intuitive ideal of
a democratic association in which the justification of the terms and con-
ditions of association proceeds through public argument and reasoning
among equal citizens.20 Citizens in such an order share a commitment to
the resolution of problems of collective choice through public reasoning
and regard their basic institutions as legitimate insofar as they establish
the framework for free public deliberation. To elaborate this ideal, I be-
19. The reasons for the phrase “in the first instance” are clarified below at 29.
20. Since writing the first draft of this section of the essay, I have read Jon Elster, “The Mar-
ket and the Forum: Three Varieties of Political Theory,” in Jon Elster and Aanund Hylland,
eds., Foundations of Social Choice Theory (Cambridge: Cambridge University Press, 1986),
103–132; and Bernard Manin, “On Legitimacy and Political Deliberation,” Political Theory 15
(1987): 338–368, which both present parallel conceptions. This is especially so with Elster’s
treatment of the psychology of public deliberation (112–113). I am indebted to Alan Hamlin for
bringing the Elster article to my attention. The overlap is explained by the fact that Elster,
Manin, and I all draw on Jürgen Habermas. See Jürgen Habermas, Legitimation Crisis, trans.
Thomas McCarthy (Boston: Beacon Press, 1975); idem, Communication and the Evolution of
Society, trans. Thomas McCarthy (Boston: Beacon Press, 1979); and idem, The Theory of Com-
municative Action, vol. 1, trans. Thomas McCarthy (Boston: Beacon Press, 1984). I have also
found very helpful the discussion of the contractualist account of motivation in T. M. Scanlon,
“Contractualism and Utilitarianism,” in Amartya Sen and Bernard Williams, eds., Utilitarian-
ism and Beyond (Cambridge: Cambridge University Press, 1982), 103–128.
21
gin with a more explicit account of the ideal itself, presenting what I shall
call the “formal conception” of deliberative democracy. Proceeding from
this formal conception, I pursue a more substantive account of delibera-
tive democracy by presenting an account of an ideal deliberative proce-
dure that captures the notion of justification through public argument
and reasoning among equal citizens and serves in turn as a model for de-
liberative institutions.
The formal conception of a deliberative democracy has five main
features:
D1 A deliberative democracy is an ongoing and independent associa-

tion, whose members expect it to continue into the indefinite fu-
ture.
D2 The members of the association share (and it is common knowl-
edge that they share) the view that the appropriate terms of associ-
ation provide a framework for or are the results of their delibera-
tion. They share, that is, a commitment to coordinating their
activities within institutions that make deliberation possible and
according to norms that they arrive at through their deliberation.
For them, free deliberation among equals is the basis of legiti-
macy.
D3 A deliberative democracy is a pluralistic association. The members
have diverse preferences, convictions, and ideals concerning the
conduct of their own lives. While sharing a commitment to the
deliberative resolution of problems of collective choice (D2), they
also have divergent aims, and do not think that some particular set
of preferences, convictions, or ideals is mandatory.
D4 Because the members of a democratic association regard delibera-
tive procedures as the source of legitimacy, it is important to them
that the terms of their association not merely be the results of their
deliberation but also be manifest to them as such.21 They prefer in-
stitutions in which the connections between deliberation and out-
comes are evident to ones in which the connections are less clear.
D5 The members recognize one another as having deliberative capac-
21. For philosophical discussions of the importance of manifestness or publicity, see Imman-
uel Kant, “Toward Perpetual Peace: A Philosophical Essay,” in Perpetual Peace and Other Es-
says, trans. Thomas Humphrey (Indianapolis: Hackett, 1983), 135–139; Rawls, Theory, 133 and
section 29; Bernard Williams, Ethics and the Limits of Philosophy (Cambridge, MA: Harvard
University Press, 1985), 101–102, 200.
22
ities, i.e., the capacities required for entering into a public ex-
change of reasons and for acting on the results of such public rea-
soning.
A theory of deliberative democracy aims to give substance to this for-

mal ideal by characterizing the conditions that should obtain if the social
order is to be manifestly regulated by deliberative forms of collective
choice. I propose to sketch a view of this sort by considering an ideal
scheme of deliberation, which I shall call the “ideal deliberative proce-
dure.” The aim in sketching this procedure is to give an explicit state-
ment of the conditions for deliberative decision-making that are suited to
the formal conception, and thereby to highlight the properties that dem-
ocratic institutions should embody, so far as possible. I should emphasize
that the ideal deliberative procedure is meant to provide a model for in-
stitutions to mirror—in the first instance for the institutions in which col-
lective choices are made and social outcomes publicly justified—and not
to characterize an initial situation in which the terms of association
themselves are chosen.22
Turning then to the ideal procedure, there are three general aspects of
deliberation. There is a need to decide on an agenda, to propose alterna-
tive solutions to the problems on the agenda, supporting those solutions
with reasons, and to conclude by settling on an alternative. A democratic
conception can be represented in terms of the requirements that it sets
on such a procedure. In particular, outcomes are democratically legiti-
mate if and only if they could be the object of a free and reasoned agree-
ment among equals. The ideal deliberative procedure is a procedure that
captures this principle.23
I1 Ideal deliberation is free in that it satisfies two conditions. First, the

participants regard themselves as bound only by the results of their
deliberation and by the preconditions for that deliberation. Their
consideration of proposals is not constrained by the authority of
22. The distinction between the ideal procedure and an initial-choice situation will be im-
portant in the later discussion of motivation formation (see 26–27) and institutions (34–36).
23. There are of course norms and requirements on individuals that do not have deliberative
justification. The conception of deliberative democracy is, in Rawls’s term, a “political concep-
tion,” and not a comprehensive moral theory. On the distinction between political and com-
prehensive theories, see John Rawls, “The Idea of an Overlapping Consensus,” Oxford Journal
of Legal Studies 7 (1987): 1–25.
23
prior norms or requirements. Second, the participants suppose

that they can act from the results, taking the fact that a certain de-
cision is arrived at through their deliberation as a sufficient reason
for complying with it.
I2 Deliberation is reasoned in that the parties of it are required to
state their reasons for advancing proposals, supporting them, or
criticizing them. They give reasons with the expectation that those
reasons (and not, for example, their power) will settle the fate of
their proposal. In ideal deliberation, as Habermas puts it, “no
force except that of the better argument is exercised.”24 Reasons
are offered with the aim of bringing others to accept the proposal,
given their disparate ends (D3) and their commitment (D2) to set-
tling the conditions of their association through free deliberation
among equals. Proposals may be rejected because they are not de-
fended with acceptable reasons, even if they could be so de-
fended. The deliberative conception emphasizes that collective
choices should be made in a deliberative way, and not only that
those choices should have a desirable fit with the preferences of
citizens.
I3 In ideal deliberation parties are both formally and substantively
equal. They are formally equal in that the rules regulating the pro-
cedure do not single out individuals. Everyone with the delibera-
tive capacities has equal standing at each stage of the deliberative
process. Each can put issues on the agenda, propose solutions, and
offer reasons in support of or in criticism of proposals. And each
has an equal voice in the decision. The participants are substan-
tively equal in that the existing distribution of power and resources
does not shape their chances to contribute to deliberation, nor
does that distribution play an authoritative role in the deliberation.
The participants in the deliberative procedure do not regard them-
selves as bound by the existing system of rights, except insofar as
that system establishes the framework of free deliberation among
equals. Instead they regard that system as a potential object of their
deliberative judgment.
I4 Finally, ideal deliberation aims to arrive at a rationally motivated
consensus—to find reasons that are persuasive to all who are com-
24. Habermas, Legitimation Crisis, 108.
24
mitted to acting on the results of a free and reasoned assessment of

alternatives by equals. Even under ideal conditions there is no
promise that consensual reasons will be forthcoming. If they are
not, then deliberation concludes with voting, subject to some form
of majority rule.25 The fact that it may so conclude does not, how-
ever, eliminate the distinction between deliberative forms of col-
lective choice and forms that aggregate non-deliberative prefer-
ences. The institutional consequences are likely to be different in
the two cases, and the results of voting among those who are com-
mitted to finding reasons that are persuasive to all are likely to dif-
fer from the results of an aggregation that proceeds in an absence
of this commitment.
Drawing on this characterization of ideal deliberation, can we say any-

thing more substantive about a deliberative democracy? What are the
implications of a commitment to deliberative decisions for the terms of
social association? In the remarks that follow I shall indicate the ways that
this commitment carries with it a commitment to advance the common
good and to respect individual autonomy.
common good a nd a u t onomy

Consider first the notion of the common good. Since the aim of ideal de-
liberation is to secure agreement among all who are committed to free
deliberation among equals, and the condition of pluralism obtains (D3),
the focus of deliberation is on ways of advancing the aims of each party to
it. While no one is indifferent to his/her own good, everyone also seeks to
arrive at decisions that are acceptable to all who share the commitment
to deliberation (D2). (As we shall see just below, taking that commitment
seriously is likely to require a willingness to revise one’s own preferences
and convictions.) Thus the characterization of an ideal deliberation pro-
cedure links the formal notion of deliberative democracy with the more
substantive ideal of a democratic association in which public debate is fo-
cused on the common good of the members.
Of course, talk about the common good is one thing; sincere efforts to
advance it are another. While public deliberation may be organized
around appeals to the common good, is there any reason to think that
25. For criticism of the reliance on an assumption of unanimity in deliberative views, see
Manin, “On Legitimacy,” 359–361.
25
even ideal deliberation would not consist in efforts to disguise personal or

class advantage as the common advantage? There are two responses to
this question. The first is that in my account of the formal idea of a delib-
erative democracy, I stipulated (D2) that the members of the association
are committed to resolving their differences through deliberation, and
thus to providing reasons that they sincerely expect to be persuasive to
others who share that commitment. In short, this stipulation rules out the
problem. Presumably, however, the objection is best understood as di-
rected against the plausibility of realizing a deliberative procedure that
conforms to the ideal, and thus is not answerable through stipulations.
The second response, then, rests on a claim about the effects of delib-
eration on the motivations of deliberators.26 A consequence of the reason-
ableness of the deliberative procedure (I2) together with the condition of
pluralism (D3) is that the mere fact of having a preference, a conviction,
or an ideal does not by itself provide a reason in support of a proposal.
While I may take my preferences as a sufficient reason for advancing a
proposal, deliberation under conditions of pluralism requires that I find
reasons that make the proposal acceptable to others who cannot be ex-
pected to regard my preferences as sufficient reasons for agreeing. The
motivational thesis is that the need to advance reasons that persuade oth-
ers will help to shape the motivations that people bring to the delibera-
tive procedure in two ways. First, the practice of presenting reasons will
contribute to the formation of a commitment to the deliberative resolu-
tions of political questions (D2). Given that commitment, the likelihood
of a sincere representation of preferences and convictions should in-
crease, while the likelihood of their strategic misrepresentation declines.
Second, it will shape the content of preferences and convictions as well.
Assuming a commitment to deliberative justification, the discovery that I
can offer no persuasive reasons on behalf of a proposal of mine may trans-
form the preferences that motivate the proposal. Aims that I recognize to
be inconsistent with the requirements of deliberative agreement may
tend to lose their force, at least when I expect others to be proceeding in
reasonable ways and expect the outcome of deliberation to regulate sub-
sequent action.
Consider, for example, the desire to be wealthier come what may. I
26. Note the parallel with Elster, “The Market and the Forum,” indicated in note 20. See
also the discussion in Habermas, Legitimation Crisis, 108, about “needs that can be communi-
catively shared,” and Habermas, Communication and the Evolution of Society, ch. 2.
26
cannot appeal to this desire itself in defending policies. The motivational

claim is the need to find an independent justification that does not ap-
peal to this desire and will tend to shape it into, for example, a desire to
have a level of wealth that is consistent with a level that others (i.e., equal
citizens) find acceptable. I am of course assuming that the deliberation is
known to be regulative, and that the wealth cannot be protected through
wholly non-deliberative means.
Deliberation, then, focuses debate on the common good. And the rele-
vant conceptions of the common good are not comprised simply of inter-
ests and preferences that are antecedent to deliberation. Instead, the in-
terests, aims, and ideals that constitute the common good are those that
survive deliberation, interests that, on public reflection, we think it legiti-
mate to appeal to in making claims on social resources. Thus the first and
third of the features of deliberative democracy that I mentioned in dis-
cussion of Rawls (17–19 above) provide central elements in the delibera-
tive conception.
The ideal deliberation scheme also indicates the importance of auton-
omy in a deliberative democracy. In particular, it is responsive to two
main threats to autonomy. As a general matter, actions fail to be autono-
mous if the preferences on which an agent acts are, roughly, given by the
circumstances, and not determined by the agent. There are two para-
digm cases of “external” determination. The first is what Elster has called
“adaptive preferences.”27 These are preferences that shift with changes in
the circumstances of the agent without any deliberate contribution by
the agent to that shift. This is true, for example, of the political prefer-
ences of instinctive centrists who move to the median position in the po-
litical distribution, wherever it happens to be. The second I shall call
“accommodationist preferences.” While they are deliberately formed,
accommodationist preferences represent psychological adjustments to
conditions of subordination in which individuals are not recognized as
having the capacity for self-government. Consider Stoic slaves, who de-
liberately shape their desires to match their powers, with a view of mini-
mizing frustration. Since the existing relations of power make slavery the
only possibility, they cultivate desires to be slaves, and then act on those
desires. While their motives are deliberately formed, and they act on
27. Jon Elster, “Sour Grapes,” in Sen and Williams, eds., Utilitarianism and Beyond, 219–
238. For an interesting discussion of autonomous preferences and political processes, see
Sunstein, “Legal Interference,” 1145–1158, and “Naked Preferences,” 1699–1700.
27
their desires, the Stoic slaves do not act autonomously when they seek to
be good slaves. The absence of alternatives and consequent denial of
scope for deliberative capacities that defines the condition of slaves sup-
ports the conclusion that their desires result from their circumstances,
even though those circumstances shape the desires of the Stoic slaves
through their deliberation.
There are then at least two dimensions of autonomy. The phenome-
non of adaptive preferences underlines the importance of conditions
that permit and encourage the deliberative formation of preferences; the
phenomenon of accommodationist preferences indicates the need for fa-
vorable conditions for the exercise of the deliberative capacities. Both
concerns are met when institutions for collective decision-making are
modeled on the ideal deliberative procedure. Relations of power and
subordination are neutralized (I1, I3, I4), and each is recognized as
having the deliberative capacities (D5), thus addressing the problem of
accommodationist preferences. Further, the requirement of reasonable-
ness discourages adaptive preferences (I2). While preferences are
“formed” by the deliberative procedure, this type of preferences forma-
tion is consistent with autonomy, since preferences that are shaped by
public deliberation are not simply given by external circumstances. In-
stead, they are the result of “the power of reason as applied through pub-
lic discussion.”28
Beginning, then, from the formal ideal of a deliberative democracy, we
arrive at the more substantive ideal of an association that is regulated by
deliberation aimed at the common good and that respects the autonomy
of the members. And so, in seeking to embody the ideal deliberative pro-
cedure in institutions, we seek, inter alia, to design institutions that focus
on political debates on the common good, that shape the identity and in-
terests of citizens in ways that contribute to an attachment to the com-
mon good, and that provide the favorable conditions for the exercise of
deliberative powers that are required for autonomy.
3
I want now to shift the focus. While I shall continue to pursue the rela-
tionship between the ideal procedure and more substantive issues about
deliberative democratic association, I want to do so by considering four
28. Whitney v. California, 274 U.S. 357 (1927).
28
natural objections to the conception I have been discussing, objections to

that conception for being sectarian, incoherent, unjust, and irrelevant.
My aim is not to provide a detailed response to the objections, but to
clarify the conception of deliberative democracy by sketching the lines
along which a response should proceed. Before turning to the objections,
I enter two remarks about what follows.
First, as I indicated earlier, a central aim in the deliberative conception
is to specify the institutional preconditions for deliberative decision-mak-
ing. The role of the ideal deliberative procedure is to provide an abstract
characterization of the important properties of deliberative institutions.
The role of the ideal deliberative procedure is thus different from the
role of an ideal social contract. The ideal deliberative procedure pro-
vides a model for institutions, a model that they should mirror, so far
as possible. It is not a choice situation in which institutional principles
are selected. The key point about the institutional reflection is that it
should make deliberation possible. Institutions in a deliberative democ-
racy do not serve simply to implement the results of deliberation, as
though free deliberation could proceed in the absence of appropriate in-
stitutions. Neither the commitment to nor the capacity for arriving at
deliberative decisions is something that we can simply assume to obtain
independent from the proper ordering of institutions. The institutions
themselves must provide the framework for the formation of the will;
they determine whether there is equality, whether deliberation is free
and reasoned, whether there is autonomy, and so on.
Second, I shall be focusing here on some requirements for “public” in-
stitutions that reflect the ideal of deliberative resolution. But there is, of
course, no reason to expect as a general matter that the preconditions for
deliberation will respect familiar institutional boundaries between “pri-
vate” and “public” and will all pertain to the public arena. For example,
inequalities of wealth, or the absence of institutional measures designed
to redress the consequences of those inequalities, can serve to undermine
the equality required in deliberative arenas themselves. And so a more
complete treatment would need to address a wider range of institutional
issues.29
29. See Cohen and Rogers, On Democracy, chs. 3, 6; Joshua Cohen, “The Economic Basis
of Deliberative Democracy,” Social Philosophy and Policy 6, 2 (1988): 25–50.
29
secta ria nism

The first objection is that the ideal of deliberative democracy is objec-
tionably sectarian because it depends on a particular view of the good
life—an ideal of active citizenship. What makes it sectarian is not the
specific ideal on which it depends but the (alleged) fact that it depends
on some specific conception at all. I do not think that the conception of
deliberative democracy suffers from the alleged difficulty. In explaining
why not, I shall put to the side current controversy about the thesis that
sectarianism is avoidable and objectionable and assume that it is both.30
Views of the good figure in political conceptions in at least two ways.
First, the justification of some conceptions appeals to a notion of the hu-
man good. Aristotelian views, for example, endorse the claim that the ex-
ercise of the deliberative capacities is a fundamental component of a
good human life and conclude that a political association ought to be or-
ganized to encourage the realization of those capacities by its members.
A second way in which conceptions of the good enter is that the stability
of a society may require widespread allegiance to a specific conception of
the good, even though its institutions can be justified without appeal to
that conception. For example, a social order that can be justified without
reference to ideals of national allegiance may nonetheless require wide-
spread endorsement of the ideal of patriotic devotion for its stability.
A political conception is objectionably sectarian only if its justification
depends on a particular view of the human good, not if its stability is
contingent on widespread agreement on the value of certain activities
and aspirations. For this reason the democratic conception is not sectar-
ian. It is organized around a view of political justification—that justifica-
tion proceeds through free deliberation among equal citizens—and not
around a conception of the proper conduct of life. So, while it is plausi-
ble that the stability of a deliberative democracy depends on encouraging
the ideal of active citizenship, this dependence does not suffice to show
that it is objectionably sectarian.
30. For contrasting views on sectarianism, see Rawls, “The Idea of an Overlapping Consen-
sus”; Ronald Dworkin, A Matter of Principle (Cambridge, MA: Harvard University Press, 1985),
part 3; Alisdair MacIntyre, After Virtue (Notre Dame: University of Notre Dame Press, 1981);
Michael Sandel, Liberalism and the Limits of Justice (Cambridge: Cambridge University Press,
1982).
30
incoherence
Consider next the putative incoherence of the ideal. We find this charge
in an important tradition of argument, including Schumpeter’s Capital-
ism, Socialism, and Democracy and, more recently, William Riker’s work
on social choice and democracy. I want here to say a word about the lat-
ter, focusing on just one reason that Riker gives for thinking that the ideal
of popular self-government is incoherent.31
Institutionalizing a deliberative procedure requires a decision rule
short of consensus—for example, majority rule. But majority rule is glob-
ally unstable: as a general matter, there exists a majority-rule path leading
from any element in the set of alternatives to any other element in the
set. The majority, standing in for the people, wills everything and there-
fore wills nothing. Of course, while anything can be the result of the ma-
jority decision, it is not true that everything will be the result. But, be-
cause majority rule is so unstable, the actual decision of the majority will
not be determined by preferences themselves, since they do not constrain
the outcome. Instead decisions will reflect the particular institutional
constraints under which they are made. But these constraints are “exoge-
nous to the world of tastes and values.”32 So the ideal of popular self-gov-
ernment is incoherent because we are, so to speak, government by the in-
stitutions, and not by ourselves.
I want to suggest one difficulty with this argument that highlights the
structure of the deliberative conception. According to the argument I just
sketched, outcomes in majority-rule institutions reflect “exogenous” in-
stitutional constraints and not underlying preferences. This suggests that
we can identify the preferences and convictions that are relevant to
collective choices apart from the institutions through which they are
formed and expressed. But that is just what the deliberative conception
denies. On this conception, the relevant preferences and convictions are
those that could be expressed in free deliberation, and not those that are
prior to it. For this reason, popular self-government premises the exis-
tence of institutions that provide a framework for deliberation; these ar-
31. See William Riker, Liberalism Against Populism: A Confrontation between the Theory of
Democracy and the Theory of Social Choice (San Francisco: W. H. Freeman, 1982); for discus-
sion of Riker’s view see Jules Coleman and John Ferejohn, “Democracy and Social Choice,”
Ethics 97 (1986): 6–25; Joshua Cohen, “An Epistemic Conception of Democracy,” Ethics 97
(1986): 26–38.
32. Riker, Liberalism Against Populism, 190.
31
rangements are not “exogenous constraints” on the aggregation of prefer-

ences but instead help to shape their content and the way that citizens
choose to advance them. And, once the deliberative institutions are in
place, and preferences, convictions, and political actions are shaped by
them, it is not clear that instability problems remain so severe as to
support the conclusion that self-government is an empty and incoher-
ent ideal.
injust ice
The third problem concerns injustice. I have been treating the ideal of
democracy as the basic ideal for a political conception. But it might be
argued that the ideal of democracy is not suited to the role of funda-
mental political ideal because its treatment of basic liberties is mani-
festly unacceptable. It makes those liberties dependent on judgments
of majorities and thus endorses the democratic legitimacy of decisions
that restrict the basic liberties of individuals. In responding to this objec-
tion, I shall focus on the liberty of expression33 and shall begin by filling
out a version of the objection, which I put in the words of an imagined
critic.34
“You embrace the ideal of a democratic order. The aim of a demo-
cratic order is to maximize the power of the people to secure its wants. To
defend the liberty of expression you will argue that that power is dimin-
ished if the people lack the information required for exercising their will.
Since expression provides information, you will conclude that abridge-
ments of expression ought to be barred. The problem with your argu-
ment is that preventing restrictions on expression also restricts the power
33. For discussion of the connection between ideals of democracy and freedom of expres-
sion, see Alexander Meiklejohn, Free Speech and Its Relation of Self-Government (New York:
Harper and Row, 1948), and John Hart Ely, Democracy and Distrust (Cambridge, MA: Harvard
University Press, 1980), 93–94, 105–116). Freedom of expression is a special case that can per-
haps be more straightforwardly accommodated by the democratic conception than liberties of
conscience, or the liberties associated with privacy and personhood. I do think, however, that
these other liberties can be given satisfactory treatment by the democratic conception, and
would reject it if I did not think so. The general idea would be to argue that other fundamental
liberties must be protected if citizens are to be able to engage in and have equal standing in po-
litical deliberation without fear that such engagement puts them at risk for their convictions or
personal choices. Whether this line of argument will work out on the details is a matter for
treatment elsewhere.
34. This objection is suggested by Dworkin, “The Forum of Principle,” in A Matter of Prin-
ciple, 61–63. He cites the following passage from a letter of Madison’s: “And a people who
mean to be their own Governors, must arm themselves with the power which knowledge gives”
(emphasis added).
32
of the people, since the citizens may collectively prefer such restric-
tions. And so it is not at all clear as a general matter that the protection
of expression will maximize popular power. So while you will, of course,
not want to prevent everyone from speaking all the time, you cannot
defend the claim that there is even a presumption in favor of the protec-
tion of expression. And this disregard for fundamental liberties is unac-
ceptable.”
This objection has force against some conceptions in which democ-
racy is a fundamental ideal, particularly those in which the value of ex-
pression turns exclusively on its role as a source of information about
how best to advance popular ends. But it does not have any force against
the deliberative conception, since the latter does not make the case for
expression turn on its role in maximizing the power of the people to se-
cure its wants. That case rests instead on a conception of collective
choice, in particular on a view about how the “wants” that are relevant to
collective choice are formed and defined in the first place. The relevant
preferences and convictions are those that arise or are confirmed through
deliberation. And a framework of free expression is required for the rea-
soned consideration of alternatives that comprises deliberation. The de-
liberative conception holds that free expression is required for determin-
ing what advances the common good, because what is good is fixed by
public determination, and not prior to it. It is fixed by informed and au-
tonomous judgments, involving the exercise of the deliberative capaci-
ties. So the ideal of deliberative democracy is not hostile to free expres-
sion; rather, it presupposes such freedom.
But what about expression with no direct bearing on issues of public
policy? Is the conception of deliberative democracy committed to treat-
ing all “non-political expression” as second-class and as meriting lesser
protection? I do not think so. The deliberative conception construes poli-
tics as aiming in part at the formation of preferences and convictions, not
just at their articulation and aggregation. Because of this emphasis on
reasoning about preferences and convictions, and the bearing of expres-
sion with no political focus on such reasoning, the deliberative view
draws no bright line between political speech and other sorts of expres-
sion. Forms of expression that do not address issues of policy may well
bear on the formation of interests, aims, and ideals that citizens bring to
public deliberation. For this reason, the deliberative conception supports
protection for the full range of expression, regardless of the content of
33
that expression.35 It would violate the core of the ideal of free deliberation
among equals to fix preferences and convictions in advance by restricting
the content of expression, or by barring access to expression, or by pre-
venting the expression that is essential to having convictions at all. Thus
the injustice objection fails because the liberties are not simply among
the topics for deliberation; they help to provide the framework that
makes it possible.36
irrel eva nce

The irrelevance objection is that the notion of public deliberation is ir-
relevant to modern political conditions.37 This is the most important ob-
jection and the one about which it is hardest to say anything at the level
of generality required by the present context. Here again I shall confine
myself to one version of the objection, though one that I take to be repre-
sentative.
The version that I want to consider starts from the assumption that a di-
rect democracy with citizens gathering in legislative assemblies is the
only way to institutionalize a deliberative procedure. Premising that, and
recognizing that direct democracy is impossible under modern condi-
tions, the objection concludes that we ought to be led to reject the ideal
because it is not relevant to our circumstances.
The claim about the impossibility of direct democracy is plainly cor-
rect. But I see no merit in the claim that direct democracy is the uniquely
suitable way to institutionalize the ideal procedure.38 In fact, in the ab-
sence of a theory about the operation of democratic assemblies—a theory
which cannot simply stipulate that ideal conditions obtain—there is no
reason to be confident that a direct democracy would subject political
35. On the distinction between content-based and content-neutral abridgements, the com-
plexities of drawing the distinction in particular cases, and the special reasons for hostility to
content-based abridgements, see Laurence Tribe, American Constitutional Law (Mineola, NY:
Foundation Press, 1978), 584–682; Geoffrey Stone, “Content-neutral Restrictions,” University
of Chicago Law Review 54 (1987): 46–118.
36. I am not suggesting that the deliberative view provides the only sound justification for
the liberty of expression. My concern here is rather to show that the deliberative view is capa-
ble of accommodating it.
37. For an especially sharp statement of the irrelevance objection, see Carl Schmitt, The
Crisis of Parliamentary Democracy, trans. Ellen Kennedy (Cambridge, MA: MIT Press, 1985).
38. This view is sometimes associated with Rousseau, who is said to have conflated the no-
tion of democratic legitimacy with the institutional expression of that ideal in a direct democ-
racy. For criticism of this interpretation, see Joshua Cohen, “Autonomy and Democracy:
Reflections on Rousseau,” Philosophy and Public Affairs 15, 3 (Summer 1986): 275–297.
34
questions to deliberative resolution, even if a direct democracy were a

genuine institutional possibility.39 In the absence of a realistic account of
the functioning of citizen assemblies, we cannot simply assume that large
gatherings with open-ended agendas will yield any deliberation at all or
that they will encourage participants to regard one another as equals in a
free deliberative procedure. The appropriate ordering of deliberative in-
stitutions depends on issues of political psychology and political behav-
ior; it is not an immediate consequence of the deliberative ideal. So, far
from being the only deliberative scheme, direct democracy may not even
be a particularly good arrangement for deliberation. But, once we reject
the idea that a direct democracy is the natural or necessary form of ex-
pression of the deliberative ideal, the straightforward argument for irrele-
vance no longer works. In saying how the ideal might be relevant, how-
ever, we come up against the problem I mentioned earlier. Lacking a
good understanding of the workings of institutions, we are inevitably
thrown back on more or less speculative judgments. What follows are
some sketchy remarks on one issue that should be taken in this spirit.
At the heart of the institutionalization of the deliberative procedure is
the existence of arenas in which citizens can propose issues for the politi-
cal agenda and participate in debate about those issues. The existence of
such arenas is a public good and ought to be supported with public
money. This is not because public support is the only way, or even the
most efficient way, of ensuring the provision of such arenas. Instead, pub-
lic provision expresses the basic commitment of a democratic order to
the resolution of political questions through free deliberation among
equals. The problem is to figure out how arenas might be organized to
encourage such deliberation.
In considering that organization, there are two key points that I want to
underscore. The first is that material inequalities are an important source
of political inequalities. The second point—which is more speculative—
is that deliberative arenas that are organized exclusively on local, sec-
tional, or issue-specific lines are unlikely to produce the open-ended de-
39. Madison urges this point in the Federalist Papers. Objecting to a proposal advanced by
Jefferson, which would have regularly referred constitutional questions “to the decision of the
whole society,” Madison argues that this would increase “the danger of disturbing the public
tranquility by interesting too strongly the public passions.” And “it is the reason, alone, of the
public that ought to control and regulate the government . . . [while] the passions ought to be
controlled and regulated by the government.” I endorse the form of the objection, not its con-
tent. Federalist Papers, ed. Clinton Rossiter (New York: New American Library, 1961), 315–317.
35
liberation required to institutionalize a deliberative procedure. Since

these arenas bring together only a narrow range of interests, deliberation
in them can be expected at best to produce coherent sectional interests,
not a more comprehensive conception of the common good.
These two considerations together provide support for the view that po-
litical parties supported by public funds play an important role in making
a deliberative democracy possible.40 There are two reasons for this, corre-
sponding to the two considerations I have just mentioned. In the first
place, an important feature of organizations generally, and parties in par-
ticular, is that they provide a means through which individuals and
groups who lack the “natural” advantage of wealth can overcome the
political disadvantages that follow on that lack. Thus they can help to
overcome the inequalities in deliberative arenas that result from material
inequality. Of course, to play this role, political organizations must them-
selves be freed from the dominance of private resources, and that inde-
pendence must be manifest. Thus the need for public funding. Here we
arrive back at the second point that I mentioned in the discussion of
Rawls’s view—that measures are needed to ensure manifest equality—
though now as a way of displaying a shared commitment to fairness. Sec-
ond, because parties are required to address a comprehensive range of
political issues, they provide arenas in which debate is not restricted in
the ways that it is in local, sectional, or issue-specific organizations. They
can provide the more open-ended arenas needed to form and articulate
the conceptions of the common good that provide the focus of political
debate in a deliberative democracy.
There is certainly no guarantee that parties will operate as I have just
described. But this is not especially troubling, since there are no guaran-
tees of anything in politics. The question is how we can best approximate
the deliberative conception. And it is difficult to see how that is possible
in the absence of strong parties supported with public resources (though,
of course, a wide range of other conditions is required as well).
40. Here I draw on Cohen and Rogers, On Democracy, 154–157. The idea that parties are re-
quired to organize political choice and to provide a focus for public deliberation is one strand
of arguments about “responsible parties” in American political-science literature. My under-
standing of this view has been greatly aided by Lee Perlman, “Parties, Democracy, and Con-
sent,” (unpublished PhD dissertation, MIT, 1987), and, more generally, by the work of Walter
Dean Burnham on the implications of party decline for democratic politics. See, for example,
Burnham, The Current Crisis in American Politics (Oxford: Oxford University Press, 1982).
36
4
I have suggested that we take the notion of democratic association as a
fundamental political ideal and have elaborated that ideal by reference
to an ideal deliberative procedure and the requirements for institutional-
izing such a procedure. I have sketched a few of those requirements here.
To show that the democratic ideal can play the role of fundamental orga-
nizing ideal, I should need to pursue the account of fundamental liber-
ties and political organization in much greater detail and to address a
wide range of other issues as well. Of course, the richer the requirements
are for institutionalizing free public deliberation, the larger the range of
issues that may need to be removed from the political agenda; that is, the
larger the range of issues that forms the background framework of public
deliberation rather than its subject matter. And the larger that range, the
less there is to deliberate about. Whether that is good news or bad news,
it is, in any case, a suitable place to conclude.
37
2
M O R A L P LU R A L I S M A N D
POLITICAL CONSENSUS
T he idea of normative consensus plays a central role in John Rawls’s

theory of justice. In a well-ordered society, he says, “everyone has a
similar sense of justice and in this respect a well-ordered society is homo-
geneous.”1 But is a consensus on fundamental norms of justice a realistic
and attractive prospect for a morally pluralistic society?2
Rawls says little about this question in A Theory of Justice. Although he
is closely attentive there to the diversity of interests and of conceptions of
good among citizens in a well-ordered society, he is generally inattentive
to the pluralism of moral conceptions that can be expected when expres-
sive and associative liberties are protected. As a consequence, he does not
consider the possibility that this pluralism might either exclude consen-
sus on justice altogether or throw its value into question by turning it into
mere compromise. Moreover, since the argument in A Theory of Justice
that justice as fairness is a realistic conception—in particular, the case for
the stability of a just society—depends on the idea that a just society fea-
tures a consensus on principles of justice, the inattention to moral plural-
ism renders the force of that argument uncertain. So justice as fairness
may be, after all, unrealistic and utopian.3
To address these concerns and show that the case for justice as fairness
This essay began as comments on an unpublished paper by John Rawls entitled “A Reason-
ably Realistic Idea of a Well-ordered Society.” I have rewritten it to address Rawls’s “The Do-
main of the Political and Overlapping Consensus,” which was originally published in New
York University Law Review 64, 2 (May 1989): 233–255, and reprinted in the book in which my
paper originally appeared. I thank Michael Hardimon, John Rawls, T. M. Scanlon, and Judith
Thomson for very helpful comments on earlier drafts of this essay.
1. John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), 263.
2. Rawls is concerned with forms of diversity that extend beyond the domain of morality, for
example, to religious and philosophical matters. Nothing turns on the limitation that I adopt
here.
3. The ideal of consensus may, of course, be unrealistic in other ways as well.
38
M O R A L P L U R A L I S M A N D P O L I T I C A L C O N S E N S U S
can be restated under more realistic assumptions, Rawls introduced the

idea of an overlapping consensus and, corresponding to this idea, a condi-
tion on the acceptability of a conception of justice that I refer to as the
“pluralistic consensus test.” A society features an overlapping consensus
on norms of justice if and only if it is a morally pluralistic society with a
consensus on norms of justice in which each citizen, holding one of the
different moralities that win adherents and persist over time in the soci-
ety, supports the consensual norms as the correct account of justice.
Norms of justice satisfy the pluralistic consensus test if and only if those
norms could provide the focus of an overlapping consensus in a society
regulated by those norms and operating under favorable conditions.4 A
conception of justice that would not be so supported by at least some of
the moral doctrines that persist within a society regulated by it, and so
could not be the focus of an overlapping consensus, fails to meet the plu-
ralistic consensus test and is, to this extent, unreasonable.5
Why unreasonable? Why (if at all) should requirements of justice be
realistic? “Because ought implies can” will not do as an answer, because
the question concerns justice, not what ought to be done, all things con-
sidered. Judgments about what ought to be done, all things considered,
must, of course, be sensitive to all sorts of practical matters, since issues of
practicality plainly are among the things to be considered. The question
is what sorts of constraints on realizability are constitutive of ideal justice.
And in matters of justice, realism is an uncertain good. By accepting the
“demands” of realism, we may be led to build an accommodation to un-
happy, grim, and even hideous facts of political life into the foundations
of political justification and into the fundamental principles of justice
themselves.
Focusing this general concern about the demands of realism on the
pluralistic consensus test, one might say that in aiming for a conception
of justice that could realistically be supported by a pluralistic consensus,
one in fact undercuts the attraction of the conception that results. Con-
sider the following elaboration of this objection:
The pluralistic consensus test asks us to evaluate a conception

of justice in part by asking whether we can realistically expect
the conception to be supported as the correct account of justice
4. I return to the issue of favorable conditions later.
5. I explain the point of the phrase “to this extent” later.
39
by the diverse moralities in a well-ordered society. But why

should we be concerned with such support? In fact, requiring
it forces an accommodation to power at the foundations of a
theory of justice—to the power of those who believe the false
and spurn the good. Accommodation to power is commonly
prudent and often recommended by our all-things-considered
judgments about the application of moral ideals to the facts of
life. We give money to the thief who threatens our life; we let
the rich get richer if that is what’s needed to get them to invest;
we pay the lion’s share of the surplus to the greedy if that is nec-
essary to motivate them to use their talents for the common
good (at least in the first case we don’t call it “justice”). And we
often frame our political arguments and proposals to win broad
acceptance, if that is what we must do to keep those who don’t
believe the true and love the good from making life worse for
those of us who do. But adjustments designed to build support
do not define ideal justice. To suppose otherwise would be to
permit the facts of power to fix the content of the fundamental
requirements of justice, thus undercutting their attraction as
basic requirements. Philosophers, above all, should resist the
confusion of justice with accommodation and a moral ideal
with a consensus on principles that accommodate the power of
thieves, pirates, and benighted souls. Because if philosophers
are not good for that, then just what are they good for?6
Responding to this objection, Rawls argues that consensus on justice is

both a realistic and an attractive prospect for a morally pluralistic society
and that subjecting conceptions of justice to the pluralistic consensus test
is not tantamount to substituting mere compromise for genuine moral
consensus and through that substitution advancing an account of justice
that is “political in the wrong way” (234).7
I agree with Rawls’s main contentions, and my aim here is to explore
the problem itself, to discuss some surrounding issues, and to clarify the
grounds of agreement. After some initial points of clarification, I offer a
6. The objection extrapolates on some points made by G. A. Cohen, in a conversation about
Rawls’s difference principle.
7. All references to “The Domain of the Political” are included parenthetically within the
text. Page numbers refer to the New York University Law Review edition cited in the unnum-
bered note at the start of this essay.
40
generic statement of the problem of moral consensus and pluralism.

Then, I discuss some historical background, linking the problem of plu-
ralism, realism, and moral consensus to a line of argument extending
from Rousseau through Hegel to Marx. Next, I discuss and criticize one
source of concern about imposing constraints of realism and, in particu-
lar, imposing the pluralistic consensus test on a conception of justice—
that the constraint of realism undermines a substantively egalitarian con-
ception of justice. Although this discussion does not address the concern
about pluralism and realism in its most generic form, I include it because
I suspect that the energy surrounding the debate about pluralism and po-
litical consensus derives importantly from alleged implications of the de-
bate for matters of equality. Finally, I argue in more general terms that
the pluralistic consensus condition does not fall prey to the objection I
have sketched here. The argument turns on taking what Rawls calls “the
fact of pluralism” (235) in a certain way. In particular, I distinguish the
fact of pluralism from the fact of reasonable pluralism and, drawing on
this distinction, I suggest that in aiming to find a conception of justice
that meets the pluralistic consensus test, we are not simply adjusting ide-
als to the facts of life and to moral pluralism as one such fact. Instead, we
are acknowledging the scope of practical reason. Put otherwise, in aim-
ing to find a conception of justice that meets the pluralistic consensus
test, we are not accommodating justice to an unfavorable condition of
human life, since, as the idea of reasonable pluralism shows, we ought
not to count moral pluralism itself among the unfavorable conditions.
The Place of Consensus

Before getting to these issues, I need to clarify one remark I made earlier.
I said that a conception of justice that fails to meet the pluralistic consen-
sus test is, to this extent, unreasonable. The phrase “to this extent” is
meant to indicate the place of the pluralistic consensus test in an account
of justice and, in particular, its role in the two-stage strategy of argument
that Rawls sketches in “The Domain of the Political.”
Rawls emphasizes that the idea of an overlapping consensus and the
pluralistic consensus test come into play at the second stage of a two-
part argument for a conception of justice. The aim of the first stage is,
roughly, to show that the content of a conception is attractive—that it or-
ganizes a set of fundamental political values in a plausible way. The aim
of the second stage is to determine whether a conception of justice that is
41
in other respects attractive is also realistic—in particular, that it is stable.

Showing that it is stable consists in part in showing that it satisfies the
pluralistic consensus test: that different people, brought up within and at-
tracted to different traditions of moral thought, might each affirm the
conception as the correct account of justice.
But how, more precisely, are we to understand the relationship be-
tween the results of the first stage and the argument at the second? What
would follow if there were problems at the second stage? Three possibili-
ties suggest themselves: (1) it is necessary that the correct account of jus-
tice satisfies the pluralistic consensus test; (2) satisfying the test is not nec-
essary, though it does provide some support for a conception of justice; or
(3) satisfying the test is a desideratum that has no bearing on the correct-
ness of an account of justice. In case (3), the pluralistic consensus test
might be interpreted as a condition on the all-things-considered reason-
ableness of a conception of justice or perhaps as a test of the legitimacy of
the exercise of state power, not as a condition on the justice of the institu-
tions through which that power is exercised. On this interpretation, if the
best understanding of justice failed to satisfy the pluralistic consensus test
even under favorable conditions, we ought to conclude that there is an
unhappy divergence between justice and legitimacy—that even under
the best conditions we can realistically hope for, it will be illegitimate to
secure justice—but not that we should revise our conception of justice.
Interpretation (3) may by suggested by Rawls’s emphasis (234) on the
importance of separating the two stages of argument and, so, distinguish-
ing questions of justice from issues about the course of the world. But it
is, in fact, ruled out by the description of the conclusions of the first stage
as “provisionally on hand” (246) and the remark that the argument is “not
complete” until the case for stability has been presented (245n27). I am
not sure which of the other two views Rawls means to endorse. But for
the purposes of this essay, I will assume that (2) is right, that satisfying the
pluralistic consensus condition does count in favor of the correctness of a
conception of justice, and that while failure to meet it is not a sufficient
reason for rejecting a conception, it would provide some reason to mod-
ify a view to bring it into conformity with that test.
Consensus and Moral Pluralism

Pluralism takes a variety of forms, and so there is, correspondingly, a vari-
ety of ways that it might raise troubles for consensus and social unity. To
42
state the specific problem of pluralism and consensus that I will be con-
sidering here, I first need to fix some terminology. Following Rawls, then,
I will say that a “well-ordered society” is a society in which it is common
knowledge that the members share an understanding of justice and a
willingness to act on that understanding. A well-ordered society, that is,
features a restricted but important moral consensus. The moral consen-
sus is restricted in that it extends only to certain basic constitutional val-
ues and principles and norms of distributive justice, and not to all aspects
of the conduct of life. Despite this limitation, the consensus that defines
a well-ordered society is a genuinely moral consensus. For the norms and
ideals on which there is consensus play a reason-giving and an authorita-
tive role in the deliberation and choices of individual citizens.8
At the same time, a well-ordered society may be morally pluralistic in
that members may have conflicting views about the fundamental norms
and ideals that ought to guide conduct in life more generally. In a mor-
ally pluralistic society, the members hold different theories about what is
valuable and worth doing. Thus understood, moral pluralism is to be dis-
tinguished both from cultural pluralism—the existence of groups of peo-
ple within a single society who share distinct histories and ways of life
and who share a common identity as members of a group—and from or-
ganizational pluralism—the existence of a plurality of organized groups
pursuing distinctive interests or ideals. These forms of pluralism are dis-
tinct phenomena and less plausibly understood as a matter of people
holding different theories. So the discussion here of moral pluralism and
consensus is limited and does not naturally translate into an account of
consensus and either cultural or organizational pluralism.
Moving now from terminology to substance: a moral consensus on po-
litical fundamentals is a fundamental good for at least three reasons.
First, for any conception of justice, the likelihood that social order will
stably conform to the conception is increased by the existence of a moral
consensus on it.9
Second, the existence of a moral consensus supports a variety of spe-
cific values of considerable importance. It increases social trust and har-
8. I identify moral reasons by their functional role in individual deliberation and choice not
by their content. There may be content restrictions as well, but I think that the functional role
characterization captures a central aspect of ordinary usage and in any case suffices for my pur-
poses here.
9. See Rawls’s “third general fact,” 235.
43
mony, supports social peace, reduces the complexity of decision making,

encourages a willingness to cooperate and so reduces the costs of moni-
toring and enforcement, and—assuming the consensus is reflected in
public debate and decisions—reduces alienation from public choices,
because citizens embrace the norms and ideals that guide those choices.
Third, a consensus on norms of justice provides a way to reconcile the
ideal of an association whose members are self-governing with an ac-
knowledgment of the central role of social and political arrangements in
shaping the self-conceptions of citizens, constraining their actions, chan-
neling their choices, and determining the outcomes of those choices.10
For when a consensus on norms and values underlies and explains col-
lective decisions, citizens whose lives are governed by those decisions
might nonetheless be said to be self-governing, because each endorses
the considerations that produce the decisions as genuinely moral reasons
and affirms their implementation.11
But not just any consensus is attractive, as is indicated by reflection on
these reasons themselves. If, for example, a moral consensus is attractive
because it provides a way to make the ideal of free association consistent
with the unavoidable chains of political connection, then the consensus
must be a free moral consensus and not simply a form of enforced homo-
geneity. A free consensus is a consensus arrived at under conditions that
ensure the possibility of individual reflection and public deliberation—
conditions in which, for instance, expressive and associative liberties are
protected.
It is at just this point that a minimal condition of realism appears to un-
dermine either the possibility of consensus or, at least, its attractions as an
ideal. For the assurance of expressive and associative liberties—an assur-
ance that is necessary if the consensus is to be free and attractive—will
also produce moral, religious, and philosophical pluralism.12 But can a
genuine moral consensus survive this “fact of pluralism” (235)? Or does
an insistence on consensus under conditions of pluralism in effect turn
political philosophy into a search for a political compromise among peo-
ple who disagree?
10. See Jean-Jacques Rousseau, On the Social Contract, trans. Judith R. Masters (New York:
St. Martin’s, 1978), Book 1, chap. 6.
11. We also need to add that everyone believes with good reason that the decisions express
the values.
12. This is what Rawls calls the “first general fact.” See 234–235.
44
Historical Excursus
These concerns about the pluralistic consensus condition ought to have
a familiar ring. Earlier I mentioned the problem of reconciling self-
government with the chains of political connection. Rousseau identified
this problem and thought it could be solved if social order were regulated
by a consensual understanding of the common good—a “general will.”
Rousseau’s solution is commonly rejected on the ground that it is inat-
tentive to differences among people and to the diversity of human inter-
ests and ideals. In the face of that diversity, according to the objection,
consensus on the common good can be achieved only through the unat-
tractive combination of a sectarian conception of virtue and, for those
who do not share that conception, enforced subordination and homoge-
neity in the name of freedom.
Hegel’s response to Rousseau was more complex. He agreed that free-
dom could be reconciled with the chains of political connection and ap-
plauded the notion of a general will as the way to achieve that reconcilia-
tion.13 But he also appreciated the force of the critique of Rousseau that I
just sketched. His conclusion was that it was necessary to reformulate the
classical ideal of a political community organized around a moral con-
sensus in light of the modern distinction between the unity of political
society and the diversity of civil society. This distinction shapes Hegel’s
own political conception in three important ways:
1. He endorsed a fundamental distinction between civic diversity and
political unity, associating that distinction with the differentiation be-
tween two spheres of social life. While the civil sphere would feature a di-
versity of aims and ideals and a range of individual and group activities
organized around those aims and ideals, the political sphere would be or-
ganized around a set of values that both claimed authority over individ-
ual concerns and were alleged to lie within the diverse aspirations of civil
life and to provide their common ground.14
2. His distinction between political unity and civic diversity is associ-
13. Hegel’s discussion of Rousseau in the History of Philosophy is more balanced than his
critical remarks in the Philosophy of Right. Compare Lectures on the History of Philosophy, vol.
3, trans. Elizabeth Haldane (New York: Humanities Press, 1968) with Philosophy of Right,
trans. T. M. Knox (Oxford: Clarendon Press, 1952), 156–157.
14. See Philosophy of Right, paragraph 261, where Hegel says that the state is both an “exter-
nal necessity,” with respect to the family and civil society, and “the end immanent within
them.”
45
ated with an acceptance of substantially inegalitarian forms of civic diver-

sity,15 as though an acceptance of that distinction and of a social sphere in
which people pursue diverse aims itself brings inegalitarian implications
in its wake.
3. Concerned to affirm the unity of the state in the face of the tenden-
cies to social fragmentation that might follow from civic diversity, he de-
fended a strong, highly centralized, executive-dominated constitutional
monarchy, featuring a corporatist form of representation and special po-
litical rights for the landed class.
At least since Marx, critics of Hegel have objected that some or all of
these gestures at reinterpreting the ideal of political unity in the face of
civic diversity represent unwanted accommodations to de facto power in
the formulation of basic political ideals. Marx, for example, objected to
all three.16 Putting Hegel’s favored form of state to the side, these allega-
tions of “accommodation” raise two questions that are relevant for our
purposes here.
First, does the reformulation of the ideal of consensual political unity
with an eye to respecting the diversity of civil society itself represent an
objectionable accommodation? Do we find unacceptable accommoda-
tion in Hegel’s reformulation of the ideal of political society to accommo-
date the diversity of aspirations characteristic of civil society or in Rawls’s
broadly parallel idea that a reasonable conception of justice should be
supportable by an overlapping consensus?
Second, does the affirmation of moral diversity lead to an accommoda-
tion of social and economic privilege? A Theory of Justice defended an
egalitarian liberalism that departed from Hegel’s accommodation to in-
egalitarian forms of civic diversity. Does this egalitarianism survive the
gesture at realism reflected in the pluralistic consensus condition? Put
otherwise: the pluralistic consensus condition presumably restricts the
content of norms of justice in some way. More demanding norms are less
plausibly the object of agreement than less demanding norms. So does
15. Hegel did acknowledge the need to regulate property in the name of the general welfare
and to avoid certain extreme cases of poverty (see Philosophy of Right, paragraphs 234–248).
But his view does not appear to countenance the regulation of economic activity with an eye to
ensuring that the final distribution of resources is not determined by differences of social back-
ground and natural ability.
16. See his “On the Jewish Question,” in Marx-Engels Reader, 2d ed., ed. Robert Tucker
(New York: Norton, 1978), 26–46; Critique of Hegel’s Philosophy of Right, trans. Joseph O’Mal-
ley (Cambridge: Cambridge University Press, 1970).
46
the importance of accommodating moral diversity lead to a thinner con-

ception of justice that lacks the critical egalitarian dimension of Rawls’s
earlier position?
Because an affirmative answer to the second question would fuel an af-
firmative answer to the first, I will begin with diversity and equality.
The Case of Equality

A number of commentators on Rawls’s post-Theory of Justice work have
noted that the many reformulations of his views about political justifica-
tion have not yet been matched by similar revisions in the substance of
the theory. My impression17 is that lots of people now think that Rawls’s
discussions of political justification—with their emphasis on the impor-
tance of realism, on the practical nature of political philosophy, and on
the associated idea of an overlapping consensus—do require a shift in
the substance of his theory of justice and, in particular, a shift in an
inegalitarian direction.18
The reasoning goes something like this: “Rawls recognizes the utopian-
ism of his earlier conception of a well-ordered society. So he now recom-
mends that political justification proceed by identifying the common
ground among the diverse moralities and conceptions of justice in our
own society. But if we follow that recommendation, we will certainly not
find support for the specifically egalitarian aspects of Theory of Justice,
since there is (to put it mildly) considerable contemporary controversy
about egalitarian political views.”
This account of the idea of an overlapping consensus, with its empha-
sis on locating common ground among current political views, is mis-
taken in several ways. Once we see where it goes wrong, we shall see as
well that the concern for realism expressed in the pluralistic consensus
test has none of the alleged implications. To make this case, I will begin
with a sketch of the egalitarian content of the theory and the strategy of
argument for it and then proceed to a discussion of the objection.
17. This impression was confirmed by conversations at the conference at which I presented
the first draft of these comments. See also the concerns about the “abstractions, vagueness, and
conservatism” of Rawls’s later work expressed in Thomas Pogge’s Realizing Rawls (Ithaca, NY:
Cornell University Press, 1989), 4.
18. See, for example, John Gray, “Contractarian Method, Private Property, and Market
Economy,” in John W. Chapman and J. Roland Pennock, eds., Markets and Justice, Nomos
XXIII (New York: New York University Press, 1989), 13–58; and William Galston, “Pluralism
and Social Unity,” Ethics 99, 4 (July 1989): 711.
47
The egalitarian content of A Theory of Justice is encapsulated in three

requirements: the fair value of political liberty, fair equality of opportu-
nity, and the maximin criterion of distributive equity. These three condi-
tions, which are meant to sever the distribution of advantage from social
background and natural difference, represent substantively egalitarian in-
terpretations of more formal and less controversial norms of equal liberty,
equal opportunity, and the common good.19 A contention common to
egalitarian liberal political conceptions generally, and advanced in A
Theory of Justice in particular, is that we are led to these substantively
egalitarian interpretations by considering the justification of those more
formal and political norms.
The basic strategy of argument for this contention is familiar and pro-
ceeds by bootstrapping. Thus, associated with the more formal require-
ments of equal liberties and assurances of opportunity is a conception of
the properties of human beings that is important for the purposes of polit-
ical justification. That conception of persons supposes that the relevant
features are not race, color, cultural creed, sex, religion, and the like.
The relevant features are certain potentialities (moral powers)—for ex-
ample, the capacity to govern one’s conduct and to revise one’s aspira-
tions—rather than the determinate form in which those potentialities are
realized. The rationale for the protection of liberties and formal opportu-
nity, for example, lies in part in the importance of assuring favorable con-
ditions for the realization of the basic potentialities. But—and here is
where the bootstrapping comes in—once we acknowledge the need for
favorable conditions for realizing the basic potentialities, we are naturally
led from the more formal to the more substantively egalitarian require-
ments, since the latter more fully elaborate the range of favorable condi-
tions.
With this quick sketch as background, I can now state more precisely
the concern already noted about the idea of an overlapping consensus.
The intuitive objection was that the need to confine fundamental politi-
cal justification to considerations that lie on common ground would un-
dercut the egalitarian components and result in an unacceptable accom-
modation to power in the formulation of principles. Is this right? Does
the requirement of proceeding on common ground deprive us of the ar-
19. See Theory of Justice, 65.
48
gumentative resources necessary for the bootstrapping argument for an

egalitarian form of liberalism?
common groun d
To see why the answer is no, it is important to note first that the bootstrap
argument for the egalitarian view is itself meant to proceed on common
ground shared by different moral conceptions in a well-ordered society
governed by it. That may seem puzzling, since the conception of potenti-
alities as morally fundamental may strike some as peculiarly Kantian. But
the contention of the argument (which I am not evaluating here) is that
those ideas will seem attractive for the purposes of political argument to
anyone who considers how best to defend the liberties, formal norms of
equal opportunity, and the requirement that public powers be exercised
for the common good.20
Noting this draws attention to a first feature of the notion of an overlap-
ping consensus that is important in assessing the objection. What lies in
the intersection of different moral conceptions is not simply a set of poli-
cies or a system of norms within which political conflict and competition
proceed.21 Nor is it simply a determinate set of moral principles. Instead,
the consensus extends to a view of persons, of the importance of fairness
and other political values, of what counts as an advantage, and of which
practices are paradigmatically evil (e.g., slavery, religious intolerance,
and racial discrimination). In short, what lies at the intersection of differ-
ent views is a (restricted) terrain on which moral and political argument
can be conducted, and not simply a fixed and determinate set of substan-
tive points of political agreement.
To show, then, that an egalitarian conception of justice meets the plu-
ralistic consensus test, one needs to show that the bootstrap argument
succeeds and that the terrain on which that argument proceeds could it-
self be the focus of an overlapping consensus in a society governed by it.
One need not deny the obvious fact of disagreement on egalitarian politi-
20. The attribution to Kant in particular of the idea that abstract human potentialities are
morally fundamental is also off the mark historically. That idea plays a central role in Rous-
seau’s view and is also suggested in Locke’s theory of natural law. The variations on this general
theme are complex, as is the evolution of the idea; fortunately, these details are not relevant
here.
21. Robert Dahl, for example, emphasizes the importance of “underlying consensus on pol-
icy” and on the basic rules of political competition in A Preface to Democratic Theory (Chi-
cago: University of Chicago Press, 1956), 75–84, 132.
49
cal ideals or the only slightly less obvious fact that such disagreement is
likely to persist even under favorable conditions.
cont emp ora ry sup p ort

When the case for an egalitarian conception of justice is understood as a
bootstrap argument and the common ground is understood in the way
that I just sketched, it is not so obvious that an appeal to a wide range of
contemporary political views will fail to support the substantively egalitar-
ian aspects of the conception. For we do not require de facto agreement
on substantively egalitarian norms, but only that the reasoning support-
ing those norms proceed on common ground. That is, we require that
the egalitarian features represent a reasonable extension of what people
do agree to—that they “extend the range of some existing consensus” by
bringing the best justification of certain fundamental points of agree-
ment to bear on unsettled and controversial matters.22 That contention is
not so implausible, because—as I noted earlier—the bootstrap argument
for the egalitarian ideals proceeds principally by reference to points of
agreement about the value of the liberties and certain formal require-
ments of equality.
rol e of ov erl a p p ing c ons e ns us

While the contention that the resources for defending an egalitarian po-
litical conception are implicit in current understandings may not, then,
be entirely implausible, it should not be identified with the thesis that an
egalitarian liberal political conception can meet the pluralistic consen-
sus test. That test does not require that we rummage through the political
culture searching for underlying points of agreement among the views
featured in it.23 Rummaging may serve an important function, and I will
say a word about it below. But the pluralistic consensus test does not itself
command a search for de facto points of agreement at all, and so the fail-
ure to find any would not undercut the force of an egalitarian conception
of justice.
Instead it formulates a test on the reasonableness of a political concep-
tion that is in other respects attractive. The test is this: consider a pro-
22. See Theory of Justice, 582. For elaboration of this strategy, see my “Democratic Equality
and the Difference Principle,” Ethics 99, 4 (July 1989): 727–751.
23. See John Rawls, “The Priority of the Right and Ideas of the Good,” Philosophy and Pub-
lic Affairs 17, 4 (Fall 1988): 275–276.
50
posed conception of justice in operation, and then consider whether the

principles, ideals, and terms of argument that figure in it provide moral
reasons within the views that could be expected to arise among those
who live in a society governed by it. Bringing this to bear on the issue of
egalitarian liberalism, then, we are to imagine a society regulated by such
a conception and existing across several generations. In such a society, we
can reasonably expect moral diversity. We also can expect widespread
agreement on the fundamental value of the liberties and on at least for-
mal understandings of equality. But then, if there were such agreement,
and if the bootstrap argument had any force, the diverse moral under-
standings would each still have the resources necessary for supporting the
substantively egalitarian conception as the correct conception of justice.
I have, of course, not tried to defend the bootstrap argument here. In-
stead, I have only argued that the pluralistic consensus test does not un-
dercut the force (whatever its magnitude may be) of that argument. The
acknowledgment of diversity underscored by the notion of an overlapping
consensus does not undercut the critique of privilege contained in the
egalitarian aspects of egalitarian liberalism.
cont emp ora ry sup p or t, a ga i n

I have been emphasizing that the pluralistic consensus test does not itself
require a search for implicit points of agreement in current moral views.
Nonetheless, the existence of such points might have a certain indirect
relevance to justification. For, given that the deliberative liberties now
receive some protection, it seems implausible to suppose that existing
moral views simply represent accommodations to current and historical
injustices and would not continue to have some hold under just condi-
tions. So it would be surprising if we could not already find the resources
available in current moral understandings for defending a view of justice
that we would also be able to defend under more favorable conditions.
And if the pluralistic consensus test is acceptable, then there is also some
rationale for taking current points of agreement seriously. But it must be
emphasized that when we understand the rationale for an examination of
current points of agreement this way, we are not letting anything about
justification turn on the mere fact of current consensus. In fact, it is
never the case—not in the gesture to current understandings of value
and not in the requirement of overlapping consensus—that de facto
agreement itself plays a role in justification.
51
With this last point, I have begun to tread on the issues of the next sec-
tion and so shall move directly to them.
Realism and Reason

Now we come to the first of the issues about accommodation that I noted
earlier: Does the pluralistic consensus test represent an unwanted ac-
commodation to power? I begin my discussion of this question with some
distinctions that will play an essential role in my (negative) answer.
rea sona bl e p l ura l i s m

Rawls refers to the fact that the deliberative liberties produce diversity as
“the fact of pluralism.” I think that this terminology may be misleading
because “fact” puts the emphasis in the wrong place.24 To explain why, I
need first to introduce the idea of reasonable pluralism.25
The idea of reasonable pluralism is that there are distinct understand-
ings of value, each of which is fully reasonable (235–238). An understand-
ing of value is fully reasonable just in case its adherents are stably dis-
posed to affirm it as they acquire new information and subject it to
critical reflection.26 The contention that there is a plurality of such un-
derstandings is suggested by the absence of convergence in reflection on
issues of value, which leaves disagreements, for example, about the value
of choice, welfare, and self-actualization; about the value of contempla-
tive and practical lives; about the value of devotions to friends and lovers
as distinct from more diffuse concerns about abstract others; and about
the values of poetic expression and political engagement.
What we ought to suppose about the truth of our beliefs about any sub-
ject matter, evaluative or otherwise, in the face of such an apparently
“irresoluble rivalry” of reasonable alternative views, is an open philosoph-
ical question.27 But among the rationally acceptable answers to that ques-
24. One reason for referring to a fact of pluralism is to distinguish the view that we need to
accommodate the diversity of values that follows on the protection of the liberties from the
view that that diversity should be accommodated because it is a good thing in itself. Nothing
that I say is meant to challenge the propriety of this usage.
25. My discussion of reasonable pluralism is in agreement with Rawls’s account of the “bur-
dens of reason” (235–238). The point of the discussion is largely to indicate the special impor-
tance of those burdens, as distinct from the four other general facts that Rawls discusses (234–
235), in explaining the pluralistic consensus test and in responding to objections to it.
26. I take this formulation from conversations with Mark Johnston.
27. See, for example, W. V. O. Quine, Pursuit of Truth (Cambridge, MA: Harvard Univer-
sity Press, 1990), 98–101, from whom I take the phrase “irresoluble rivalry” and the term “sec-
tarian” as it is used in the next sentence.
52
tion is that it is permissible, even with full awareness of the fact of re-
flective divergence, to take the sectarian route of affirming one’s own
view; that is, believing it as a matter of faith. And since believing is believ-
ing true, a rationally permissible (though not mandatory) response to an
apparently irresoluble rivalry of evaluative conceptions is to affirm that
one’s own view contains the whole truth, while the truths in other views
are simply the subsets of those views that intersect with one’s own. This
being one of the options, and the option that creates the most trouble for
the pluralistic consensus test, I will frame the rest of my discussion so that
it is consistent with it.
These remarks about reasonable pluralism suggest two different ways
to understand the fact of pluralism:
The simple fact of pluralism: The protection of the deliberative liber-
ties will result in moral pluralism.
The fact of reasonable pluralism: The protection of the deliberative
liberties will result in moral pluralism, and some of the moral con-
ceptions will fall within the set of fully reasonable conceptions.
The reasonable pluralism interpretation does make a factual claim. The
asserted fact, however, is not simply that the protection of deliberative
liberties will result in a plurality of conceptions of value but, further, that
a number of those conceptions will be reasonable and permissibly taken
by their adherents to be true.
th e rea sona bl e p l ura l ism e x p l a na t i on

Consider now a conception of justice that we wish to subject to the plu-
ralistic consensus test. We imagine a society regulated by that conception
and in which the condition of reasonable pluralism obtains. The pluralis-
tic consensus test requires that the values and principles used to autho-
rize the exercise of power by the state must be restricted to those that are
compelling to the different reasonable moral views adhered to in the so-
ciety. Consider some people—call them “us” (or “we”)—who hold one
such view and think that others believe what is false about the domain of
value. Should we think that the pluralistic consensus test, which prevents
us from relying on the whole truth in authorizing the use of power, is
simply an accommodation to the de facto power of those others? It de-
pends, and what it depends on is clarified by the distinction between sim-
ple and reasonable pluralism.
Suppose that we are impressed by the lack of reflective convergence in
53
understandings of value, that we acknowledge the idea of reasonable plu-

ralism, and at the same time that we embrace (not unreasonably) the sec-
tarian view that our moral views are true. Because these are consistent
positions, our sectarianism does not require that we condemn as unrea-
sonable everyone who believes what we take to be false. And this provides
a rationale for formulating a conception of justice that is confined to con-
siderations that others also take to be moral reasons.
In particular, when we restrict ourselves in political argument to the
subset of moral considerations that others who have reasonable views ac-
cept as well, we are doing three things. First, we are advancing consider-
ations that we take to be genuine moral reasons; the adherents of each of
the views that support the overlapping consensus hold that nothing but
the truth lies in the overlapping consensus.
Second, in restricting ourselves to a subset of the true moral reasons—
appealing to nothing but the truth, though not to the whole truth—we
are not simply acknowledging that those who believe the false and spurn
the good have the power to make their voices heard or to make our lives
miserable if we fail to heed those voices. Instead, we are acknowledging
that their views are not unreasonable, even if they do believe what is
false. In short, we are moved not by their power but by an acknowledg-
ment that they are reasonable.
Third, we are taking cognizance of a peculiarity in insisting on the
whole (sectarian) truth in the face of our acknowledgment of the idea of
reasonable pluralism. For suppose we acknowledge it and affirm the di-
vergence of moralities under reflection. Then we must see that if we were
to appeal to the whole truth, that appeal would be, from the standpoint of
others who we take to be reasonable, indistinguishable from simply ap-
pealing to what we believe. But we already acknowledge that the mere
appeal to what we believe carries no force in justification.28
Suppose, for example, we believe that welfare is the sole ultimate
28. For elaboration of this point, see Thomas Nagel, “Moral Conflict and Political Legiti-
macy,” Philosophy and Public Affairs 16, 3 (Summer 1987): 215–240. Joseph Raz has criticized
Nagel’s point, suggesting that it rests on an untenable distinction between the position of the
speaker who advances a justification and the listener to whom it is addressed. See his “Facing
Diversity: The Case of Epistemic Abstinence,” Philosophy and Public Affairs 19, 1 (Winter
1990): 37–39. I am not persuaded by Raz’s contention. He is right that the positions of speaker
and listener are parallel. But taking up the point of view of the person to whom a justification is
addressed is simply a heuristic for understanding the limited force of an argument that appeals
to the whole truth. So, far from undermining Nagel’s point, the parallelism is essential to draw-
ing the right conclusions from the use of the heuristic.
54
good, and we understand that view to imply that choice is not an inde-
pendent final value. In the course of political argument, we affirm: “It is
true that welfare is the sole ultimate good.” Now others ought not to sup-
pose that what we mean is equally well captured by “We believe that wel-
fare is the sole ultimate good.” The indistinguishability at issue is not se-
mantical. The point, rather, is that if others accept the idea of reasonable
pluralism, then they notice what we also notice, namely, that what lies
between our taking our views to be reasonable (about which there may
be no disagreement) and our taking them to be true (about which there
is disagreement) is not a further reason, but simply our (rationally permis-
sible) belief in those views. Because there is nothing else that lies in be-
tween, an appeal to the whole truth will seem indistinguishable from an
appeal to what we believe.
t he simp l e-fa ct ex p l a na t i on
Following the reasonable pluralism interpretation, then, when we restrict
ourselves to common ground in the face of the fact of diversity, we are ac-
knowledging that reason does not mandate a single moral view and then
are refraining from imposing ourselves on others who are prepared to be
reasonable. This account of whom we need to accommodate turns on
our willingness to acknowledge that some people with whom we funda-
mentally disagree are not unreasonable. That is why we are not simply
accommodating principles to power when we are concerned, to ensure
that the conception of justice is acceptable to them as well. This explana-
tion of the pluralistic consensus test might be clarified by contrasting it
with another explanation, which is suggested by some of Rawls’s remarks
but which is not persuasive.
As I indicated at the outset, Rawls emphasizes the importance of real-
ism in the formulation of reasonable ideals. And he suggests that when
we confine ourselves to considerations that are reasons for others as well,
we are simply adjusting to certain general facts about the social world.
Here the emphasis is on the need to be realistic, to find common ground
because disagreement is a basic fact of life under free conditions.
To see why this explanation of the need to accommodate diversity is
not right, notice that it is a plausible general fact that there will always be
people with unreasonable views. But the fact that there are some people
with unreasonable views does not require that we adjust our conception
of justice so that it can be supported by an overlapping consensus that
55
will appeal to them. While we need to take the fact of disagreement into
account in some way in deciding what to do, the pluralistic consensus
condition is certainly not the only way to do that and is not mandated by
the recognition that there are and will always be such people. Further-
more, if we did embrace the requirement that a conception of justice be
able to bring everyone on board—that it restrict itself to reasons em-
braced by all understandings of value—then it is hard to see what the re-
sponse would be to the objection that the requirement of an overlapping
consensus simply forces an accommodation to power.
The problem with this explanation is that it makes too much of the de
facto diversity highlighted in the simple-fact interpretation. The first ex-
planation—which draws essentially on the idea of reasonable plural-
ism—does not deny the relevance of the fact that under conditions of de-
liberative liberty there will be diversity. But the response to that fact is not
undiscriminating and, in particular, is controlled by the distinction be-
tween reasonable and unreasonable understandings of value.
Ensuring that a conception of justice fits the fact of diversity under
conditions of deliberative liberty is not, then, an unacceptable accommo-
dation to power. But the reason that it is acceptable is not because diver-
sity is a fact of life, as the simple-fact interpretation of pluralism states,
and not because adjustment to general and unalterable facts of social life
is always to be distinguished from accommodation to power. Instead, that
adjustment is reasonable because some forms of diversity are the natural
consequence of the free exercise of practical reason. Once we agree that
they are, we will not be inclined to count moral diversity among the un-
favorable facts of human life nor to confuse a concern to find a concep-
tion of justice consistent with it with a willingness to compromise justice
in the face of the course of the world.
ex cl usion
Answering the charge of unwarranted accommodation, then, commits us
to the view that we need not accommodate the unreasonable. Indeed,
given the explanation for this view, if we did accommodate the unreason-
able in the formulation of fundamental principles, then we would be un-
acceptably adjusting principles to de facto power.
But this brings me to a different concern about power and political
consensus: that the promise of consensus is associated with the practice
of arbitrary exclusion. In view of the problem of securing general agree-
56
ment on anything, claims to speak on behalf of all of the reasonable de-

pend, it will be argued, on drawing arbitrary boundaries around the com-
munity of the reasonable.29 So the charge is that any appeal to the ideal of
consensus, in fact, rests on the power to exclude, exercised in this case
through the pretense of discovering that some people are unreasonable.
In the case at hand, the exclusion is of a special kind. It does not
amount to a deprivation of liberties or of what are conventionally under-
stood to be the advantages of social cooperation. Instead, exclusion lies in
the fact that the arguments used to justify the exercise of power depend
on norms, values, and ideals that are rejected by some people whose
views will as a consequence not belong to an overlapping consensus. Al-
though this does not violate the ideal of consensus, which requires that
justification proceed by reference to reasons located on the common
ground occupied by all who are prepared to listen to reason, it is exclu-
sion all the same. And it is of a troublesome form. Its implication is that
some people will reject the values, ideals, principles, and norms that
serve, at the most fundamental level, to justify the exercise of power over
them.
These are extremely important and complicated matters, not least be-
cause the charge of unreasonableness is commonly a ponderous way to
express simple disagreement, or, in the distinctively American political
idiom, a thinly disguised signal that one’s opponents are poor or female
or black. But as important as these issues are, I must be very brief here
and intend my comments as a way only to mark out certain issues for fur-
ther examination and to introduce some doubts about the alleged arbi-
trariness of characterizations of views as unreasonable.
Consider, then, some views that might end up being excluded in this
way, in particular those that would deny the protection of liberties on the
basis of the doctrine that “outside the church there is no salvation.” Rawls
discusses this case and states that it is “unreasonable” to use public pow-
ers to enforce this doctrine. I agree. But it is important to distinguish
two ways that such enforcement might be unreasonable. Distinguishing
them will help illustrate what is involved in exclusion on the grounds of
unreasonableness.
The first case is presented by a “rationalist fundamentalist.” This is the
29. I am indebted to Uday Mehta for many discussions of these issues. For discussion of a va-
riety of different strategies of exclusion, see his “Liberal Strategies of Exclusion,” Politics and
Society 18, 4 (December 1990): 427–454.
57
person who denies the idea of reasonable pluralism, affirming instead

that it lies within the competence of reason to know that salvation is the
supreme value, that there is a single path to salvation, that there is no sal-
vation among the damned, and, therefore, that liberty of conscience is to
be condemned. This is not a common view, if only because it claims for
reason, territory usually reserved for faith.30 But if someone were to ad-
vance the view, then one ought to say that she is simply mistaken. Even if
the elements of rationalist fundamentalism are all rationally permissible,
reason surely does not mandate them, and in insisting that it does, the ra-
tionalist fundamentalist is not acknowledging the facts.
This response will not do in the second case. These are the non-
rationalist fundamentalists who accept the limited competence of reason
but deny that reason is controlling in the authorization to use power. By
contrast with the rationalist fundamentalists, they agree that an apprecia-
tion of the value of salvation and of the conditions for achieving it fall
outside the competence of reason and that grasping the truth about the
proper conduct of life depends on faith. But they affirm that truths acces-
sible only through faith are sufficient to authorize the legitimate exercise
of power. What is important is that they are truths, and not the mode of
access to them available for finite human creatures. Faced with non-
rationalist fundamentalists, it will not do to state the case for the idea of
reasonable pluralism; they know that case, celebrate the limited compe-
tence of reason as a guide in human affairs, and lament the self-imposed
disabilities of those who insist on proceeding within its narrow compass.
Still, what they are prepared to do is to impose on those who are out-
side the faith in a way that—so far as those others can tell—is indistin-
guishable from the concededly irrational practice of imposing in the
name of their beliefs. To resist such imposition is not simply to affirm a
disagreement with the nonrationalist fundamentalist. Instead, it is to
complain about this fundamental form of unreasonableness. And finding
them unreasonable in this way, is sufficient to show that the exclusion is
not arbitrary.
Conclusions
I noted earlier that Rawls’s problems—the reasonableness of the ideal of
a consensual order and of the pluralistic consensus condition—echo a set
30. It is an analog to “creation science,” operating in the domain of salvation. The proper re-
sponse is the same in both cases.
58
of concerns familiar from Hegel’s political philosophy and critical discus-

sion of it. Returning now to these concerns, what conclusions about
them can we draw from the discussion here?
First, in A Theory of Justice, Rawls proposed a formulation of the dis-
tinction between political and civil society and a conception of justice
that was meant to accommodate that distinction without carrying the
inegalitarian implications that some have thought intrinsic to it. What-
ever the merits of that earlier defense of egalitarian liberalism, the plural-
istic consensus condition does nothing to weaken it.
Second, Hegel thought that an account of the ideal of a consen-
sual polity suited to modern conditions needed to accommodate the
diversity of values and attachments characteristic of civil society.31 While
Hegel emphasized that the universal-particular distinction, and its in-
stitutionalization in the separation of civil and political spheres, is a dis-
tinguishing feature of modern societies, he did not suppose it to be sim-
ply a brute fact about post-Reformation Europe. Instead, his rationalism
led him to suppose that this peculiarity represented a historically situated
discovery about the operation of practical reason. In accommodating the
diversity institutionalized in civil society, then, political philosophy was
not simply accommodating the bare fact that people differ in aims and
aspirations. Instead, it was acknowledging the diverse promptings of prac-
tical reason itself, even as it sought to find within that diversity, the seeds
of the set of common values underlying political society.32 In short, some
form of civil-political society distinction is an unavoidable aspect of any
attractive ideal, once we see the scope and competence of practical rea-
son. In a Hegelian Doppelsatz: we need to accommodate the ideal to the
real because the real manifests the ideal.
Rawls’s talk about the fact of pluralism, the role of the Reformation in
prompting acknowledgment of that fact, and the need for an overlapping
consensus can be taken in this same spirit. If we accept the idea of rea-
sonable pluralism, then moral diversity is not simply a bare fact, even a
bare general fact about human nature, but, rather, indicates something
31. I am not confident that Hegel held the view I attribute to him in this paragraph. It does
fit with and make sense out of various pieces of his view, including his account of the relation-
ship between civil society and the state, his conception of the role of reason in history, and his
views about the rationality of modern social arrangements. But he does not state it anywhere in
the way that I put it here. If I am wrong in thinking that he held it, nothing else in the essay
would need to change.
32. See Philosophy of Right, paragraph 261.
59
about the operation and powers of practical reason. With this account of
diversity, we have a response to the contention that accommodating dif-
ferent understandings of value in the formulation of basic moral princi-
ples for the political domain is tantamount to supposing that justice com-
mands that we turn our money over to thieves. The response is that we
are accommodating basic principles not to the reality of power but,
rather, to the way that social reality reveals the powers of practical reason.
60
3
A S S O C I AT I O N S A N D D E M O C R A C Y
with Joel Rogers
S ince the publication of John Rawls’s A Theory of Justice, normative

democratic theory has focused principally on three tasks: refining
principles of justice, clarifying the nature of political justification, and
exploring the public policies required to ensure a just distribution of edu-
cation, health care, and other basic resources. Much less attention has
been devoted to examining the political institutions and social arrange-
ments that might plausibly implement reasonable political principles.1
Moreover, the amount of attention paid to issues of organizational and
institutional implementation has varied sharply across the different spe-
cies of normative theory. Neoliberal theorists, concerned chiefly with
protecting liberty by taming power and essentially hostile to the af-
This essay is an abbreviated version of “Associations in Democratic Governance” (with Joel
Rogers), Politics and Society 20, 4 (December 1992): 393–472, published with replies by Ellen
Immergut, Andrew Levine, Jane Mansbridge, Philippe Schmitter, Wolfgang Streeck, Andrew
Szasz, and Iris Young. That essay was published, along with the replies, in Associations and De-
mocracy (London: Verso, 1995). Versions of the longer essay were presented at meetings of the
American Political Science Association, the Princeton University Political Theory Collo-
quium, the Social Organization Colloquium at the University of Wisconsin-Madison, the Soci-
ety for Ethical and Legal Philosophy, the UCLA Center for History and Social Theory, the
University of Chicago Colloquium on Constitutionalism, the University of Maryland Seminar
on Political Theory, PEGS (Political Economy of the Good Society), and CREA (Ecole
Polytechnique); drafts have also been presented at the conference on “Post-Liberal Democratic
Theory” held at the University of Texas at Austin and at the “Associations and Democracy”
conference held at the University of Wisconsin-Madison. We are grateful to participants in
those discussions for many useful comments and suggestions, and especially to Bruce
Ackerman, Suzanne Berger, Owen Fiss, Charles Sabel, Wolfgang Streeck, and Erik Olin
Wright, for the same. We also thank the editors of Social Philosophy & Policy for comments on
an earlier draft of this essay.
1. See John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971),
whose own work is an exception to the generalization made in the text. Another prominent ex-
ception is Roberto Unger’s False Necessity, vol. 2 of Politics (Cambridge: Cambridge University
Press, 1987).
61
firmative state,2 have been far more sensitive to such issues than egalitar-
ian-democratic theorists, who simultaneously embrace classically liberal
concerns with choice, egalitarian concerns with the distribution of re-
sources, and a republican emphasis on the values of citizen participation
and public debate (we sketch such a conception below in Section 1). Ne-
glect of how such values might be implemented has deepened the vul-
nerability of egalitarian-democratic views to the charge of being unrealis-
tic: “Good in theory but not so good in practice.”
In this essay we address this vulnerability by examining the construc-
tive role that “secondary”3 associations—labor unions, employer associa-
tions, citizen lobbies and advocacy groups, private service organizations,
other private groups—can play in a democracy. Our central contention is
that, as a practical matter, implementing democratic norms requires a
high level of secondary group organization of a certain kind. Roughly
speaking, the “level” required is one in which all citizens, irrespective of
their initial endowment, enjoy the political benefits of organization; the
“kind” required is one which delivers those benefits in ways consistent
not only with political equality but also with other democratic norms.
The problem is that the required level and kind of group activity do not
arise naturally, and those groups that do arise often frustrate, rather than
advance, democratic aspirations. Our proposed solution to this problem
is to supplement nature with artifice: through politics, to alter the envi-
ronment, incidence, activity, and governing status of associations in ways
that strengthen democratic order. We call this deliberate politics of asso-
ciations “associative democracy.”4
We would recommend the pursuit of “associative democracy” for a
wide range of administrative and property regimes. Here, however, we as-
sume the context of modern capitalism, where markets are the primary
mechanism of resource allocation and private, individual decisions are
the central determinant of investment. Admitting the limits this context
2. For examples of the institutional program of “neoliberal constitutionalists” hostile to the

affirmative state, see Friedrich A. Hayek, The Constitution of Liberty (Chicago: University of
Chicago Press, 1960); idem, The Mirage of Social Justice, vol. 2 of Law, Legislation, and Liberty
(Chicago: University of Chicago Press, 1976); and James M. Buchanan, The Limits of Liberty:
Between Anarchy and Leviathan (Chicago: University of Chicago Press, 1975).
3. So called because they are, by convention, the large residual of the “primary” organiza-
tions of family, firm, and state.
4. We share the term “associative democracy” with John Mathews, Age of Democracy: The
Political Economy of Post-Fordism (New York: Oxford University Press, 1989). But we arrived at
the term independently.
62
A S S O C I A T I O N S A N D D E M O C R A C Y
places on the satisfaction of egalitarian-democratic norms, our argument

is that associative democracy can improve the practical approximation to
those norms.
Our argument is animated by concerns about the likely future of even
such approximation. Due principally to changes in the organization of
capitalism, many of the most important institutional sources of egalitar-
ian achievement under modern capitalism—from strong unions and em-
ployer organizations to a variety of popular political organizations—have
recently fallen into disarray. The egalitarian project is weakened by a
widening organizational deficit at its base. Recognizing that most social
clocks cannot be turned backward, that new as well as revived institu-
tional structures are needed, we offer associative democracy as a strategy
to rebuild that base—to provide egalitarian democracy with necessary as-
sociative supports.
We sketch the associative conception in four steps. First, to identify the
need for an associative strategy of democratic reform, we note three barri-
ers to egalitarianism and indicate how each implicates questions of asso-
ciative order. Second, to underscore the potential contribution of groups
to democratic governance, we distinguish four general types of contribu-
tion and then draw from comparative experience to illustrate how the po-
tential has been realized in different areas of public policy. Third, we de-
fend the associative strategy for netting this contribution against two
objections: that it is impossible, because groups are intractable to reform;
and that it is undesirable, because the increase in group power needed to
secure contributions poses unacceptably high risks of group abuse of
power. Fourth and finally, we illustrate the associative strategy by discuss-
ing how it might be used to guide reforms of industrial relations and vo-
cational training in the United States.
1. Why Associative Reform?

Associative democracy aims to further an egalitarian-democratic view of
politics defined by simultaneous respect for norms of political equality,
popular sovereignty, distributive equity, and deliberative politics and the
operation of society for the general welfare. We interpret these norms in
the following ways. Political equality requires a rough equality across citi-
zens in their chances to hold office and to influence political choices.
Popular sovereignty requires that the authorization of state action be de-
termined (within the limits set by fundamental civil and political liber-
63
ties) by procedures in which citizens are represented as equals.5 Distribu-

tive equity obtains when inequalities of advantage, if they exist, are not
determined by differences of inherited resources, natural endowments,
or simple good luck. Collective choice is deliberative when it is framed
by different conceptions of the common good and public initiatives are
defended ultimately by reference to a conception of the public interest.
Society operates for the general welfare when there is both economic and
governmental efficiency.
These norms are routinely frustrated in the everyday politics of con-
temporary mass democracies. While there are many sources of frustra-
tion, here we note three that are of special relevance to our discussion of
associations.6
three p robl ems of ega l ita ri a n gov e rna nc e

The first problem is that government programs directed to achieving a
more equitable distribution of advantage (e.g., welfare services, active la-
bor-market policies, much economic and social regulation) are widely
perceived as unacceptably costly and inefficient. Whatever their theoreti-
cal attractions, critics assert, in practice such programs generate eco-
nomic rigidities, and a wasteful expansion of government aims beyond
government capacities. During a period of slowed productivity growth
and intensified economic competition, this makes egalitarianism at best
an unaffordable indulgence, at worst a betrayal of government obliga-
tions to “promote the general welfare.”
While claims of government inefficiency are often grossly exagger-
ated, they have sufficient basis in fact to give popular resonance to their
constant amplification.7 And especially in more liberal societies—where
choices about social governance are seen largely as choices between
5. This procedural formulation of the idea of popular sovereignty does not assume a people
with a single will and thus is immune to the criticisms directed against that assumption by, for
example, William Riker, Liberalism against Populism: A Confrontation Between the Theory of
Democracy and the Theory of Social Choice (San Francisco: W. H. Freeman, 1982).
6. Among the fundamental issues we will put to the side here are intense national and reli-
gious divisions and the destructive conflicts associated with them.
7. For discussion of some prominent exaggerations, see George W. Downs and Patrick D.
Larkey, The Search for Government Efficiency: From Hubris to Helplessness (New York: Ran-
dom House, 1986). In the United States, increased public doubt about government capacity to
achieve egalitarian ends is coincident with increased support for those ends. The “politics of
happiness” that some saw in the reformist projects of the 1960s has been succeeded by a “poli-
tics of sadness,” in which the public knows that it is not getting what it wants but has no con-
fidence that government can provide it.
64
states and markets, and no associative alternative is perceived—popular

acceptance of those claims is devastating to the practical pursuit of egali-
tarian ends. Most publics are unwilling to forgo economic growth in the
interest of equality. None enjoys literally wasting its tax dollars. So if state
programs are successfully defined as inimical to growth and wasteful, and
if market governance is the only alternative, egalitarianism is politically
doomed.
The second problem is that egalitarian efforts are deeply compro-
mised by representational inequalities. Capitalist property relations are, of
course, defined by inequalities in economic power, and political power is
materially conditioned. So economic inequalities characteristically trans-
late into political inequalities in violation of the norm of political equal-
ity. Until recently, however, at least in most rich, Western, liberal socie-
ties, it was possible to speak of a relatively steady advance in the social
democratization of capitalist societies. Gains in political equality accrued
from gains in the political representation of economically disadvantaged
interests.8
Today, any such optimistic assessment needs to be revised. Unions and
virtually all other mass popular organizations representing working peo-
ple are in palpable decline, while success in the organized representation
of the interests reflected in the “new social movements” of feminism, en-
vironmentalism, and racial justice is distinctly limited.9 With a widening
range of interests lacking an effective voice, the defining idiom of much
politics is not equality but exclusion.
The third problem is that those whose voice is organized often speak
with a strident particularism. On both sides of the many lines of privilege,
the narrow assertion of group interest is very nearly a norm. Whether moti-
vated by simple selfishness or fear of cooperation that comes from weak-
ness, the result is a politics of group bargaining that, undisciplined by
respect for the common good, inevitably conflicts with norms of popular
sovereignty and deliberative politics.10 Group particularism makes demo-
8. Many saw this as irreversible. See, for example, Jürgen Habermas, The Legitimation Cri-
sis of Late Capitalism (Boston: Beacon Press, 1973).
9. On unions, see Jelle Visser, “Trends in Trade Union Membership,” OECD Employment
Outlook, July 1991, 97–134.
10. For the American case, see the classic characterization of the resulting “interest group
liberalism” offered by Theodore J. Lowi, The End of Liberalism: The Second Republic of the
United States, 2d ed. (New York: Norton, 1979).
65
cratic governance more difficult, and it lessens the appeal of inclusive

politics by inspiring doubt that inclusion in fact enhances democracy.
The problems of government incompetence, political inequality, and
particularism feed one another. Inequalities in representation diminish
support for any egalitarian effort. The particularism of existing groups
prompts substantial reliance on statist means in those efforts. The adop-
tion of such means, even where ends could in theory be better accom-
plished by or with the aid of associations, compromises government ef-
ficiency. And the fact and the perception of government inefficiency,
working directly or through the consequent erosion of political support,
weaken those efforts and thus underscore inequality.
a n a ssocia t iv e st ra t egy of re f orm

The idea of associative democracy is to break this cycle by curing the
associative disorders that help to fuel it. Using conventional tools of pub-
lic policy (taxes, subsidies, legal sanctions), as applied through the famil-
iar decision-making procedures of formal government (legislatures and
administrative bodies, as overseen by the courts), it would promote asso-
ciative reform in each of the three problem areas.11 Where manifest in-
equalities in political representation exist, it recommends promoting the
organized representation of presently excluded interests. Where group
particularism undermines democratic deliberation or popular sover-
eignty, it recommends encouraging the organized to be more other-re-
garding in their actions. And where associations have greater competence
than public authorities for achieving democratic ends, or where their
participation could improve the effectiveness of government programs, it
recommends encouraging a more direct and formal governance role for
groups.
This last point may be the most immediate. In many areas of economic
and social concern—from the environment and occupational safety and
health to vocational training and consumer protection—egalitarian aims
are badly served by the state-market dichotomy that still dominates main-
stream debate about how those aims should be pursued. Often, the right
11. Throughout, respect for the associational liberties of group members, recognition of the
resistance of many groups to change, and rejection of concessionist views of associations mean
that the strategy stops well short of legislating associative practice or its relation to the state. As-
sociative democracy is not a distinct form of order, but a strategy to reform aspects of current
practice.
66
answer to the question “Should the state take care of the problem, or
should it be left to the market?” is a double negative.
This seems to be so in three ideal-typical classes of regulatory prob-
lems. In the first, nonmarket public standards on behavior are needed,
and government has the competence to set them, but the objects of regu-
lation are so diverse or unstable that it is not possible for the government
to specify just how those standards should be met at particular regulated
sites. Much environmental regulation presents problems of this sort. In
the second, public standard-setting is needed, and government has the
competence to do it, but the objects of regulation are sufficiently numer-
ous or dispersed to preclude serious government monitoring of compli-
ance. Consider the problems of occupational safety and health enforce-
ment. In the third, uniform public standards are needed, but it lies
beyond the competence of either markets or governments to specify and
secure them, as doing either requires the simultaneous coordination of
private actors and their enlistment in specifying the behavior sought.
Here, consider the difficulties of getting private firms to agree on stan-
dards for vocational training and to increase their own training efforts.
Where these sorts of problems are encountered, associative governance
can provide a welcome alternative or complement to public regulatory
efforts because of the distinctive capacity of associations to gather local
information, monitor behavior, and promote cooperation among private
actors. In such cases, the associative strategy recommends attending to
the possibility of enlisting them explicitly in the performance of public
tasks.
In sum, the idea of the associative strategy is to encourage the use of
associations to address concerns about unequal representation, partic-
ularism, and excessive cost and inefficiency of egalitarian programs and,
through that address, to satisfy more fully egalitarian-democratic norms.
In the next two sections, we will explore in more detail the features of as-
sociations that provide foundations for the strategy and underlie our as-
sessment of its promise.
2. The Potential Contribution of Groups

The cornerstone of the argument for associative democracy is that groups
have a significant contribution to make to democratic governance. In
the ordinary operation of mass democracies, groups are generally ac-
knowledged to be capable of performing at least four useful, democracy-
67
enhancing functions: providing information, equalizing representation,

promoting citizen education, and implementing alternative governance.
Information. Associations can provide information to policy makers on

the members’ preferences, the impact of proposed legislation, or the im-
plementation of existing law. As the state has become more involved in
regulating society and has extended the reach of its regulation to more di-
verse sites, technically complex areas, and processes subject to rapid
change, this information function has arguably become more important.
Good information is needed to assess the effectiveness of a myriad of
state policies, commonly operating at some distance from the monitoring
of state inspectorates, and to adjust policies to changed circumstances
or behaviors. This is especially so given social and policy interdepen-
dence—the interaction of social welfare policy and economic growth, for
example, or environmental regulation and technical change—which un-
derscores the value of accurate, timely intelligence on policy effects. Be-
cause of their proximity to those effects, groups are often well positioned
to provide such information. When they do, they contribute to satisfying
the norm of popular sovereignty, since good information improves citi-
zen deliberation, facilitates the enforcement of decisions, and clarifies
the appropriate objects of state policy.
Equalizing representation. Politics is materially conditioned, and in-

equalities in material advantage, of the sort definitive of capitalism, trans-
late directly to inequalities in political power. Groups can help remedy
these inequalities by permitting individuals with low per-capita resources
to pool those resources through organization. In making the benefits of
organization available to those whose influence on policy is negligible
without it, groups help satisfy the norm of political equality. Similarly,
groups can promote a more equitable distribution of advantage by cor-
recting for imbalances in bargaining power that follow from the unequal
control of wealth. Groups can also represent interests not best organized
through territorial politics based on majority rule. These include func-
tional interests, associated with a person’s position or activity within a so-
ciety; “categoric” interests of the sort pursued by the new social move-
ments; interests whose intensity is not registered in voting procedures;
and, at least in systems without proportional representation, the interests
68
of political minorities. Here, groups improve an imperfect system of in-

terest representation by making it more fine grained, attentive to prefer-
ence intensities, and representative of diverse views. This, too, furthers
political equality.
Citizen education. Associations can function as “schools of democracy.”

Participation in them can help citizens to develop competence, self-
confidence, and a broader set of interests than they would acquire in a
more fragmented political society. Alexis de Tocqueville provides the
classic statement of this educative power of associations: “Feelings are re-
cruited, the heart is enlarged, and the human mind is developed only by
the reciprocal influence of men on one another,” and under democratic
conditions, this influence can “only be accomplished by associations.”12
In performing this educative function, associations help foster the “civic
consciousness” on which any egalitarian order, and its deliberative poli-
tics, depends. That is, they promote a recognition of the norms of demo-
cratic process and equity and a willingness to uphold them and to accept
them as fixing the basic framework of political argument and social coop-
eration, at least on condition that others do so as well.
Alternative governance. Associations can provide a distinctive form of so-

cial governance, alternative to markets or public hierarchies, that permits
society to realize the important benefits of cooperation among member
citizens. In providing a form of governance, associations figure more as
problem solvers than simply as representatives of their members to au-
thoritative political decision makers, pressuring those decision makers on
behalf of member interests. They help to formulate and execute public
policies and take on quasi-public functions that supplement or supplant
the state’s more directly regulatory actions.
Such associations facilitate cooperative dealings in two ways. First,
their sheer existence reduces the transaction costs of securing agreement
among potentially competing interests. The background of established
forms of communication and collaboration they provide enables parties
to settle more rapidly and reliably on jointly beneficial actions. Second,
groups help to establish the trust that facilitates cooperation. They effec-
12. Alexis de Tocqueville, Democracy in America (New York: Vintage, 1945), vol. 2, p. 117.
69
tively provide assurances to members that their own willingness to coop-

erate will not be exploited by others. Often directly beneficial to society,
associative governance can also support public efforts to achieve egalitar-
ian aims.
Lessons from comparative experience. While examples of all these sorts of

group contributions can be found in the United States, in recent years it
is students of comparative politics, in particular the politics of Western
Europe, who have been especially attentive to these positive features of
associations. They have argued more particularly that certain sorts of
group organization play a central role in resolving, in egalitarian fashion,
problems of successful governance in mass democracies.
The rediscovery in the 1970s of liberal “corporatist” systems of interest
representation in Northern European democracies was the key to one
such argument.13 Students of liberal corporatism suggested that the in-
corporation of organized interests into the formation of economic pol-
icy helped produce, simultaneously, better satisfaction of distributive
concerns, improved economic performance, and gains in government ef-
ficiency. Of particular note was the negotiation and compromise be-
tween organized business and organized labor within such systems,
which appeared to permit their joint realization of many gains from co-
operation.
The Scandinavian social democracies of Norway and Sweden provided
a particularly advanced example of such labor-business cooperation.
There, encompassing union and employer federations, both speaking for
virtually all of their respective populations of interest, would meet regu-
larly to negotiate the terms of their essentially peaceful coexistence, with
the state serving to ratify and support those terms. Unions exchanged
wage restraint for guarantees of low unemployment and a high social
wage. Employers traded employment security and industrial upgrading
for union moderation. The state, backed by both “social partners,” cali-
brated fiscal policy to stabilize employment, social policy to provide in-
13. See Philippe C. Schmitter, “Still the Century of Corporatism?,” Review of Politics 36
(1974): 85–131; Suzanne Berger, ed., Organizing Interests in Western Europe: Pluralism,
Corporatism, and the Transformation of Politics (Cambridge: Cambridge University Press,
1981); and John H. Goldthorpe, ed., Order and Conflict in Contemporary Capitalism (Oxford:
Clarendon Press, 1984).
70
surance against market misfortune, and industrial policy to maintain

competitiveness in foreign markets.14
More recent discussions, even as they have dissented from claims
made about corporatism or have paused to note its devolution or col-
lapse, have also stressed the importance of associative activity to eco-
nomic performance. Students of the successful alternatives to mass
production that are marked, simultaneously, by high wages, skills, pro-
ductivity, and competitiveness have argued that this success requires a
dense social infrastructure of secondary association and coordination.
This organizational infrastructure provides the basis for cooperation be-
tween management and labor, among firms, and between firms and gov-
ernment on issues of work organization, training, technology diffusion,
research and development, and new product ventures. And that coopera-
tion, it is argued, is essential to ensuring economic adjustment that is
both rapid and fair.15
The reemerging (or more newly visible) regional economies of West-
ern Europe—Italy’s Emilia-Romagna, Sweden’s Småland, Germany’s
Baden-Württemburg, Denmark’s Jutland peninsula—provide particu-
larly striking examples of such associative economic governance. They
feature complex public-private partnerships on training and technology
diffusion, flexible manufacturing networks that facilitate inter-firm coop-
eration in performing discrete and varied production tasks, more formal-
ized consortia and industry associations to realize economies of scale in
some functions (e.g., marketing or research and development) among
otherwise competing firms, joint training activities among firms, occupa-
tional credentialing of labor through industry-wide or regional labor and
14. For useful description and analysis of such coordination in Scandinavia, see Walter
Korpi, The Democratic Class Struggle (London: Routledge and Kegan Paul, 1983); Gøsta
Esping-Andersen, Politics against Markets (Princeton: Princeton University Press, 1985); for a
good comparative treatment of the Swedish and German cases, and the role played by
corporatist institutions in facilitating wage stability and industrial upgrading, see Peter
Swenson, Fair Shares: Unions, Pay, and Politics in Sweden and West Germany (Ithaca: Cornell
University Press, 1989); and Lowell Turner, Democracy at Work: Changing World Markets and
the Future of Labor Unions (Ithaca: Cornell University Press, 1991). For a general review of
problems that have beset social democracies since the mid-1970s, see Fritz W. Scharpf, Crisis
and Choice in European Social Democracy (Ithaca: Cornell University Press, 1991).
15. See Charles F. Sabel, “‘Flexible Specialization and the Re-emergence of Regional Econ-
omies,” in Paul Q. Hirst and Jonathan Zeitlin, eds., Reversing Industrial Decline: Industrial
Structure and Policy in Britain and Her Competitors (Oxford: Berg, 1989), 17–70; and Wolfgang
Streeck, “On the Institutional Conditions of Diversified Quality Production,” in Egan
Matzner and Wolfgang Streeck, eds., Beyond Keynesianism: The Socio-Economics of Produc-
tion and Employment (London: Edward Elgar, 1991), 21–61.
71
management associations, and close linkages between regional develop-

ment and welfare policies. Indeed, the object of state economic develop-
ment policy in most of these regions prominently includes efforts to
build the private associative framework upon which such efforts rely.16
For an example of the sorts of associations being fostered, consider
CITER, an association of small knitwear firms in the town of Carpi, in
Emilia-Romagna. Its six hundred dues-paying member firms are gener-
ally tiny, averaging fewer than eight workers each. But by pooling re-
sources in the association itself, as well as in countless joint-production
schemes facilitated by the association, they are able to flourish in the
fiercely competitive and unstable business of international fashion.
Through CITER, they share information on trends in technology, pro-
duction processes, and emerging markets, underwrite a sophisticated
forecasting service on fashion trends, gain access to and training in the
use of sophisticated business software, and enjoy other services no one
firm could afford on its own. CITER is not a cartel. Its member firms still
compete with one another. They simply do not forsake the obvious gains
to all that can come from associative cooperation.17
The virtues of associative forms of governance are, however, not con-
fined to economic cooperation. Associative governance has also been
credited with achieving more effective social regulation and welfare de-
livery. Within the heavily procedural and litigious “command and
control” regulation favored in more liberal systems and particularly dom-
inant in the United States, groups commonly appear to frustrate regula-
tory efficiency. Evidence from systems in which associations are assigned
a more central and open governance function, however, suggests that
they can powerfully contribute to the success of regulatory programs.
Instead of acting only or chiefly as “special interests” intent on either cap-
turing public powers or limiting their efficacy, groups supplement tradi-
tional public authority by helping to define policy, to monitor its imple-
16. For examples of state policy, see Stuart A. Rosenfeld, Technology Innovation and Rural
Development: Lessons from Italy and Denmark (Washington: Aspen Institute, 1990). We em-
phasize that state policy is needed in all these cases. The appropriate infrastructure does not
emerge naturally from the interactions of economic actors or from favorable cultural tradition.
For further discussion, see Section 3 below.
17. For this and other examples of “flexible manufacturing networks,” see C. Richard Hatch,
Flexible Manufacturing Networks: Cooperation for Competitiveness in a Global Economy
(Washington: Corporation for Enterprise Development, 1988).
72
mentation, and to enforce it. Rather than acting as obstructions, they

serve as private multipliers on public capacities.
Associations have been shown to play this role for a wide range of regu-
latory purposes, extending from the enforcement of occupational safety
and health, wage and hour, and environmental regulation to the promo-
tion of curricular reform and better learning opportunities in education
and training systems. Admitting variations in national success, the gen-
eral result appears to be a style of regulation and the affirmative promo-
tion of egalitarian ends, at once more effective, flexible, and efficient
than command and control or simple state administration of programs.
Consider occupational safety and health. Instead of relying exclusively
on a centralized state inspectorate to enforce occupational safety and
health laws, virtually all European systems supplement their inspector-
ates with mandated workplace health and safety committees. These com-
mittees operate with delegated public powers: they monitor, and in some
measure are empowered to enforce, compliance with the regulatory re-
gime. While bringing new costs in its train (e.g., the costs of training
worker deputies), the general result of this strategy is a health and safety
policy more effective and efficient than an inspectorate-alone approach.
It is more effective because it supplements public capacities for monitor-
ing compliance with the capacities of workers themselves. It is more ef-
ficient because it permits public efforts to be left largely to standard set-
ting and enlists the local knowledge of regulated actors in devising the
least costly means, in particular settings, of satisfying such standards.18
Or consider the use of associations in education. A striking example is
provided by the German system of youth apprenticeship. Employer asso-
ciations and unions determine training standards and requirements,
monitor the provision of training at both school and work, and provide
much of the workplace-based instruction. The role of the state is essen-
tially to inform the social partners about emerging labor-market trends,
ratify the results of their deliberations, help enforce the occupational
18. For a review of worker participation in safety regulation focusing on Europe, see the con-
tributions to Sebastiano Bagnara, Raffaello Misiti, and Helmut Wintersberger, eds., Work and
Health in the 1980s: Experiences of Direct Workers’ Participation in Occupational Health
(Berlin: Edition Sigma, 1985); for a particularly useful country study, see Bjørn Gustavsen and
Gerry Hunnius, New Patterns of Work Reform: The Case of Norway (Oslo: Universitetsforlaget,
1981); for the contrast with the United States, see Charles Noble, Liberalism at Work: The Rise
and Fall of OSHA (Philadelphia: Temple University Press, 1986); and Eugene Bardach and
Robert Kagan, Going by the Book (Philadelphia: Temple University Press, 1982).
73
standards that result, and encourage widespread participation in the asso-

ciative effort. From top to bottom, the system is driven by the associa-
tions, albeit acting in concert with public authority. The result is gener-
ally recognized as the most successful and inclusive vocational training
program in the developed Western world.19
Taken together, these different investigations and examples underscore
the range of important contributions associations can make to a function-
ing democratic order. What assures that contribution, moreover, is not
the sheer “quantity of associability” found in such systems but the care of
public authorities within them in matching the qualitative characteris-
tics of different groups to public functions and in working with groups to
encourage the appropriate qualitative characteristics.20 The deliberate
conditioning of state fiscal and welfare assistance on the outcomes of
wage bargaining under corporatism; the explicit state efforts to build the
associative infrastructure of regional economies; the laws mandating the
workplace safety committees; the support provided by the German state
to the social partners in education: here we have examples of the sort of
public encouragement of appropriate group forms recommended by the
associative strategy.
3. Impossibility and Undesirability: A Response to Objections

Thus far we have discussed problems of government incompetence,
political inequality, and particularism that now thwart egalitarian-
democratic politics; we have proposed that a partial remedy for those
problems lies in an improved organization of secondary associations pur-
sued through a politics of associations; and we have presented some ana-
lytical considerations and comparative experience to support and illus-
trate our proposal. We want now to consider a pair of related objections
to it. Both objections accept (at least for the sake of argument) the attrac-
tiveness of egalitarian-democratic norms and both agree that associations
can contribute to the satisfaction of those norms. But they reject the use
19. For a close examination of the different public powers enjoyed by the “social partners” in
the German case, see Wolfgang Streeck, Joseph Hilbert, Karl-Heinz van Kevelaer, Frederike
Maier, and Hajo Weber, The Role of the Social Partners in Vocational Training and Further
Training in the Federal Republic of Germany (Berlin: European Center for the Development
of Vocational Training, 1987).
20. The phrase and the point come from Philippe C. Schmitter, “Interest Intermediation
and Regime Governability in Contemporary Western Europe and North America,” in Berger,
ed., Organizing Interests, 285–327.
74
of an associative strategy to engender the “right” sort of associative envi-

ronment. According to the first objection, it is not possible to create a fa-
vorable associative environment through politics; according to the sec-
ond, efforts to create such an environment are more dangerous than the
disease they aim to cure.
imp ossibil it y
The argument for impossibility begins with the assumption that groups
are a product of nature or culture or some other unalterable substrate of a
country’s political life. Just as some countries are blessed with good top-
soil or a temperate climate, others are blessed with the “right” kinds of
groups, at the right level of organization. In countries that are so blessed,
group contributions of the sort we note are observed. But since patterns
of group organization and behavior lie beyond politics, the observation
provides no support at all for an associative strategy for addressing the
problems of egalitarianism. Indeed, precisely by highlighting the impor-
tance of a favorable social basis for egalitarian democracy, they explain
why equality does not travel well.
We think that this objection exaggerates the fixity of the associative en-
vironment. Groups are, after all, in important ways political artifacts.
Their incidence, character, and patterns of interaction are not merely the
result of natural tendencies to association among citizens with like prefer-
ences. They reflect structural features of the political economy in which
they form—from the distribution of wealth and income to the locus of
policy making in different areas. And they reflect variations across the
members of that society, along such dimensions as income, information,
and density of interaction. Existing political institutions and “culture”
may crystallize around certain structural features and patterns of varia-
tion along these dimensions. But those features and variations are in no
sense natural. They can be changed through public policy.
Public policy can, for example, make the background distribution of
wealth and income more or less uneven. It can shift the locus of public
decision making from regional to national levels or concentrate it in a
single department, in ways that encourage different sorts of group forma-
tion and discourage others. The availability of information can be wid-
ened or constricted. The density of interaction among similarly situated
citizens can be increased or decreased. The cost of administering joint ef-
forts, or navigating the negotiation antecedent to them, can be subsidized
75
or not. Those subsidies can simply be provided to the most powerful or

tied to antecedent satisfaction of certain requirements of behavior. Con-
sistent with the continued supremacy of formal political institutions,
groups can also be assigned public functions, including the power to is-
sue complaints for violations of administrative regulation, to take emer-
gency action in correcting violations, to establish standards for licensing
and training in different occupations and industry standards on produc-
tion, to establish eligibility criteria for receipt of other sorts of benefits
(including welfare benefits), and to apply such licensing procedures,
standards, and eligibility criteria as part of a general regulatory regime.
All such changes in the environment of group formation, the incen-
tives available to individual groups, and the governing status of groups
can manifestly change the group system.
The experience of countries that are now recognized as having the
“right” kinds of groups, moreover, bears out the importance of such de-
liberate efforts to shape the group environment. While corporatist sys-
tems of wage bargaining and peak negotiation may have benefited from
preexisting religious solidarities, they were commonly built, deliberately,
on the wreckage of much more contentious industrial relations. While
regional economies may be furthered by the social linkages of indepen-
dent agrarian communities, today those linkages are fabricated by efforts
to seed joint projects and lower information costs. While apprenticeship
vocational training may draw on longstanding traditions of craft produc-
tion and employer obligation, the organizational base of such training,
and the base of craft production itself, is secured through legally required
memberships in organizations and protection of small producers. There
is nothing “natural” about such efforts to secure appropriate associative
ends and nothing in “nature” that has precluded their success.
undesira bil it y
Still, efforts to enlist associations in democratic governance may be unde-
sirable. While groups can contribute to democratic order, and while
their contribution can be secured through public policy, they can also
work to undermine democratic order. This threat of “faction” was evi-
dent in our own inventory, offered earlier, of the practical problems now
faced by democratic egalitarianism. Each problem suggested an impair-
ment of democracy produced by the existing system of secondary associa-
tion. If our associative strategy entails the further cultivation of groups,
76
and the ceding to them of further public powers, does it not risk making
faction truly ruinous?
Before addressing this question, we need to enter some background re-
marks aimed at clarifying the issues it raises.
The problem of faction has been a particular preoccupation of Ameri-
can politics and democratic theory ever since James Madison announced
it as the key issue of American constitutional design.21 But it must be
faced by any liberal order, by reason of one of the defining features of
such order: the protection of associative liberties. Once associative liber-
ties are protected, associations inevitably follow. And, inevitably, legiti-
mately, and without malfeasance, some of those associations will use
their powers in pursuit of their aims in ways that frustrate the satisfaction
of basic democratic norms. They will represent members in ways that un-
dermine political equality; they will capture areas of policy in ways that
undermine popular sovereignty and promotion of the general welfare; in
“doing their job” of advancing member interests, they will inevitably pro-
mote particularism in place of deliberative politics. The threat of faction
is, then, inescapable in any regime with associative liberties. Moreover,
since those liberties are fundamental, the issue is how to mitigate that
threat, not how to remove it.
The characteristic forms of faction were suggested earlier. There is,
first, a pathology of inequality. Given inequalities in organization arising
naturally from the background of market capitalism, group efforts to
represent the interests of members may simply compound political in-
equality rather than relieve it. Political inequality may then compound
material inequality, as groups use their political powers to improve their
material position, in a vicious cycle of privilege. Second, there is a pa-
thology of particularism. Groups are, by their very nature, to some degree
particularistic. Only some citizens are represented in them, group lead-
ers are (at best) accountable to their members and not others, and the in-
terests and ideals of groups are not shared by all citizens. Representing
their members faithfully, particular groups thus often seek policies that
impose costs to the society at large, even as they provide gains for their
own members, and promote a politics of narrow advantage and bar-
gaining that corrupts the ideal of public deliberation about the com-
mon good. Commonly, inequality and particularism both thrive, as over-
21. See James Madison, Federalist 10, in The Federalist (New York: G. P. Putnam’s Sons,
1907), 51–60. We are concerned here only with what Madison called “minority” faction.
77
represented interests bargain with one another, divide the political spoils,
and so preserve their privileges until the next round of bargaining begins.
The problem of faction is serious, then; it is also inevitable so long as
associative liberties are preserved. Since threats of faction are inevitable,
it would be a mistake to attribute them to the associative strategy or to ex-
pect that strategy to eliminate faction. But since those problems are seri-
ous, it would be objectionable if the associative strategy increased the
threat of faction. The question raised by the second objection, then, is
whether pursuit of associative strategy would make the problem of fac-
tion worse.
To address this question fully, we would need to consider the likely ef-
fects of the associative strategy on each of the defining norms of egalitar-
ian democracy: popular sovereignty, political equality, distributive equity,
deliberative politics, and the operation of society for the general welfare.
In the interests of space, we propose to focus here solely on the norm of
popular sovereignty, though our treatment of it will suggest the shape of
our more general response. Recall that that norm requires that the autho-
rization of state action be determined (within the limits set by fundamen-
tal civil and political liberties) by procedures in which citizens are repre-
sented as equals. Our question then becomes: Would the pursuit of our
associative strategy undermine the ultimate authority of the people in the
formation of policy?
In answering this question, we assume that all associations, includ-
ing those vested with quasi-public powers, will operate within a political
system with encompassing formal institutions organizing representation
along traditional territorial lines. We assume, then, a possibility of “exit”
from the group-based system of interest representation to the more tradi-
tionally organized system. Moreover, we assume that the group system is
itself regulated by the traditional system. Final formal authority resides
with traditional institutions. Associations will depend on them for autho-
rizations of certain of their powers and for material support in carrying
such authorizations out.
With these background assumptions in mind, we want first to indi-
cate four sorts of positive-sum relationships between associations and the
democratic state—four ways, that is, that the fuller and more explicit in-
corporation of groups into governance roles might actually enhance the
exercise of popular sovereignty through the traditional institutions and
practices of territorial representation.
78
First, groups provide the state with information, thus permitting better
definition of problems and greater precision in the selection of means for
addressing them. By thus sharpening policy instruments and enabling
them to be applied with greater precision, groups promote the capacity
of the people to achieve their aims. Second, groups provide additional
enforcement power, thus increasing the likelihood that decisions made
by the people will be implemented.22 Third, in mitigating enforcement
problems, groups remove one important constraint on political debate.
Instead of proposals being short-circuited with the claim that they are
unenforceable, a wider range of proposals can be seriously discussed.
Fourth, a more open politics of associations makes explicit a condition,
which is already a standing feature of even the most liberal of societies,
namely, that secondary associations do in fact perform a variety of func-
tions that affect the conditions of political order. The associative strategy
“exposes and brings out into the open, it institutionalizes a factor in law-
making that we have, eagerly in fact, attempted to obscure.”23 By bringing
the role of associations “into the open,” it would make the exercise of
power by associations more accountable. In combination, better and
more flexible means, better enforcement, less constrained debate about
ends and their achievement, and more openness and accountability in
the exercise of power all count as important gains for popular sovereignty.
These four contributions are, however, accompanied by three sources
of serious concern—of negative-sum relations between the powers of as-
sociations and the egalitarian-democratic order.
First, there are problems of disjunction of interest between the leader-
ships of groups and their members—the problem of the “iron law of oli-
garchy.” A dense world of association may make the government more in-
formed about, and more responsive to, the interests of group “oligarchs”
but not those of its members. Second, there is the problem of indepen-
dent powers—what might be called the “Frankenstein” issue. Endowed
with quasi-public status, and commonly subsidized by the state, groups
that at one point in time contribute to decent policy may continue to ex-
ercise power after outgrowing their usefulness, use that power to freeze
their position, and so work to distort future debate and choice. Third, in-
22. See, for example, the discussion of “fire alarm” enforcement in Mathew D. McCubbins
and Thomas Schwartz, “Congressional Oversight Overlooked: Police Patrols vs. Fire Alarms,”
American Journal of Political Science 28 (1984): 165–179.
23. Louis Jaffe, “Law-Making by Private Groups,” Harvard Law Review 51 (1937): 220–221.
79
creasing the extent of policy making outside of formal legislative arenas

increases threats of improper delegation. In particular, powers delegated
to associations are bound to be vague. As in the context of legislative dele-
gations to administrative agencies, then, there are problems about the
abuse of the discretion permitted by such vagueness.
What are we to make of these problems? To make the case for associa-
tive democracy, it should be clear, we do not need to show that the strat-
egy will solve these problems. They already exist and will remain in
place so long as freedom of association is guaranteed. It is enough to
show that associative democracy will not plausibly make the problems
worse. Moreover, if the same deliberate politics of association that har-
nesses group contributions can mitigate the threat of faction, that should
count as an added support for the argument. In considering the three
problems just noted, it appears to us that this burden can be carried and
that the promise of actual advance on curbing faction can be redeemed.
Beginning with internal democracy, the chief threat of the associative
strategy appears to be its potential encouragement of large, encompass-
ing, bureaucratic associations of the sort capable of taking broad respon-
sibility for the coordination of social interests. These, it might be thought,
are likely to suffer from even greater problems of internal responsiveness
than the existing population of organizations. A recurrent example used
in critical discussions is the distant, professionalized leadership of cen-
tralized trade union federations, whose “social responsibility” in dealings
with employers and the state is seen to come at the expense of the con-
cerns of actual members.
Given the decline of centralized union bargaining, the example may
be of diminished empirical relevance. But it suffices to carry the con-
cern. And it remains an instructive test of the intuitive assumption that
responsiveness of leadership to group membership must decline as group
encompassingness, size, and social responsibility increase. For, in fact, it
suggests that the intuitive assumption is without foundations. There is no
correlation between the opportunities for voice and exit that encourage
responsiveness and the conditions necessary for peak bargaining. On a
variety of measures of internal union democracy, for example, the Nor-
wegian union movement, among the most centralized and encompass-
ing in the world, is more internally democratic than unions in the United
Kingdom, comprising one of the least centralized union movements,
80
which are in turn more democratic than the unions of West Germany,
which are intermediate in their level of centralization.24
If the union case is credited, internal responsiveness need not come at
the expense of external capacity. Moreover, internal responsiveness can
be designed into large organizations through their internal procedures.
In combination, these points suggest that oligarchy is more plastic than
the “iron law” suggests. More immediately, they suggest a natural re-
sponse to the problem of disjunction: require greater use of internal dem-
ocratic procedures among groups that are granted quasi-public status.
Operationally, the requirement should be that groups accorded this sta-
tus provide evidence that they in fact represent their members by show-
ing that they actually use some mechanism of responsiveness. Infinite
gradations in degree and differences in judgment are certainly imagina-
ble here, just as they are in ongoing disputes over the representativeness
of electoral systems. But as the case of electoral systems also suggests, it is
possible to articulate a general principle of legitimacy, in this case inter-
nal responsiveness, and to use that general principle to guide debate
about specific proposals.
Our second problem, the “Frankenstein” problem of independent pow-
ers, also carries a natural response, namely some variant of “sunset legis-
lation.” The quasi-public status of groups (and subsidies to them) should
be reviewed on a regular basis, with a rebuttable presumption that the
status (or subsidies) will be withdrawn or amended as group behavior, or
perceived social needs, warrants. The general requirements are reason-
ably clear, though their precise elaboration is not. On the one hand, the
threat of withdrawal must be sufficiently credible, and the gains associ-
ated with public status sufficiently great, to induce groups to meet ac-
countability requirements and other conditions on their conduct. On the
other hand, since continuity in bargaining relations is an important pre-
requisite of gains, the requirements must not be so exacting as to make
them impossible to satisfy.
Of course, the ultimate guard against independent powers is the vital-
24. See Peter Lange, Union Democracy and Liberal Corporatism: Exit, Voice, and Wage Reg-
ulation in Postwar Europe, Cornell Studies in International Affairs, Occasional Paper No. 16.
The measures include rules governing election to union councils, intermediate organizations,
and national office; incidences and support of informal caucuses; and procedures for debate
and vote on strikes, contracts, and other sorts of concerted action.
81
ity of the system dispensing the powers in the first place. This fact is pre-
cisely what gives normative force to our assumption, above, that systems
relying heavily on group-based representation still rest final authority in
encompassing territorial organizations. For evaluating associative democ-
racy, the narrow issue here is whether, ceteris paribus, that system is made
more or less vital by the increase in its democratic capacity that would
follow on its enlistment of the energies of representative groups. And to
ask that question is to answer it.
Finally, we offer two thoughts on the third problem noted above: the
problem of vague delegations of power and the attendant risks of abused
discretion. The first of these is simply a plea for realism and fairness in
evaluation. The threat of vague delegations of powers in our associative
scheme should be contrasted not with some ideal world but with the one
that exists and the alternative reform proposals for that world. When it is,
the contrast does not seem particularly damning. In the existing world,
there is already much vague delegation to and exercise of discretion by
administrative agencies. lf we consider a scheme of more limited govern-
ment as a means to cabin discretion, then we need to keep in mind that
such a scheme is unlikely to serve the egalitarian democratic aims at is-
sue here. If we consider a scheme with stronger legislative controls—less
vagueness in delegating and more sharpness in formulating legislative
standards—then we should consider familiar cautions that it may lead to
an unwelcome politicization of legislative instruction, reflected in unrea-
sonable goals, improbable deadlines on their achievement, or simple leg-
islative deadlock.25 Nor is there any reason to think that such reasonable
requirements as clarity in the statement of statutory goals will be incon-
sistent with the associative scheme.
Moving now to a more positive engagement with the issue, we pro-
pose to address the problem of delegation through performance crite-
ria. Where associations are involved in the enforcement and administra-
tion of policy, public institutions should formulate clear performance
standards for groups to enforce and administer, while avoiding detailed
specification of the means to be used in meeting those standards. For ex-
ample, in the area of workplace health, there might be performance stan-
25. These effects are noted in Cass Sunstein, “Constitutionalism after the New Deal,” Har-
vard Law Review 101 (1987): 480–481: “The movement toward increased congressional control
is not without risks of its own [since] . . . undue specificity may produce regulation riddled by
factional tradeoffs.”
82
dards in the form of permissible exposure limits for hazardous chemicals,

while decisions about the means for implementing those limits would
fall to health and safety committees. When associations are involved in
the formation of policy, the discretion ingredient in grants of quasi-public
status can again be addressed by setting performance criteria—for exam-
ple, minimum standards for skills, knowledge, courses, and examinations
in vocational training programs whose operation is coordinated by labor
and business in particular sectors. Even where groups do not enjoy subsi-
dies for their performance of quasi-public duties, they should be regu-
lated in the conduct of those duties. Where they are officially granted
quasi-public status, or material state assistance, performance criteria can
be more exacting.
In sum, then, our response to the undesirability objection is that dan-
gers of faction in the area of popular sovereignty could be mitigated by
requirements on internal democracy, legislative and judicial oversight,
sunset laws that threaten a group with competition for its position, and
performance standards. Moreover, we think that similar measures of in-
ternal accountability, external oversight, and competition could be de-
ployed to mitigate problems of faction that arise on the other dimen-
sions of democracy (political equality, for example). But we have not, of
course, argued this here, and to that extent the discussion of faction is im-
portantly incomplete. It might, for example, be argued that an associative
strategy for equalizing political representation would generate cartels or
other concentrations of economic power that would, in turn, present in-
tolerable threats to economic efficiency. We disagree with this objection
and think that some of the comparative evidence discussed in Section 2
speaks against it. Nevertheless, we think it raises a serious problem and
that a fuller discussion of the associative idea would need to show in de-
tail how it could be met.
4. The American Case

Thus far we have argued that associative solutions are, in the abstract, at-
tractive ways of advancing democratic ideals and that the factional poten-
tial of such solutions can be tamed by the same strategy of constructive
artifice that enlists group contributions. Still, the idea of associative de-
mocracy may seem of little relevance to the United States. More than
any other economically advanced mass democracy, the United States has
a strongly anti-collectivist political culture, a weak state, and a civil soci-
83
ety dominated by (relatively disorganized) business interests. The poten-

tial for artifice granted, this context poses obvious problems for the asso-
ciative strategy. At best, it might be thought, the absence of any initial
favoring conditions makes the strategy irrelevant. There is simply not
enough to get started down the path of democratic associative reform. At
worst, it might be feared, pursuit of the strategy under these conditions
would be a political nightmare. Giving new license to a congeries of
group privilege and particularism, it would exacerbate inequalities and
further corrupt and enfeeble the state.
Such concerns have considerable force and deserve a fuller answer
than we can provide here. Briefly, however, while we acknowledge the
anti-collectivism of much American political culture, we also see consid-
erable experimentation now going on with associative solutions to policy
problems in such areas as regional health and welfare service delivery, lo-
cal economic development, education and training, and environmental
regulation, among many others.
There is, for example, a tradition of delivering many welfare and so-
cial services through secondary associations—community organizations,
churches, volunteer agencies, and the like. While such organizations of-
ten have substantial autonomy in designing the appropriate service mix
for the communities they are asked to serve, they are also increasingly in-
extricably dependent on government fees for such services for their own
survival.26 Much “public” input in local economic development is de-
cided, for good or ill, in “community development corporations,” heavily
subsidized by government grants, representing different admixtures of in-
dependent neighborhood associations and business firms.27 In education,
parent-teacher associations are commonly vested with substantial powers
in determining the budget and curriculum of elementary and secondary
public schools, and those schools increasingly look to local business in-
terests for support in setting standards on student performance.28 In train-
26. For an instructive discussion of the role of nonprofit organizations in welfare-state ser-
vice delivery, emphasizing the increased dependence of many of these agencies on their ties to
government, see Steven Rathgeb Smith and Michael Lipsky, Nonprofits for Hire: The Welfare
State in the Age of Contracting (Cambridge, MA: Harvard University Press, 1995).
27. A useful (though not impartial) survey of local economic development strategies is pro-
vided in R. Scott Fosler, Local Economic Development (Washington: International City Man-
agement Association, 1991).
28. For an enthusiastic review of some of the emerging linkages between schools and private
business associations, see Anthony Carnevale, Leila Gainer, Janice Villet, and Shari Holland,
Training Partnerships: Linking Employers and Providers (Alexandria: American Society for
Training and Development, 1990).
84
ing, the largest single training program in the United States, the Job
Training Partnership Act (JTPA), is almost wholly administered through
“private industry councils” dominated, by statute, by local business inter-
ests.29 In environmental regulation, from the deliberate promotion of bar-
gaining among industry and environmental groups as a prelude to stan-
dard setting at the federal level, to the promotion of bargaining between
business and community organizations over the appropriate implementa-
tion of environmental standards in local neighborhoods and regions, pol-
icy is rife with secondary associations exercising de facto public powers.30
Some of these efforts display the great strengths of associative gover-
nance; others display its many dangers. Our point here is simply that
such governance in fact goes on, widely, even in this liberal culture. And
its incidence provides a natural basis for more deliberate, and demo-
cratic, associative strategies.
Moreover, while we acknowledge the weakness of the American state,
we think that at least some sorts of associative reforms can make it stron-
ger. Particularly given a weak state, it is important that group empower-
ment proceed in a way that is reliably positive-sum with state power. But
this merely requires judgment in the choice of associative strategies. It
does not generally bar pursuit of them. And while we acknowledge,
finally, the overwhelming business dominance of the American polity,
we think this again simply constrains choice in the groups that are advan-
taged through the associative strategy. If business is too powerful, then as-
sociative resources should be provided to labor or other nonbusiness-
dominated groups; the current imbalance is not an argument for aban-
doning the general idea.
Most generally, we agree that the United States has high levels of in-
equality, a less-than-competent government, and weak cooperative insti-
tutions—that, in brief, it does not work well as a democracy. This, in fact,
29. JTPA has been widely criticized as insufficiently accountable to public needs. Among
others, see John D. Donahue, Shortchanging the Workforce: The Job Training Partnership Act
and the Overselling of Privatized Training (Washington: Economic Policy Institute, 1989);
United States General Accounting Office (GAO), Job Training Partnership Act: Inadequate
Oversight Leaves Program Vulnerable to Waste, Abuse, and Mismanagement, GAO/HRD-91–
97 (Washington: General Accounting Office, 1991).
30. Some of the federal experience is reviewed in Charles W. Powers, The Role of NGOs in
Improving the Employment of Science and Technology in Environmental Management (New
York: Carnegie Commission on Science, Technology, and Government, May 1991); the experi-
ence of local communities in fostering such environmental bargaining among organized
groups is reviewed in Valjean McLenighan, Sustainable Manufacturing: Saving Jobs, Saving
the Environment (Chicago: Center for Neighborhood Technology, 1990).
85
is the very problem that provides our point of departure. We move, then,
to some examples of how an associative strategy might proceed from this
point of departure in this distinctive polity. We offer two illustrations of
the general look and feel of the associative project: the reform of worker
representation and industrial relations in the United States, and the re-
form of vocational training. In each case we sketch some problems that
need to be addressed, indicate the ways that a richer associational setting
might help in addressing them, and discuss some measures that might
now be taken to promote that setting.
worker rep resenta t i on

Our goal here—controversial and surely bitterly contested—would be
to improve the organization of American workers. Such improvement
would plausibly contribute to the satisfaction of democratic norms in a
variety of ways. By extending and deepening the benefits of organized
representation to those who are now unorganized or under-organized, it
would advance the goal of political equality. It would also have a fair
chance of improving distributive equity and economic performance in
the United States. At the same time, properly structured worker organiza-
tion is of particular importance, because work is important. The associa-
tive framework that determines how it is organized, distributed, and re-
warded sets the background and tone for associative action throughout
much of the society. So other reforms are more likely to succeed if re-
forms here succeed.31
The system of worker organization in the United States currently suf-
fers from two related problems. First, very few substantive benefits are
provided to workers in their role as citizens. We have a low “social” wage.
Most benefits are instead provided through individual firms. But benefits
are costly and firms compete. So there are obvious incentives to skimp on
the provision of benefits. The result is comparatively low and uneven
substantive protection for workers.
Second, the system discourages cooperation between employers and
employees. Part of the reason for this is the generally low level of worker
organization. Genuine cooperation is based on mutual respect, which
typically depends on recognition of mutual power. With the disorganiza-
tion of workers limiting their power, however, employees are commonly
31. The force of this claim will emerge in our discussion of the role of associations in voca-
tional training.
86
incapable of extracting from employers the sorts of institutionalized re-

spect for their interests (e.g., a serious commitment to job security or
consultation in advance of work reorganization) needed to elicit genu-
ine cooperation. The other part of the reason has to do with the structure
of union organization. In general, mimicking the decentralized benefit
system, unions themselves are highly decentralized. Where they have
power, then, they have incentives to free-ride on the interests of others
and to seek maximum reward for their particular labor. Decentralization
does permit wildcat cooperation. More commonly, however, it—in con-
junction with the low social wage—promotes an economistic job-control
unionism unfavorable to cooperation. Altogether, then, an environment
featuring low social wage, low union density, and highly decentralized
union organization is dense with incentives for collectively irrational
conflict.32
This diagnosis suggests four related steps of associative reform of this
system: (1) lower the barriers to unionization, (2) encourage alternative
forms of self-directed worker organization, (3) raise the social wage, and
(4) promote more centralization in wage bargaining, while permitting
high levels of decentralization in bargaining over specific work condi-
tions. We consider these in turn.
Even within the current framework of U.S. labor law—which centers
on collective bargaining between elected and exclusive worker represen-
tatives (unions) and employers—strategies for reducing barriers to worker
representation are clear enough. Elections of representatives could be
simplified and expedited, bargaining obligations could attach early and
survive the arrival of successor employers, the right to use economic
force could be enhanced, and, throughout, violations of labor regulation
could be remedied with compensatory damages rather than with tooth-
less “make whole” remedies. In a more ambitious scheme of reforms,
representation might be awarded upon a simple demonstration of sup-
port from a majority of affected workers rather than upon the elaborate
demonstration now required; individual rights of workplace members of
unions without majority status might be enhanced; restraints on the coor-
32. For a general review of the U.S. industrial relations system emphasizing these interac-
tions, see Joel Rogers, “Divide and Conquer: ‘Further Reflections on the Distinctive Character
of American Labor Law,’” Wisconsin Law Review (1990): 1–147. For a review of the state of the
American labor movement, see the contributions to George Strauss, Daniel G. Gallagher, and
Jack Fiorito, eds., The State of the Unions (Madison: Industrial Relations Research Association,
1991).
87
dination of unions in using economic force could be relaxed; greater at-

tention to the practical requirements of union “security” in maintaining
a workplace presence could be given; current restraints on the use of
member dues for organizing the unorganized, and for political action,
could be relaxed.33
Even with such reforms in place, however, most of the economy would
remain nonunion, leaving most workers without representation. We
would suggest, then, that forms of workplace representation alternative
to, though not in direct competition with, unions also be encouraged.
This could be achieved directly through a mandate of workplace com-
mittees with responsibilities in, for example, occupational health and
safety, training, or areas of concern apart from wages. Alternatively, or as
a supplement, government purchasing contracts might be used to en-
hance worker voice. Eligibility for such contracts could be conditioned
on successful employer demonstration of the existence of a works council
or some other acceptable form of autonomous employee representation
with real powers in the administration of the internal labor market.
The increased levels of worker organization that could be expected to
follow on these two changes would mitigate one of the barriers to cooper-
ation noted earlier, namely the weakness of labor organization. With la-
bor stronger, it is possible to imagine a new social contract in the internal
labor market, one that will promote cooperation. The terms of the con-
tract are simple enough: labor offers flexibility on internal labor-market
work rules and greater job commitment in exchange for management’s
commitment to consultation and heightened job security.
To ensure fairness, however, and to promote the stability of associa-
tions that contributes to their beneficial effects, a system of multiple
worker organizational forms would require an increase in the social
wage—our third initiative. For workers, an increased social wage would
provide some assurances of fair treatment and security external to the
firm. Aside from its direct distributional benefits, this increase would re-
lieve pressures for the internal rigidity and the defensiveness associated
with job-control unionism. It would make more flexible, productivity-
enhancing strategies of work organization more appealing. For employ-
33. There are many such statements of possible labor-law reform. A good guide to the issues
involved, containing both more and less ambitious recommendations for reform, is provided
by Paul Weller, Governing the Workplace: The Future of Labor and Employment Law (Cam-
bridge, MA: Harvard University Press, 1990).
88
ers, the mitigation of job-control consciousness (and the likely reduction

of labor costs) among organized workers would remove one powerful in-
centive to resist worker association in their firms.
Finally, greater coordination of wage contracts would be needed to
overcome a second barrier to cooperation and to reap the full benefits for
economic performance. As noted earlier, the American system of con-
tract negotiation is highly decentralized. It is unreasonable to expect the
United States to approximate the corporatist peak bargaining of the late
1970s (especially since corporatist systems themselves no longer approxi-
mate that). Still, some measures could be undertaken to encourage more
encompassing associations than now exist, thus generating an environ-
ment better suited to some greater centralization and coordination of
wage negotiations (at least on a regional basis).
One step would be to amend the law governing multi-employer bar-
gaining, shifting the presumption away from the voluntariness and insta-
bility of such arrangements toward their requirement. In addition, pres-
sures within the union movement for consolidation could be
strengthened by selective incentives, for example, in the form of funds
for (re)training, conditioned on inter-union cooperation. Government
support for business cooperation—for example, consortia pursuing joint
research and development strategies—could be conditioned on efforts to
consolidate wage policies. Or, following common practice in most sys-
tems, “extension laws” on bargaining contracts could be enacted, gener-
alizing their results to nonunion settings.
The effect of this combination of increasing the social wage and pro-
moting more generalization of wage patterns across firms would be to dis-
criminate more sharply between the focus of bargaining within the firm
and the focus of bargaining outside it. Within the firm, unions would
come to look more like employee-participation schemes, and employee-
participation schemes would look more like unions. Worker represen-
tation would be secured, but with a particular focus on regulating the
internal labor market and increasing productivity within it, through inno-
vation on issues of job design, work organization, access to training on
new firm technology, and the like. Outside the firm, more encompassing
organizations suited to handling matters affecting workers in general
rather than workers in a particular firm—would be more empowered to
pursue that object. They would focus more on securing generalizable
wage agreements and the content of the social wage.
89
Such a system, which relies on associative empowerment and artifice

throughout, would likely be a vast improvement on current American in-
dustrial relations. It would improve representation, increase productivity,
generalize the benefits of cooperation, and better integrate the industrial
relations system with state economic and welfare policies.
v oca t iona l t ra in i ng
Our second example of constructive group artifice comes from the area
of vocational training. In the United States, as in most other rich coun-
tries, intensified international competition and rapid technological
change have underscored the need for improvements in workforce skills.
To preserve living standards in the face of low-wage competition from
abroad, labor must be made substantially more productive, and firms
must become increasingly adept at such “nonprice” aspects of product
competition as quality, variety, customization, and service. Success here
will require, inter alia, that “frontline” production and nonsupervisory
workers be equipped with substantially higher and broader skills than
they presently possess.
The vocational training problem in the United States consists in the
fact that such skills are being provided in insufficient quality and quantity
by schools and firms, and insofar as they are provided, they are directed
to college-bound youths and managers. In the public school system, very
little occupational training is provided for the “forgotten half” of each
high school cohort that does not go on to college or the “forgotten three-
quarters” of each cohort that does not complete it. And U.S. employers
provide their frontline workforce with far less training than do leading
foreign competitors. Moreover, the training they do provide is generally
narrower than is desirable—for the economy as a whole, for innovative
firms drawing from the external labor market, and for individual workers,
who typically change employers several times in their working lifetimes.34
With skills more essential than ever to compensation, the failures of U.S.
training have powerfully contributed to the decline in production and
34. For general reviews of U.S. training problems, making all these points, see U.S. Con-
gress, Office of Technology Assessment, Worker Training: Competing in the International
Economy, OTA ITE-457 (Washington: Government Printing Office, 1990); and Commission
on the Skills of the American Workforce, America’s Choice: High Skills or Low Wages! (Roches-
ter: National Center on Education and the Economy, 1990).
90
nonsupervisory worker wages experienced over the last generation and to

rising inequality in U.S. market incomes.35
The problems in the American training system lie on both the “de-
mand” and “supply” sides. We will concentrate here on the supply side,
focusing in particular on two central issues.36
First, the quality of public school vocational training is limited by the
absence of effective linkages with the economy itself. Most public voca-
tional training in the United States is essentially “stand alone” classroom-
based instruction, and while such instruction is certainly important for
any training system, it has intrinsic limits.37 As a general matter, the sys-
tem will lag behind industry practice in its provision of skills. It will be
baffled by the need to make large expenditures on capital equipment, of
the sort needed to replicate factories inside schools. And it will have dif-
ficulty conveying to students the active knowledge they need to flourish
in, and can only acquire from, real-world production situations.
35. For a good review of wage trends in the United States and the more general decline in
living standards among nonsupervisory workers, see Lawrence Mishel and David M. Frankel,
The State of Working America, 1990–91 edition (Armonk: M. E. Sharpe, 1990).
36. A word of explanation on the focus. Demand by American employers for high and broad
frontline workforce skills is extremely weak and uneven. Unless this changes, supply-side inno-
vations geared to improving skill delivery to frontline workers will risk having all the effect of
“pushing on a string.” Moreover, competitive pressures acting alone cannot be counted on to
change the structure of employer demand in the desired way, since employers can choose to
respond to those pressures by reducing wages, increasing firm productivity through changes in
work organization that “‘dumb down” most jobs while increasing the human-capital compo-
nent of a well paid few, or simply moving away from high-end markets. Most U.S. firms, in
fact, have chosen some combination of these “low wage, low skill” competitive strategies. To
remedy the demand-side problem, it is essential to foreclose this option. The most obvious way
to do this is to build stable floors under wages and effective linkage between productivity im-
provements and wage compensation, thus forcing employers to be more attentive to strategies
for increasing the productivity of their labor (e.g., skill upgrading). Direct state action can help
here, by increasing minimum-wage floors. As regards more specifically associative reform, how-
ever—and this is why we do not linger on the demand side—we believe the most important
actions are those already outlined in the recommendations just made on improving industrial
relations. Deeper and more encompassing worker organizations, especially ones shaped by so-
cial interests in improved cooperation, would help create the needed wage floors, wage-pro-
ductivity linkages, and pressures within firms to upgrade. Moreover, they could be expected to
do so in a way that not only raises the aggregate demand for skills and their compensation but
also improves the distribution of both. The basic problem on the demand side is that the inter-
ests of the bulk of the population, workers, are simply not now centrally in the picture. They
are barely represented in the economy and only very imperfectly represented in the state. The
basic solution to under-representation is to improve the conditions of their organization in
ways consistent with other democratic norms.
37. The importance of these limits rises where, as in the United States, the public training
system lacks any effective industry-based-training complement.
91
To remedy these problems, denser linkages must be forged between

schools and students, on the one side, and between employers and their
workers, on the other. Through such linkages can flow that which the
classroom system now lacks: up-to-date knowledge on industry trends,
loans and grants of current equipment on which to train, and, all impor-
tant, access to actual workplaces, and their principals, for work-based in-
struction complementary to what goes on in the classroom.
Second, while the quantity of training supplied by government could
be expected to increase as a result of the reform of worker representation
discussed earlier, the effort by employers must also be substantially in-
creased and improved. Here, the problem is in part that employers are
uncertain about the sorts of broad-banded skills that would be appropri-
ate to provide and in part that they have no confidence that they would
capture the returns to providing training in such skills. Employer training
suffers, that is, both from a lack of agreed-upon standards for coordinated
training and from the positive externalities that accompany an open ex-
ternal labor market, in which workers are able to move freely among
firms, so that one firm’s trainee can become another firm’s asset. The ex-
ternalities problem is particularly acute for high and broad skills. Since
such skills are, by definition, of use in a wide variety of work settings,
their possession increases the potential mobility of workers, enabling one
firm to appropriate the benefits of another firm’s training efforts. This is
part of the reason that when firms do train, they train narrowly, in job-
specific or firm-specific skills.
To remedy the problem of coordination, a mechanism for setting com-
mon standards and expectations is necessary. To remedy the externality
problem, there are two basic solutions. One is to reduce worker mobility
across firms. This permits firms to train workers with the confidence that
they will recoup any investments made. In effect, this is what is done in
Japan. The other solution is to socialize the costs of private-firm training,
so that individual employers will not care about worker mobility. This
can be done with the assistance of the tax system—for example, in the
form of “train or tax” rules, requiring firms either to train or to pay into
some general fund. Or it can be done through the private collective orga-
nization of employers to a point that they can discipline free riders or, at
high levels of joint participation (where close to all relevant competitors
or poachers train), become indifferent to them. In effect, this is what is
done in successful European training systems, which, like those in the
92
United States, operate with relatively open external labor markets and
high rates of inter-firm worker mobility.
As the second European strategy makes clear, the presence of com-
petent, encompassing employer and labor associations immensely aids
both in addressing the problem of linkage between the worlds of school
and work and in increasing the level and quality of employer-sponsored
training.
Facilitating linkage, associations provide the state with timely informa-
tion on emerging industry trends and practices, new technologies, skill
needs, and access to the insides of firms. They permit industries to speak
with a unified voice to public training providers, to negotiate authorita-
tively with the state over training curricula, access to firms, require-
ments on skills certification, rules on the use of equipment, and the like.
They permit the state to get closure and enforcement on decisions once
made—“If you don’t like it, talk to your association” being a far more ef-
fective retort to second-guessing firms than, “Well, that’s just what we de-
cided to do”—while providing monitoring and enforcement capacities to
supplement any public training effort. And being broad in their represen-
tation, and accountable to members, associations are natural vehicles for
developing general standards, of wide applicability, of the sort that pro-
tect the training investment made by employees themselves.
Facilitating employer training efforts, industry associations help in part
by setting general standards on skills—something no single firm can do.
The identification of commonly desired competencies assures workers
that acquiring those competencies will improve their position on the ex-
ternal labor market. This leads to increased take-up rates on training, as-
suring employers of a large pool of workers with high and common skills.
And this assurance encourages more proactive industry strategies of up-
grading and inter-firm cooperation in implementing those strategies.
But associations also act to facilitate employer training efforts by miti-
gating the externality problem that discourages those efforts. They re-
quire training as a condition of membership, or of receipt of its benefits.
They monitor the training that goes on, relieving fears of “suckering.”
They ease the flow of information about new technology and work prac-
tices among members, providing a natural vehicle for voluntary industry
benchmarking that creates upward pressures on existing standards. They
share training facilities and curricula among themselves, reducing per
capita training costs. More elusive, but not less important, they help de-
93
fine and sustain—through means ranging from social gatherings and

award dinners to insider gossip and plum subcontracting deals—com-
mon norms of “accepted practice.” As such norms congeal into obliga-
tory industrial cultures, those who undersupply training come to be seen
less as clever businessmen than as social pariahs, to be punished with loss
of status and business. This can powerfully discourage even temptations
to defect, making the consideration of cooperation more familiar, extend-
ing and securing its reach, and lowering monitoring costs. In all these
ways, a strong employer association, especially one “kept honest” by a
strong union, can provide a powerful boost to the quality and the extent
of firm training efforts.
How might associative supports be enlisted for a revamped vocational
training system in the United States? In general terms, the problems and
the instruments at hand to solve them are clear enough. Both labor and
employer associations are relatively weak in the United States. Both need
to be strengthened, at least in their capacity to discipline their own mem-
bers and to deal with one another and with the state effectively, on train-
ing matters. Very little public money now goes directly to these purposes,
even though the lessons of comparative experience clearly indicate their
virtues. Public supports—in the form of direct cash assistance, technical
assistance, greater participation in curriculum development, increased
legal powers to enforce obligations against their own members—can be
provided in exchange for help in carrying out the important public task
of training the workforce.
For example, significant improvement in the quality of vocational
training will require some recognized occupational standards. But out-
side a few specialized trades, these do not exist. Joining with public train-
ing providers, existing unions and employer associations could be in-
vited, on an industry-by-industry basis, to develop such standards. Their
work could be facilitated by the state, in the form of modest financial
supports and technical assistance. And it should not be accepted by the
state without independent evaluation. But some product should finally
be accepted and enforced as a standard. Such enforcement will naturally
be advanced by the primary authors themselves. Employers would look
to demonstrated competence, according to these standards, when award-
ing jobs in internal labor markets. Unions would center on them in wage
negotiations or in rules governing job assignments in those markets. But
such private actions could also be supplemented through public means.
94
The standard could be made applicable to all federally funded voca-

tional training programs, for example, and adopted as a standard in arbi-
tration and judicial decisions in labor and employment law.38
The competency of labor and trade associations to provide training ser-
vices to members may be explicitly promoted by public policy as well.
Public subsidies and technical assistance to such organizations for this
purpose, utterly routine in other countries and already tried with some
success with a handful of trade and labor organizations in the United
States, would be a natural supportive policy. Antitrust law could be re-
laxed for joint training activities of member firms;39 additional amend-
ments might be needed in labor law, to permit union-management coop-
eration in training activities involving nonunion firms.40
Both of the examples just presented involve efforts to improve training
by strengthening existing associations. But the formation of new associa-
tions around training might be encouraged as well. Industry or regional
training consortia composed of firms and unions, for example, could be
encouraged through demonstration grant assistance, technical aid, and
discounts on public training services provided to their members.41 These
supports would properly be conditioned on those associations providing
training services, participating in setting standards, mounting outreach
programs to public schools, providing such schools with technical assis-
tance, expanding existing apprenticeship programs (the best, albeit much
neglected, example of vocational training in the United States), and oth-
erwise cooperating with public providers, and each other, to move to
more aggressive and inclusive training agenda. The goal again would be
to bring more order, and a critical mass, to private training efforts and to
improve effective linkages to schools.
Given the present weakness of associations in the United States, ad-
dressing the externality problem probably requires direct government ef-
38. The Department of Labor’s Office of Work-Based Learning is already making qualified
moves in this direction—“qualified” in that outside more heavily unionized industries, it re-
mains unclear what, if any, organized voice workers in the industry will have.
39. Following current practice for joint research and development activities.
40. Recommendations on how to do this are made in Margaret Hilton, “Shared Training:
Learning from Germany,” Monthly Labor Review 114, 3 (March 1991): 33–37.
41. An experiment along these lines is under way in Milwaukee, where several firms (non-
union and unionized), unions, and public training providers have come together around a
Wisconsin Manufacturing Training Consortium designed to do just these things. See Joel Rog-
ers and Wolfgang Streeck, Recommendations for Action (Madison: Center on Wisconsin Strat-
egy, 1991).
95
forts at socializing costs—through unqualified payroll levies or “play or

pay” levy structures. The revenues, however, could be used in ways that
strengthen future private capacities for self-governance. Funds might, for
example, be given to associations for redistribution. The effect would be
to create enormous temptations to associations to organize themselves to
take a more active role in training and for firms and unions to join associ-
ations—in effect, an inducement to encompassingness of the sort de-
sired. Or, in a “play or pay” scheme, tax relief could be granted to firms
that demonstrate that the training they provide conforms to the standards
set by industry associations. This would have the same effect of strength-
ening a collective associative hand in setting standards and strengthening
the associations themselves.
There are many paths to virtue, but this should be enough to make the
point. In principle, at least, the associative supports for a more success-
ful vocational training system could be achieved in the United States
with fairly standard policy instruments. Those supports would benefit
both workers and “better” firms (i.e., those firms interested in upgrad-
ing). And, far from engendering further corruption of the state, they
would strengthen public capacities to address problems of manifest pub-
lic concern.
Conclusion
The examples just given provide no more than a couple of illustrations of
the directions an associative democratic strategy might take in the United
States. But they suffice to underscore the sorts of concerns that define
that strategy and the considerations relevant to its execution. What we
have argued in this essay, and what is displayed in the examples just
given, is straightforward enough. To proceed, egalitarian politics must
once again be shown to work. To work, it requires associative supports.
Those supports can be developed. And developing them, and realizing
their contribution to democratic governance, requires neither a naive
view of associations as free from the threat of faction nor a dangerous
view on the surrender of encompassing public authority. Faction can be
mitigated through the same artifice that enlists associative contributions,
and the strength and the competence of public authorities can be en-
hanced by this enlistment.
More broadly, by assuring greater equality in organized representation
96
among private citizens, and by more effectively recruiting the energies of

their organizations into public governance, the aim of the associative
strategy is to forge an egalitarian-democratic order without an oppres-
sive state. That is nice work if you can get it—and we have suggested that
you can.
97
4
FREEDOM OF EXPRESSION
Introduction
In April 1989, students at the University of Michigan walked into a class
and were faced with a blackboard that read: “A mind is a terrible thing
to waste—especially on a nigger.” This message followed closely on the
appearance of a flier at the University declaring “open season on Blacks.”
A month later, an African student at Smith College found a message
slipped under her door, which read: “African nigger do you want some
bananas? Go back to the jungle.”1
Responding to a pattern of such incidents and the longstanding Ameri-
can traditions of racial hatred and violence reflected in them, a substan-
tial number of colleges and universities have adopted codes regulating
racist and other forms of hate speech. These regulations have been the
object of intense controversy. Denounced by some as the work of “ten-
ured radicals,”2 they have also been the target of more serious criticism.
I have presented talks based on earlier drafts of this essay at Haverford College, the Univer-
sity of California (Davis), the John F. Kennedy School of Government, Wellesley College, the
University of Illinois (Chicago), Northwestern University, the Jerusalem Philosophical En-
counter, Amherst College, New York University, the Inter-Africa Group Symposium “On the
Making of the New Ethiopian Constitution,” and the Society for Ethical and Legal Philoso-
phy. I am grateful to audiences at these talks for their criticisms and suggestions. I would also
like to thank C. Edwin Baker, Randall Forsberg, John Rawls, John Simmons, and Cass
Sunstein for comments on previous versions, Sunstein for making available successive drafts of
his Democracy and the Problem of Free Speech (New York: Free Press, 1994), and Archon Fung
for his research assistance. More generally, I am very much indebted to Tim Scanlon’s papers
on freedom of expression, in particular his “Freedom of Expression and Categories of Expres-
sion,” University of Pittsburgh Law Review 40 (1979): 519–550.
1. See Charles Lawrence, “If He Hollers Let Him Go: Regulating Racist Speech on Cam-
pus,” Duke Law Journal (1990): 431–483; Mari Matsuda, “Public Response to Racist Speech:
Considering the Victim’s Story,” Michigan Law Review 87 (1989): 2320–2381; Richard Delgado,
“Campus Antiracism Rules: Constitutional Narratives in Collision,” Northwestern University
Law Review 85 (1991): 343–387.
2. See for example George Will, “Curdled Politics on Campus,” Newsweek, 6 May 1991, 72;
Chester E. Finn, “The Campus: ‘An Island of Repression in a Sea of Freedom,’” Commentary,
September 1989, 17–23.
98
F R E E D O M O F E X P R E S S I O N
The University of Michigan’s own speech code was found constitution-

ally infirm by Judge Avern Cohn.3 Considering the University’s record in
implementing that code, Cohn’s objections were well-taken.4
Still, critics commonly sweep too widely. The United States is, after
all, unique internationally in its legal toleration of hate speech.5 And the
Michigan rule is not the only model. Consider, for example, Stanford’s
regulation on discriminatory harassment. The Stanford behavior code
regulates “speech or other expression” that:
(1) is intended to insult or stigmatize individuals on the basis of their

sex, race, color, handicap, religion, sexual orientation, or national
and ethnic origin.
(2) is addressed directly to the individual or individuals whom it in-
sults or stigmatizes.
(3) makes use of insulting or “fighting words” or non-verbal symbols
that are “commonly understood to convey direct and visceral
hatred or contempt for human beings on the basis of their sex,
race, etc.”
Expression is only regulable if it meets all three conditions. So here we

have a not very restrictive regulation that can be endorsed consistent with
a strong commitment to freedom of expression and to the toleration asso-
ciated with that commitment.6 It does restrict some expression, but it is
not very restrictive.7 There is no violation if a student in a course or at a
political rally says, “The Holocaust is a Zionist fraud,” or “Slavery was a
great civilizing influence.” Indeed, the regulation does not prohibit very
3. Doe v. University of Michigan, 721 F. Supp., 852 (E.D. Michigan, 1989).
4. See ibid. for discussion of cases of enforcement against comments made in the course of
classroom discussion.
5. See David Kretzmer, “Freedom of Speech and Racism,” Cardozo Law Review 8 (1987):
445–513; Eric Stein, “History Against Free Speech: The New German Law Against the
‘Auschwitz’—and Other ‘Lies,’” Michigan Law Review 85 (1986): 275–324; Kenneth Lasson,
“Racism in Great Britain: Drawing the Line of Free Speech,” Boston College Third World Law
Journal 6 (1987): 161–181; Robert Sedler, “The Constitutional Protection of Religion, Expres-
sion, and Association in Canada and the United States: A Comparative Analysis,” Case Western
Reserve Journal of International Law 20 (1988): 577–621; Matsuda, “Public Response to Racist
Speech”; Delgado, “Campus Antiracism Rules.”
6. But it does appear to be inconsistent with the current view of the Supreme Court on per-
missible forms of state regulation of speech. See R. A. V. v. St. Paul, 122 S. Ct. 2538 (1992). I will
discuss this view in Section 5.
7. In saying that it is not very restrictive, I do not mean to say that it is, therefore, an accept-
able restriction. “No one whose first name includes the letters ‘z’ and ‘y’ may criticize Impres-
sionist painting” is not very restrictive, but it is also not acceptable.
99
much at all—for example, probably not the Michigan or Smith cases I

mentioned at the outset.
Putting the extent of prohibition to the side, what is the rationale for it
and similar regulations? The aim is not to encourage civility, to shelter
people from offensive comments, or to punish malign ignorance. The
regulations are motivated instead by various costs associated with discrim-
inatory harassment: direct psychological injury, indirect injury from en-
couragement to assaults on targeted groups, and, in particular, damage to
prospects of equality that comes from undermining equality of educa-
tional opportunity within the university and from contributing to an envi-
ronment in which unacceptable forms of discrimination seem reason-
able. Judge Cohn’s opinion in Doe v. Michigan gives special notice to
concerns about equality. He begins by noting that it is an “unfortunate
fact of our constitutional system that the ideals of freedom and equality
are often in conflict.” Responding to this unfortunate fact, he indicates in
the concluding section of his opinion that the Court is “sympathetic to
the University’s obligation to ensure equal educational opportunities for
all of its students” but emphasizes that “such efforts must not be at the ex-
pense of free speech.”8
Why not? What is this “expense” of regulating free speech? Why is
the expense of such magnitude that, in the face of it, concerns about
such substantial values as equality rise only to the level of “sympathetic”
concern?
My aim here is to address these and related questions. To that end, I
leave aside for now the immediate controversies about speech codes,
though I will return to the Stanford code at the end, indicating why a
pallid endorsement of it is consistent with an affirmation of stringent
protections of expressive liberties. Principally, however, I will be arguing
for the pallor of the endorsement by discussing some reasons for the
protections. The discussion will show that support for such regulations
need not reveal a disdain for the values of freedom of expression and that
a lack of enthusiasm for them need not reveal indifference to the destruc-
tive potential of hate speech. To claim otherwise—to draw a line of prin-
ciple around regulations of this kind—is to provoke a divisive and unnec-
essary division between liberal and egalitarian commitment.
I start (Section 1) by describing what I mean by “stringent protections
of expressive liberties.” Then (in Section 2) I sketch and criticize two
8. Doe v. Michigan, 721 F. Supp., 853, 868, emphases added. Cohn did not indicate the sorts
of measures that might be consistent with the First Amendment.
100
strategies for defending such stringency. The first, which I will call “mini-
malist,” holds that expression deserves stringent protection, not because it
is so valuable but because it is costless (“just speech”), or because the
costs it imposes cannot permissibly be taken into consideration by the
state, or because government is especially untrustworthy when it comes
to regulating expression: the common thread running through the sev-
eral variants of minimalism is that the defense is to proceed without re-
course to the thesis that expression has substantial value. “Maximalist”
views, by contrast, concede the costs of stringent protections but argue
that the transcendent value of expression guarantees that it trumps the
costs (except when they are of equally transcendent value).
Minimalism and maximalism are not formal theories about freedom of
expression. Still, each represents an important tendency of thought in
this area.9 Moreover, their attractive simplicity encourages the assump-
tion that they exhaust the field of justifications. Because neither is com-
pelling, nihilism about freedom of expression lives parasitically off their
defects—the nihilism urged, for example, in Stanley Fish’s claim that
“there’s no such thing as free speech and it’s a good thing, too.”10 Put less
colorfully, the nihilist claims that all there really is—all there could be—
when it comes to decisions about restricting or permitting speech is an ad
hoc weighing of costs and benefits in particular cases using the scales pro-
vided by “some particular partisan vision.”11 No general presumption in
favor of protection can withstand inspection.
But minimalism and maximalism do not exhaust the strategies of ar-
gument for stringent protections.12 The central burden of my argument
(in Sections 3 and 4) is to present an alternative to maximalist and mini-
malist outlooks and, thereby, to defuse some of the temptations to nihil-
ism. Less simple than the alternatives, this view proposes that stringent
protections emerge as the product of three distinct considerations:
(1) That certain fundamental interests—expressive, deliberative, and

informational—are secured by stringent protections of expressive
liberty.
9. As much in informal conversations about these issues as in the legal and philosophical lit-
erature.
10. See Boston Review 17 (January-February 1992): 3–4, 23–26. I mean “nihilistic” not as a
tendentious label or a term of criticism, but rather as a term that captures the “there is no such
thing” point in the quotation in the text. The view could be called “pragmatism” about expres-
sion, but this misses its critical edge.
11. Ibid., 26.
12. For a response to nihilism, see below, 128–131.
101
(2) That the costs of expression can, in an important range of cases,

be addressed through—as Justice Brandeis put it—“more speech.”
(3) That certain features of human motivation render expression vul-
nerable to underprotection, and so rigid protections are recom-
mended for it.13
Stringent protections, then, help to advance a set of fundamental inter-
ests and are recommended principally by the importance of those inter-
ests, by the prospects of using expression as a preferred strategy for com-
bating the costs of expression, and secondarily—but only secondarily—
by concerns about our tendency to underprotect expression or mistrust
government regulations of it.14 I see no rationale that is at once simpler
and as compelling. To be sure, the complexity opens the view to the
charge of manipulability—of providing a set of relatively unstructured
elements that, with suitable adjustments, can be made to deliver any re-
sult. I do not think the view has quite that defect. In any case, I think
things are just this complex and see no gain in substituting an arbitrary
truncation of relevant considerations for a complex but hard-to-manage
structure.15
One feature of the account that I want especially to emphasize is that
the defense of stringent protections of expression does not depend on a
free-standing preference for liberty over all competing values—not, in
particular, on a free-standing preference for liberty over equality and an
associated condemnation of any restrictions of expression that (like hate-
speech regulations) are undertaken in the name of the value of equality.16
The idea that a commitment to freedom of expression depends on a free-
standing preference for liberty over equality is, I believe, a serious mis-
take, because it fosters an unnecessary and destructive hostility to free-
dom of expression among friends of equality and an unnecessary and
13. See, for example, Vincent Blasi, “The Pathological Perspective and the First Amend-
ment,” Columbia Law Review 85 (1985): 449–514.
14. For an argument based more fundamentally on mistrust, see Richard Epstein, “Property,
Speech, and the Politics of Distrust,” University of Chicago Law Review 59 (1992): 41–90. For
criticisms, see Frank Michelman, “Liberties, Fair Values, and Constitutional Method,” Univer-
sity of Chicago Law Review 59 (1992): 91–114.
15. Cass Sunstein criticizes my view along these lines. He suggests that the legal principles
that emerge from it might be “simply too complex, ad hoc, and unruly.” See Democracy and
the Problem of Free Speech, 146. I try to address this concern in my discussion of specific issues
in Sections 4 and 5.
16. For the suggestion that it does so depend, see Ronald Dworkin, “Two Concepts of Lib-
erty,” in Edna Ullman-Margalit and Avishai Margalit, eds., Isaiah Berlin: A Celebration (Lon-
don: Hogarth Press, 1991), 100–109.
102
destructive hostility to equality among friends of expressive liberty.

Where reconciliation is possible, it promotes division; where disagreement
is possible on common ground, it insists on drawing false lines of principle.
This point bears special notice because of the state of debate about
freedom of expression. For much of the past century, egalitarians of the
political left have been among the most insistent defenders of stringent
protections of expressive liberty, arguing that freedom of expression is
both an intrinsic aspect of human liberation and a precondition of popu-
lar democratic politics. Over the past three decades, this conjunction of
egalitarian and libertarian commitment has been subjected to increas-
ingly severe strain. Regulations of political spending aimed at enhancing
the voice of less wealthy citizens have been condemned as unacceptable
abridgements of expressive liberty. And free-speech values have been ad-
vanced as obstacles to regulating pornography and hate speech. Because
these regulations, too, are in part about promoting equality, the sug-
gestion has emerged that egalitarian and libertarian commitment have
come to a parting of the ways. I disagree, and aim to state a case for strin-
gent protections of expressive liberty in the tradition of free-speech egali-
tarianism.
Finally, I explore some of the implications of the view. In particular,
the basic framework of argument for stringent protections suggests a dif-
ferent treatment of hate-speech regulations than that advanced in Justice
Scalia’s opinion in the 1992 case of R. A. V. v. St. Paul. So in Section 5, I
discuss some reasons for rejecting Scalia’s reasoning and explore as well
the consistency of a certain style of pornography regulation with the view
advanced here. Finally, I return to the Stanford regulation (Section 6),
indicating how an endorsement of it is consistent with the bases for strin-
gent protections of freedom of expression advanced here.
One last introductory point: throughout, I help myself freely to exam-
ples, terms, and ideas drawn from First Amendment law.17 My aim, how-
ever, is not to present an interpretation of the Constitution but to present
a rationale for stringent protections.
1. Stringent Protections
I will begin by explaining what I mean by “stringent protections of ex-
pressive liberties.” My explanation proceeds by setting out four familiar
themes suggested by the free-speech tradition. Nothing I say about these
17. In this connection I wish to thank C. Edwin Baker for pointing out a number of blun-
ders that marred the penultimate draft of the paper.
103
themes will be original or even unfamiliar; each could be expressed in

different ways; and not much will turn on the particular formulations.
But I do need some statement of themes at hand to fix the idea of strin-
gent protections sufficiently to be able to consider the bases for it.
presump t ion a ga inst cont e nt re gul a t i on

It is common to distinguish regulations of expression that focus on con-
tent—including viewpoint and subject matter—from those that are con-
tent-neutral. A prohibition on advocating adultery restricts viewpoint; a
prohibition on discussing adultery restricts subject matter; a prohibition
on debating the merits of adultery (or anything else) on my street at
3:00 a.m. is content-neutral. The first theme, then, is that there is an es-
pecially strong (if rebuttable) presumption against regulating expression
in virtue of subject matter and, still more particularly, viewpoint: a pre-
sumption against regulations animated by a concern for what a person
says or otherwise communicates, or consequences that flow from what he
or she says.18
ca t egoriz a t ion
Despite this general presumption, some kinds of content regulation seem
intuitively less troubling—as with regulations of express, direct incite-
ment, truth in advertising, private libel, fighting words, bribery, espio-
nage, and nonobscene child pornography.19 Because content regulation
is in general objectionable, it is important to confine these exceptions.
So a second main theme recommends a special approach to handling
18. The classic statement of the general concern is Justice Marshall’s in Police Department
of Chicago v. Mosley, 408 U.S. 92, 95–96 (1972): “Above all else, the First Amendment means
that government has no power to restrict expression because of its message, its ideas, its subject
matter, or its content.” On viewpoint discrimination, see Texas v. Johnson, 491 U.S. 397, 414
(1989). For discussion, see John Hart Ely, “Flag Desecration: A Case Study in the Roles of Cat-
egorization and Balancing in First Amendment Decisions,” Harvard Law Review (1975): 1482–
1508; Geoffrey Stone, “Restrictions of Speech Because of its Content: The Peculiar Case of
Subject-Matter Restrictions,” University of Chicago Law Review 46 (1978): 81–115, and “Con-
tent-Neutral Restrictions,” University of Chicago Law Review 54 (1987): 46–120; and T. M.
Scanlon, Jr., “Content-Regulation Reconsidered,” in Judith Lichtenberg, ed., Democracy and
the Mass Media (Cambridge: Cambridge University Press, 1990), 331–354.
19. For example, on incitement, see Brandenburg v. Ohio, 395 U.S. 444 (1968); on commer-
cial speech, Virginia State Board of Pharmacy v. Virginia Citizens Consumer Council, 425
U.S. 748 (1976); Central Hudson Gas and Electric v. Public Service Commission of New York,
447 U.S. 557 (1980); Posadas de Puerto Rico Associates v. Tourism Company of Puerto Rico,
106 S. Ct. 2968 (1986); on fighting words, Chaplinsky v. New Hampshire, 315 U.S. 568 (1942);
on child pornography, Ferber v. New York, 458 U.S. 747 (1982).
104
content regulations. Sometimes called “categorization,”20 the approach

singles out a small set of categories of expression—in First Amendment
law, for example, child pornography, commercial speech, obscenity,
fighting words, and express incitement—for lesser protection, specifying
conditions for permissible regulation of expression in each category.21 For
content-neutral regulations, by contrast, the second theme asserts that
there ought to be a more or less explicit balancing, with a thumb in
the scale for speech, and an especially heavy thumb when the burden
of a content-neutral regulation is especially great for groups with re-
stricted means to convey their views.22 (I will revisit this last point about
“weighted balancing” below, in discussing fair access.)
cost l y p rot ect io ns

Expression sometimes has unambiguous costs.23 It is sometimes offen-
sive, disgusting, or outrageous; it produces reputational injury and emo-
tional distress; it requires protection from hecklers; when it is delivered
through leaflets, someone has to clean up the mess; and, concentrated in
sufficient numbers on billboards, telephone poles, and buses, it can add
to the general ugliness of an urban environment. But—here is the third
theme—the presence of such costs does not as a general matter suffice to
remove protection from expression. Neither offense, nor cleanup costs
for taxpayers, nor reputational injury, nor emotional distress, for exam-
ple, suffice by themselves to deprive expression of protection.24
I am not suggesting that all libel law is inconsistent with stringent
protections of expression, or that the intentional infliction of emotional
distress always deserves protection, or that fines for littering always offend
the ideal of freedom of expression.25 I mean only that even uncontested
20. I emphasize that I am using this protean term exclusively as a label for the approach to
content regulation described here.
21. For doubts about the virtues of categorization, and corresponding skepticism about cate-
gorically formulated prohibitions on content regulation, see John Paul Stevens, “The Freedom
of Speech,” Yale Law Journal 102 (1993): 1293–1313.
22. For a subtle discussion of the structure of argument about content-neutral regulations,
and of the extent to which the thumb gets put in the balance for different sorts of regulations,
see Stone, “Content-Neutral Restrictions.”
23. By “costs,” I mean, quite generically, conditions that it is reasonable to want to avoid.
24. See Schneider v. State, 308 U.S. 147 (1939) (on cleanup costs); New York Times v.
Sullivan, 376 U.S. 254 (1964) (on reputational injury); Cohen v. California, 403 U.S. 15 (1971)
(on offense); Hustler v. Falwell, 485 U.S. 46 (1988) (on intentional infliction of emotional dis-
tress).
25. Say, a fine for leaving a pile of leaflets sitting on a bench for people to pick up.
105
facts of reputational injury or emotional distress or mess are not always

sufficient to deprive expression of protection—as when the target of ex-
pression is a public figure, or when the expression focuses on a subject of
general interest. When, for example, New York Times v. Sullivan required
a showing of “actual malice” in order for a public figure to win a libel
judgment,26 or when Hustler v. Falwell required actual malice in cases of
the intentional infliction of emotional distress,27 there was no suggestion
that actual malice is necessary for reputational injury or emotional dis-
tress. Instead it was held, in effect, that the values associated with a sys-
tem of free expression outweighed those injuries.
fa ir a ccess
A system of stringent protections of expressive liberties must assure fair
opportunities for expression: that is, that the value of expressive liberties
must not be determined by a citizen’s economic or social position.28
Taking the unequal command of resources as a fact, a system of stringent
protections must include measures aimed expressly at ensuring fair ac-
cess to expressive opportunities. Such measures might include keeping
traditional public forums (parks and streets) open and easily accessible;
expanding the conception of a public forum to include airports, train
stations, privately owned shopping centers, and other places of dense
public interaction; affirming the importance of diverse broadcast mes-
sages and the role of fair access in contributing to such diversity; financ-
ing political campaigns through public resources; and regulating private
political contributions and expenditures. The requirement of fair access
supports a strong, general presumption against content-neutral regula-
tions that have substantially disparate distributive implications—when, as
with regulations on the distribution of handbills, or on the use of parks
and sidewalks, they work to disproportionately burden those who other-
wise lack the resources to get their messages out.
Several preliminary comments on this inclusion of fair access in the
account of stringent protections are in order.
26. 376 U.S. 254, 279–280 (1964).

27. 485 U.S. 46, 50 (1988).
28. On the idea of the value of liberty, see John Rawls, A Theory of Justice (Cambridge, MA:
Harvard University Press, 1971), 204–205. On the rationale for requiring a fair value for political
liberty in particular, and the permissibility of (content-neutral) regulations of political speech
in order to ensure that fair value, see John Rawls, “The Basic Liberties and Their Priority,” Po-
litical Liberalism (New York: Columbia University Press, 1993), lecture 8, secs. 7, 12.
106
First, the measures I listed for ensuring fair access are all content-
neutral, and all are addressed to remedying problems of unfair access
that reflect inequalities of material resources. But it is a serious and open
question whether, and to what extent, fair access can be assured through
content-neutral remedies. For a lack of fair access—social and political
exclusion—is sometimes said to result precisely from what others say
and not from the distribution of resources.29 This tension between de-
mands of content-neutrality and fair access lies at the heart of Catharine
MacKinnon’s argument for the regulation of pornography on grounds
that its content silences women and so prevents fair access.30 Here I want
simply to call attention to this concern. Later, I will suggest some ways to
address it and so broaden the range of cases in which values of fair access
and content-neutrality can be reconciled (135–140).
Second, it might be objected that the inclusion of requirements of fair
access abuses the phrase “stringent protection”—that ensuring fair access
is really a matter of “positively” expanding expressive opportunities rather
than “negatively” protecting expressive liberties. I have a more formal
and a more substantive response to this objection.
The formal response is that my four points define “stringent protec-
tion” for the purposes of this essay. So the terminological issue does not
interest me very much. More substantively, I disagree that this represents
an abuse or stretching of the term “protection.” When owners of shop-
ping malls wish to prevent people from leafleting on the premises, and
the state bars them from doing so, the state is protecting at least some
expression from efforts (by the owners) to silence it. It is tendentious to
describe this as an effort by the state to expand opportunities for the
leafleteers (rather than as an effort to protect their liberty from intrusion),
because that description imports a presumptive right of owners to ex-
clude into the distinction between protection and expansion.31 The real
29. See Frank Michelman, “Universities, Racist Speech, and Democracy in America,” Har-
vard Civil Rights-Civil Liberties Law Review 27 (1992): 352.
30. See, for example, Catharine MacKinnon, “Francis Biddle’s Sister,” in her Feminism
Unmodified (Cambridge, MA: Harvard University Press, 1987), 163–197; for discussion of the si-
lencing argument, see Frank Michelman, “Conceptions of Democracy in American Constitu-
tional Argument: The Case of Pornography,” Tennessee Law Review 56 (1989): 291–319; and the
critical appraisal in Ronald Dworkin, “Liberty and Pornography,” New York Review of Books 38,
14 (1991): 12–15.
31. For an extended elaboration of the importance of this point for free-speech doctrine, see
Cass Sunstein’s discussion of a New Deal for speech in “Free Speech Now,” The University of
Chicago Law Review 59 (1992): 255–316, esp. 263–277, 316.
107
issue—about which I do not claim to have said anything thus far—is

whether fair access to expressive arenas ought to be ensured as a matter of
right to citizens, including those who otherwise lack the resources for
participating in such arenas. In grouping these four themes together as
“protections,” I do not mean to have answered or even to have addressed
that question.
So here we have four components of a system of stringent protections
of expressive liberty: strong presumption against content regulation; cate-
gorization as a method for handling such regulation; rejection of the
view that a showing of costs suffices to remove protection from expres-
sion; and assurances of a fair distribution of expressive opportunities.
Let’s now consider some reasons for endorsing a scheme of stringent
protections, thus understood.
2. Two False Starts

Earlier I briefly sketched minimalist and maximalist styles of argument
for stringent protections. Taking “stringent protections” now to be de-
fined by the four features I presented in the last section, I want to discuss
these strategies in more detail.
minima l ism
Generically described, minimalism aims to defend stringent protections
without attaching any elevated importance to expression, instead con-
centrating, so to speak, on the magnitude of the evil those protections
prevent rather than on the magnitude of the good they protect. One fa-
miliar minimalist strategy—I will call it “no-cost minimalism”—rests on
a distinction between expression and action. Relying on that distinction,
the minimalist argues that expression—as distinct from action—is not in
itself costly or harmful and that the harms that may flow from it in con-
junction with its surrounding conditions can always be addressed without
abridging expression. The no-cost case does not rest on attaching an es-
pecially significant value to expression itself: the harm principle suffices
to generate the protections.
Other minimalists emphasize as well the remedial side of stringent
protections, arguing that they are required by the pervasive tendency of
people generally (or, in some versions, of political officials) to silence ex-
pression for insubstantial or impermissible reasons: for example, to pro-
tect officials from criticism or for reasons that are moralistic or in other
108
ways demeaning to the targets of regulation. Ronald Dworkin, for exam-

ple, has argued that a right to consume pornography is one implication
of a general ban on the enforcement of preferences about the proper way
for other people to conduct their lives; the right serves as a protective de-
vice against the legal imposition of moralistic preferences, such imposi-
tion constituting a demeaning denial of the abstract right of citizens to be
treated as equals.32 This defense of the right to consume pornography is
minimalist because it turns not on any special value of expression gener-
ally, or of sexual expression in particular, nor on the claim that restric-
tions of expression are exceptionally burdensome, but on only the ab-
stract right to be treated as an equal, the claim that that right is violated
by the legislative imposition of external preferences, and the factual as-
sumption that regulations of expression that emerge from the democratic
process commonly are rooted in such preferences.
Minimalism makes two important points: it registers a concern about
tendencies to excessive abridgement; and it emphasizes the importance
of avoiding the injuries of expression by means other than the restriction
of expression, where that is possible. Both points will figure in my own
account. But minimalism generally, and “no-cost” minimalism in partic-
ular, is pretty much hopeless as a foundation for stringent protections.
Consider for example the third element in the scheme of stringent pro-
tection: using “expression” in its ordinary English sense, expression is
sometimes harmful, and so protecting it has a price.33 Denying the cost is
simply insulting to those who pay it. Moreover, protecting people with
unpopular messages and assuring outlets for expression is costly: some-
times you have to pay for police protection or to sweep the streets to
clean up leaflets. It is not clear how no-cost minimalism proposes to cap-
ture these components of the idea of a scheme of stringent protections.
The minimalist might, of course, be understood as introducing a new
32. “Do We Have A Right to Pornography?” in A Matter of Principle (Cambridge, MA: Har-
vard University Press, 1985): 335–372. For an excellent discussion of the limits of this argument,
see Rae Langton, “Whose Right? Ronald Dworkin, Women, and Pornographers,” Philosophy
and Public Affairs 19 (1990): 311–359. Dworkin’s recent discussions of expressive liberty seem
less minimalist. In “What is Equality? Part 3: The Place of Liberty,” he ties the value of expres-
sive liberties to the formation of “authentic preferences.” See Iowa Law Review 72 (1987): 34–
36. And in “The Coming Battles over Free Speech,” he notes the importance of an active side
to personal responsibility. See New York Review of Books, 11 June 1992, 57.
33. For criticism of the project of founding an account of freedom of expression on a prior
expression-action distinction, see Scanlon, “A Theory of Freedom of Expression,” Philosophy
and Public Affairs 1 (1972): 205–208.
109
technical sense of expression—call something “expression” only if it car-

ries no costs. But then minimalism will offer no help in understanding
the rationale for stringent protections of expression as characterized here,
since they protect expressive liberty in a much wider sense than the tech-
nical one just noted—that is, even when expression has costs.34
Or consider the style of minimalism that supplements the case for
stringent protections by emphasizing a concern for tendencies to restrict
expression for demeaning reasons. This still seems insufficient as a ratio-
nale for a system of free expression. It is difficult, for example, to see the
rationale for a “thumb-in-the-scale” for expression in the case of content-
neutral regulations or in the face of a wide range of costs of expression,
unless we premise an affirmative value for expression and not simply a re-
quirement to abjure demeaning justifications for restrictions of liberty.
Consider some content-neutral reasons for restricting expression: to keep
the streets clean, the clutter under control, the noise level down, and
traffic flowing smoothly. Nothing here seems to involve a troubling, de-
meaning failure to treat people as moral equals. The problem with, for
example, sharp restrictions on political demonstrations enacted for these
reasons is that they give insufficient weight to the value of expressive lib-
erty.35 It is true, as I indicated earlier, that content-neutral regulations are
sometimes troubling because of their disparate impact, as they are espe-
cially burdensome on citizens who lack the means to get their message
out. And it might be thought such unequal burdensomeness signals the
presence of a demeaning rationale for the regulations. But in the absence
of an antecedent reason for treating expressive liberties as fundamental, I
doubt that the conclusion can be supported, unless all forms of disparate
impact are demeaning.36
Finally, none of the forms of minimalism seems to provide a good ra-
tionale for the fourth feature of stringent protections: assurances of fair
access to expressive opportunity.
ma x ima l ism
Maximalism inverts the minimalist strategy. Generically described, the
maximalist proposes that expression merits stringent protection because
34. Indeed, as Scanlon emphasizes, the central task for a theory of freedom of expression is
to explain why this should be so. See ibid., 204.
35. See Schneider v. State, 308 U.S. 147, 161 (1939).
36. To put the point in a constitutional slogan: you cannot derive the First Amendment
from the equal protection clause of the Fourteenth Amendment.
110
its great value guarantees that the benefits of protection trump the costs.37
The maximalist might, for example, argue that the dignity of human be-
ings as autonomous and responsible agents is so immediately at stake in
any act of expression and/or so immediately threatened by any regulation
of expression—or at least any regulation of expression on grounds of its
communicative impact—that abridgements of it represent intolerable vi-
olations of human dignity.38
The maximalist view has something right, and I will say what it is when
(in Section 3) I discuss fundamental expressive and deliberative interests.
Still, maximalism is too simple to capture the contours of freedom of ex-
pression. In its simplicity, it either exaggerates the stakes in particular
cases of the regulation of expression or else manipulates the notion of au-
tonomy to make it fit the complexity of the terrain.39
For example, maximalism does not help us to understand why there
are cases in which considerations of costs do seem relevant to the jus-
tification of regulations: why regulations of group libel might be more
problematic than restrictions on individual libel, why it might make
sense to distinguish the treatment of reputational injury to public and
nonpublic figures, or why autonomy does not simply trump reputational
injury altogether. Similarly, maximalism does not seem to be a promising
route to understanding why false or misleading advertising seems less
worthy of protection than false or misleading claims offered in the course
of political or religious argument.40 In each of these cases, maximalism
37. The maximalist need not hold that the value is intrinsic, nor that there is just a single
value associated with expression. I am indebted to Connie Rosati for urging this clarification.
38. See, for example, Scanlon’s listener-autonomy theory in “A Theory of Freedom of Ex-
pression,” 204–226; and his criticisms of that theory in “Freedom of Expression and Categories
of Expression,” 534–35. In “Persuasion, Autonomy, and Expression,” Columbia Law Review 91
(1991): 334–371, David Strauss aims to rescue a version of Scanlon’s theory from these criti-
cisms. Strauss condemns restrictions of speech justified by reference to the harmful results of
the speech’s persuasive power as inconsistent with listener autonomy.
39. This complaint is registered in Scanlon, “Freedom of Expression and Categories of Ex-
pression”; Steven H. Shiffrin, The First Amendment, Democracy, and Romance (Cambridge,
MA: Harvard University Press, 1990), chap. 4. On the contrast between the complexity of the
terrain and the simplicity of familiar theories, see Harry Kalven, A Worthy Tradition: Freedom
of Speech in America, ed. Jamie Kalven (New York: Harper and Row, 1988), 23.
40. In “Persuasion, Autonomy, and Expression,” Strauss argues that persuasion is “a process
of appealing, in some sense, to reason” (335), and that we ought not to regulate expression
when its harmful effects come from its power to persuade. Thus false advertising gets reduced
protection because it does not work by persuasion (as though we were in no danger of being
persuaded by liars); nonobscene pornography is protected because it does work by persuasion
(as though the scenes of the Washington Monument and the American flag featured at the
start of some triple-X movies were representative). Here I lose hold of his conception of persua-
sion and so of the argument about commercial speech.
111
has troubles with an intuitive idea, or distinction. Perhaps there is, in the
end, nothing more to these “intuitions” than second nature masquerad-
ing as first. But they do have some presumptive weight and so raise trou-
bles for maximalism.
Furthermore, if considerations about the transcendent value of expres-
sion are understood to provide only reasons for rejecting regulations on
grounds of communicative impact, then they will provide no limit at all
on content-neutral regulations—no weighted balancing—and no basis
for a concern with fair access. On the other hand, if considerations of au-
tonomy are understood to ground a uniform presumption against all reg-
ulation of expression because of the uniform connection of expression
with the value of autonomy, then either the uniform presumption will be
very low and the protections will be weak, or the uniform presumption
will be very high, and we will all have lots of listening to do.
More fundamentally, the main idea behind the variant of maximalism
I have focused on here is that expression always trumps other values be-
cause of its connection with autonomy. And this suggests that a commit-
ment to freedom of expression turns on embracing the supreme value of
autonomy. But this threatens to turn freedom of expression into a sectar-
ian political position. Is a strong commitment to expressive liberties re-
ally available only to those who endorse the idea that autonomy is the
fundamental human good—an idea about which there is much reason-
able controversy? I am not doubting that such a strong commitment is
available to those whose ethical views are of this kind, but I reject the
claim that such views are really necessary. The force of this concern
about sectarianism will become clearer as I describe an alternative to
minimalism and maximalism. Suffice it to say for now that it would be
desirable to frame an account of the values at stake that is capable of re-
ceiving wider support, an account that would free the doctrine both from
the insulting idea that expression is costless and from the sectarian idea
that it is priceless.
3. An Alternative Strategy: Foundations

The difficulties with maximalist and minimalist strategies recommend
a different angle of approach: one which gives stringent protections as
a conclusion but does not assume that expression is costless or price-
less. More precisely, I will present a view that not only gives more weight
to the value of expression than minimalism, while retaining its emphasis
on the desirability of nonrestrictive remedies for harms and its concern
112
with tendencies to overregulate, but also discriminates more finely than

maximalism, while preserving its emphasis on the importance of expres-
sion. I propose that three kinds of consideration work together to gener-
ate upward pressures for protection and so to provide the basis for the
scheme of stringent protections.
(1) An idea of the fundamental interests that are protected by a system
of freedom of expression.
(2) An account of the structure of the costs of these protections.
(3) A set of more or less commonsense factual claims that I will refer
to as fundamental background facts.
I will consider each of these in turn, beginning with the fundamental
interests, and then in the next section discuss the case for protection pro-
duced by their joint operation.
int erest s
Freedom of expression is commonly associated with such values as
the discovery of the truth, the self-expression of individuals, a well-
functioning democracy, and a balance of social stability and social
change.41 I do not wish to dispute these associations but rather to connect
more transparently the importance of expression with certain fundamen-
tal interests.
In particular, I distinguish three interests that are protected by stringent
assurances of expressive liberty and whose importance makes the de-
mand for substantial protection reasonable. I will call them the expres-
sive, the deliberative, and the informational interests. Before describing
those interests, however, I want to highlight the background of my ac-
count of them.
Earlier I accused maximalism of sectarianism. Because I want to steer
clear of that sectarianism, my presentation of these interests, and of their
importance, is framed to accommodate what I will call “the idea of rea-
sonable pluralism.”42 In brief, the idea of reasonable pluralism is that
41. See, in general, Thomas Emerson, The System of Freedom of Expression (New York: Vin-
tage, 1971). Lee C. Bollinger emphasizes as well the importance of encouraging tolerance in
The Tolerant Society (Oxford: Oxford University Press, 1986); and Vincent Blasi examines the
role of freedom of expression as a check on official misconduct in “The Checking Value in
First Amendment Theory,” American Bar Foundation Research Journal 3 (1977): 521–649.
42. For fuller discussion, see my “Moral Pluralism and Political Consensus,” in The Idea of
Democracy, ed. David Copp, Jean Hampton, and John Roemer (Cambridge: Cambridge Uni-
versity Press, 1993), 281–287 [essay 2 in this volume]; also see John Rawls, Political Liberalism,
35–38.
113
there is a plurality of distinct, conflicting, fully reasonable understand-

ings of value. An understanding of value is fully reasonable—which is
not the same as true43—just in case its adherents are stably disposed to
affirm it as they acquire new information and test it through critical rea-
soning and reflection.44 I emphasize that “test through critical reasoning
and reflection” is itself a normative notion: so a view is not reasonable
simply because of the dogged persistence of its adherents, who preserve
their disposition to affirm it after hearing (though not listening to) all the
arguments. The contention that there is a plurality of such understand-
ings is suggested by the absence of convergence in reflection on issues of
value—the persistence of disagreements, for example, about the values
of autonomy, welfare, and self-actualization; about the value of devotions
to friends and lovers, as distinct from more diffuse concerns about ab-
stract others; and about the values of poetic expression and political en-
gagement.
Acknowledging the pluralism of reasonable evaluative conceptions has
important implications for political justification. It suggests in particular
that we ought to conduct such justification in terms of considerations
that provide compelling reasons within other views as well. For when we
restrict ourselves in political argument to the subset of moral consider-
ations that others who have reasonable views also accept, we are acknowl-
edging that their views are not unreasonable, even if they do believe what
we take to be false.
Premising reasonable pluralism, then, I look to characterize inter-
ests whose importance provides a basis for stringent protections and that
are located on common ground shared by different reasonable con-
ceptions. Because different views disagree in their substantive character-
ization of what is valuable, the basic interests will inevitably be pre-
sented in abstract terms. But this abstractness is no metaphysical or
philosophical predilection; instead, it is the natural consequence of tak-
ing seriously the diversity that issues from the free exercise of practical
reason.
First, then, there is the expressive interest: a direct interest in articulat-
ing thoughts, attitudes, and feelings on matters of personal or broader hu-
man concern and perhaps through that articulation, influencing the
43. There is, for example, this logical distinction: two inconsistent views may both be fully
reasonable, though they cannot both be true.
44. I take this formulation from Mark Johnston.
114
thought and conduct of others.45 When we think of expression quite gen-

erally as a matter of outwardly indicating one’s thoughts, attitudes, feel-
ings (or at least what one wants others to believe those inner states are),
then the importance of the expressive interest may seem elusive. Draw-
ing some distinctions within the general category of expression, however,
will clarify the asserted importance of the interest and one source of the
burdensome quality of regulations of expression.
A feature shared by different evaluative conceptions is that the concep-
tions themselves single out certain forms of expression as especially im-
portant or urgent; the conception itself implies that the agent has weighty
reasons for expression in certain cases or about certain issues.46 The fail-
ure to acknowledge the weight of those reasons for the agent—even if
one does not accept them—reflects a failure to appreciate the fact of rea-
sonable pluralism. Consider in particular three central cases in which
agents hold views that state or imply that they have very strong, perhaps
compelling, reasons for expression, and so three central cases illustrating
the importance of the expressive interest.
(1) In a range of cases, the limiting instance of which is a concern to
“bear witness,” the agent endorses a view that places her under an obliga-
tion to speak out, to articulate that view—and perhaps to urge on others a
different course of thought, feeling, or conduct. Restricting expression in
such cases would prevent the agent from fulfilling what she takes to be
an obligation; it would impose conditions that the agent reasonably takes
to be unacceptable. Here, expressive liberty is on a footing with liberty of
conscience, and regulations are similarly burdensome.47
(2) In a second class of cases, expression addresses a matter of politi-
cal justice. Here the importance of the issue—indicated by its being a
matter of justice—provides a substantial reason for addressing it. The
45. I say “perhaps” because expression often has nothing to do with communication. See C.
Edwin Baker, Human Liberty and Freedom of Speech (Oxford: Oxford University Press, 1989),
51–54. I am grateful to Randall Forsberg for her helpful comments on an earlier draft in which
I characterized the expressive interest much too narrowly.
46. My emphasis on reasons in the description of the expressive interest distinguishes my
treatment from conventional discussion of the value of self-expression and self-fulfillment.
When someone fulfils what they take to be an obligation (as specified by their moral views, for
example), it is wrong to treat this as a matter of self-expression or self-fulfillment.
47. Here I follow a suggestion advanced in Rawls’s discussion of liberty of conscience: that
the rationale for liberty of conscience lies in obligations that religious and moral views assign
to those who hold them, and that this rationale can, in some measure, be extended to other lib-
erties. See Rawls, Theory of Justice, 206.
115
precise content and weight of the reason is a matter of controversy.

Brandeis, for example, urged that “public discussion is a political duty.”48
Perhaps so, but even if expression on such issues is not a matter of duty,
still, it is a requisite for being a good citizen—in some cases, for sheer de-
cency—and as such, is characteristically supported by substantial reasons
within different moral-political conceptions, even though those concep-
tions may disagree about the precise importance of civic engagement and
about the occasions that require it.
(3) In a third class of cases, expression is not a matter of personal obli-
gation, nor does it address issues of justice; rather, it is moved by con-
cerns about human welfare and the quality of human life. The evident
importance of those concerns provides substantial reasons for the expres-
sion. A paradigm here is expression about sexuality—say, artistic expres-
sion (whether with propositional content or not)49 that displays an antipa-
thy to existing sexual conventions, the limited sensibilities revealed in
those conventions, and the harms they are perceived as imposing. In a
culture that is, as novelist Kathy Acker says, “horrendously moralistic,” it
is understandable that such writers as Acker challenge understandings of
sexuality “under the aegis of art, [where] you’re allowed to actually deal
with matters of sexuality.”50
Another paradigm is social satire (or analogously, caricature). Lenny
Bruce’s biographer described him as a “man with an almost infantile at-
tachment to everything that was sacred to the lower-middle class. He
believed in romantic love and marriage and fidelity and absolute hon-
esty and incorruptibility—all the preposterous absolutes of the unquali-
fied conscience . . . Lenny doted on human imperfection: sought it out,
gloated over it—but only so he could use it as a memento mori for his
ruthless moral conscience . . . The attempt to make . . . him a hip-
pie saint or a morally transcendent artiste, was tantamount to missing
the whole point of his sermons, which were ferociously ethical in their
thrust.”51
There are further important cases here—including an interest in creat-
48. Whitney v. California, 274 U.S. 357, 375 (1927) (Brandeis concurring).
49. See, for example, Barnes v. Glen Theatre, Inc., 111 S. Ct. 2456 (1991) (White dissenting).
50. See Kathy Acker, “Devoured by Myths: An Interview with Sylvère Lotringer,” in
Hannibal Lecter, My Father (New York: Semiotext(e), 1991); and the interview of Acker in An-
gry Women, ed. Andrea Juno and V. Vale (San Francisco: Re/Search, 1991), 184–185.
51. Cited in Edward de Grazia, Girls Lean Back Everywhere: The Law of Obscenity and the
Assault on Genius (New York: Random House, 1992), 459–460.
116
ing things of beauty. But the three I have mentioned are central cases of
the expressive interest and suffice to underscore the basis of its impor-
tance. They work outward from the case of fully conscientious expres-
sion, the paradigm of expression supported by substantial reasons from
the agent’s point of view. To be sure, diverse evaluative conceptions carry
different implications about what is reasonable to say and do. But com-
mon ground is that they assign to those who hold them substantial rea-
sons for expression, quite apart from the value of the expression to the au-
dience, even if there is no audience at all.
One alternative line of argument about freedom of expression focuses
entirely on public discussion and locates the contribution of expression
to public debate at the core of the ideal of freedom of expression. Such
views miss the parallels between expressive liberty and liberty of con-
science. As a result, they are insufficiently attentive to the weight of the
expressive interest and are likely to be too narrow in the scope of their
protections.
Cass Sunstein, for example, has recently restated the case for a two-tier
conception of freedom of expression, in which political speech occupies
the upper, stringently protected tier.52 Although Sunstein’s immediate fo-
cus is the proper interpretation of the First Amendment, his case rests in
part on general political values and so intersects with my concerns here.53
Sunstein defines speech as political when “it is both intended and re-
ceived as a contribution to public deliberation about some issue.”54 This
conception of political speech is very broad and is understood to encom-
pass “much art and literature,” because much “has the characteristics of
social commentary.”55 It is not boundless, however, in that it excludes
from highest level protection commercial speech, bribery, private libel,
and obscenity.
Because of the breadth of this conception of political speech, the prac-
tical differences between Sunstein’s approach and mine may, in the end,
turn out to be rather subtle. Still, it strikes me as a mistake to make core
protection contingent on the role of expression in contributing to public
52. See his “Free Speech Now” and his Democracy and the Problem of Free Speech.
53. See in particular his four reasons for special protection for political speech, only the first
of which is concerned specifically with constitutional interpretation, Democracy and the Prob-
lem of Free Speech, 132–137. Of course, his view is also controversial as constitutional interpre-
tation.
54. Ibid., 130.
55. Sunstein, “Free Speech Now,” 308.
117
discussion, in particular on how it is received. Should the level of protec-

tion of, for example, Kathy Acker’s literary exploration of sexuality be
made to depend on whether people find her Hannibal Lecter, My Father
or Blood and Guts in High School challenging or instructive rather than
offensive, disgusting, or, simply, out-of-control, post-modernist, identity-
deconstructive raving?56 Should the level of protection of a doctor’s con-
scientious efforts to advise a pregnant patient on the alternatives available
to her depend on that advice being intended or received “as a contribu-
tion to public deliberation” about reproductive choice?57 Expression of these
kinds is often supported by very substantial reasons, quite apart from how
it is received. As my discussion of the expressive interest indicates, an ac-
count of freedom of expression ought not to disparage those reasons.
In response it might be urged that the justification for establishing
an upper tier occupied by political speech does not depend on an assess-
ment of the relative value of different sorts of speech but rather on an
assessment of their relative vulnerabilities: because the government has
such strong incentives to regulate political speech, it is especially vul-
nerable; because it is so vulnerable, it requires especially strong pro-
tections.58
This response is not convincing. The evidence of special vulnerability
is at best uncertain.59 And in any case, it is a mistake to reduce the special
protection for expression to a matter of vulnerability. As I have indicated
in my discussion of the expressive interest, there are very substantial in-
terests at stake. I see no compelling reasons—of political theory, gen-
eral constitutional theory, or American constitutional tradition—to de-
emphasize the weight of those interests and shift focus to assessments of
vulnerability.60
56. Just for the record, I think it is neither instructive nor raving.
57. Expression that falls outside the upper tier is not for that reason without protection.
Thus the formulation in terms of “level of protection.”
58. See Sunstein, Democracy and the Problem of Free Speech, 136.
59. Absent a precise delineation of the category of political speech, the empirical issue is
hard to adjudicate. But some reasons for doubting the case for special vulnerability are sug-
gested in de Grazia, Girls Lean Back Everywhere; and in William Noble, Bookbanning in
America (Middlebury: Erikson, 1990).
60. On the issue of American constitutional tradition: Sunstein associates his own concep-
tion of freedom of expression with Brandeis’s focus on deliberation rather than Holmes’s mar-
ketplace of ideas. See Democracy and the Problem of Free Speech, 23–28. But Brandeis’s con-
currence in Whitney v. California, 274 U.S. 357, 373–378 (1927), is perhaps the classic
statement of the very great constitutional weight of the interests protected by the right to free-
dom of expression; it is not about the special vulnerability of political speech to government re-
striction.
118
Let us proceed, then, to the second basic interest: the deliberative in-
terest. This interest has two principal aspects. The first is rooted in the ab-
stract idea—shared by different evaluative conceptions—that it is impor-
tant to do what is best (or at least what is genuinely worthwhile) not
simply what one now believes best (or what one now believes to be
worthwhile). For this reason, we have an interest in circumstances favor-
able to finding what is best, or at least what is worthwhile: that is, to find-
ing out which ways of life are supported by the strongest reasons.
The second aspect of the deliberative interest is rooted in the idea that
it is important that one’s evaluative views not be affirmed out of igno-
rance or out of a lack of awareness of alternatives. So alongside the inter-
est in doing what is in fact supported by the strongest reasons, there is
also an interest in understanding what those reasons are and the nature of
the support they give. This, too, leads to an interest in circumstances fa-
vorable to such understanding.
The connection between these two aspects of the deliberative interest
and expression lies in the familiar fact that reflection on matters of hu-
man concern typically cannot be pursued in isolation. As Mill empha-
sized, it characteristically proceeds against the background of an articula-
tion of alternative views by other people.61 So here, again, there is an
interest in circumstances suited to understanding what is worth doing
and what the reasons are that support it—e.g., circumstances featuring a
diversity of messages, forcefully articulated.62
Finally, and most straightforwardly, I assume a fundamental informa-
tional interest in securing reliable information about the conditions re-
quired for pursuing one’s aims and aspirations.
Having described these three interests, I want to return to the com-
61. This is the force of Mill’s contention that censorship robs the human race, and that for
this reason, it does not matter whether all censor one or one censors all. Mill does not focus on
the harm or robbery to the person who is censored. See On Liberty (Indianapolis: Hackett,
1978), 16 (chap. 2, par. 1).
62. Robert Post has suggested a tension among various conditions required for satisfying the
deliberative interest in a diverse community. In particular, deliberation depends on civility, but
requiring civility puts the community in danger of making one particular understanding of ci-
vility authoritative for the community. I am not sure how deep this tension goes. To be sure, ci-
vility has its place in public deliberation. But so do anger, disgust, bitter criticism, and open ex-
pressions of hostility. Post’s immediate concern is with a parody of Reverend Jerry Falwell in
Larry Flynt’s Hustler magazine. Suffice it to say here that the parody of Falwell was, in my
view, a contribution to public debate, even if it was not civil and was not an invitation to
Falwell to have a conversation with Larry Flynt. See Robert Post, “The Constitutional Concept
of Public Discourse: Outrageous Opinion, Democratic Deliberation, and Hustler Magazine v.
Falwell,” Harvard Law Review 103 (1990): 601–686.
119
plaint I registered earlier about the sectarianism of autonomy-based,

maximalist views of freedom of expression. For it may now seem that my
own view is not, after all, so sharply distinct from them. I will respond
briefly by noting three sorts of difference.
First, autonomy has a capaciousness that strikes me as a vice in an ac-
count of expressive liberty. Each of the basic interests that I have men-
tioned is sometimes included within the value of autonomy, but each is
importantly different. Bringing out these differences helps both to clarify
the ways that those interests can be seen to be important within different
evaluative conceptions and to provide the basis for a theory of expressive
liberty that is able to capture intuitive distinctions among different sorts
of expression. Second, I am not supposing—with the maximalist—that
the three interests always trump other values. Nor, third, do I assume that
the interests are uniformly implicated in different sorts of expression. For
example, I do not think that they are equally at stake in commercial, po-
litical, and artistic expression. (I will return to this point in my discussion
of categorization in Section 4.)
There are, then, at least these three basic interests rooted in diverse, de-
terminate evaluative conceptions and in the second-order concerns col-
lected under the deliberative rubric.63 A first component of the case for
stringent protection, then, lies in the ways that such protection secures fa-
vorable conditions for advancing these fundamental interests. In the case
of the expressive interest, the grounds for protecting expression lie in the
importance of the expressive activity itself, as specified by the agent’s rea-
sons; in the case of the deliberative and informational interests, the
grounds for protecting expression lie in the importance of the interests to
which expression contributes. In short: the reasons for protection are
partly intrinsic, partly instrumental. I see no basis for deciding (nor any
reason to decide) which is more fundamental.
I will return later to a more detailed discussion of connections. But,
first, the costs and background facts.
63. Joseph Raz proposes that part of the case for a right to freedom of expression turns on the
“fundamental need for public validation of one’s way of life” together with the fact that acts of
expression serve the purpose of “validation” in three ways: (1) they inform the public about
“ways of life common in certain segments of the public,” (2) they reassure “those whose ways of
life are being portrayed that they are not alone,” and (3) they provide a “stamp of public ac-
ceptability” for those ways of life. See “Free Expression and Personal Identification,” Oxford
Journal of Legal Studies 11 (1991): 311, 324. I agree with Raz about the importance of the interest
in validation, but I believe that it can be accounted for in terms of expressive, deliberative, and
informational interests.
120
cost s
What then of the costs of expression? Commentators since Justice
Holmes have noted that protection for expression cannot be premised on
faith in its impotence.64 As Harry Kalven put it, “Speech has a price. It is
a liberal weakness to discount so heavily the price. [It] is not always cor-
rect to win [the protection of speech] by showing danger [it threatens]
has been exaggerated.”65 Underscoring Kalven’s point about the price of
speech and the weakness of characteristic arguments for protection, re-
cent “outsider” jurisprudence has portrayed the injuries that hate speech
imposes on its targets by narratively recounting those injuries.66 If we ab-
jure both the minimalist denial of the price and the sectarian route of
maximalism, then the idea of stringent protections may seem just inde-
fensible, and the skeptical response—“there’s no such thing”—may seem
a natural alternative.
What then are some of the kinds of costs that expression imposes? In
answering this question, I want to organize costs along just one axis, dis-
tinguishing three types by the pattern of their etiology.
First, there are direct costs. Here I have in mind cases in which, intu-
itively, nothing intervenes in between the expression and its price, where
“the very utterance inflicts injury”:67 I shriek at a neurasthenic with a
weak heart; disrupt the peace and quiet with loud shouting; falsely tell an
elderly parent that her child has just died; spread defamatory falsehoods
about a colleague; use offensive language in a public setting; offer a raise
or a higher grade in return for sex. When I have said my piece, the dam-
age is done, and it is done by what I said—and in the latter four cases, by
its content.
A second category of costs are “environmental.” Thus expression may
help to constitute a degraded, sickening, embarrassing, humiliating, ob-
trusively moralistic, hypercommercialized, hostile, or demeaning envi-
ronment. It might, for example, combine with other expressive actions to
64. Abrams v. U.S., 250 U.S. 616, 629 (1918) (Holmes dissenting).
65. Cited from Kalven’s notes in the “Editor’s Introduction” to A Worthy Tradition, xxii.
66. See in particular Matsuda, “Public Response to Racist Speech”; Charles Lawrence, “If
He Hollers Let Him Go”; Delgado, “Campus Antiracism Rules.”
67. Chaplinsky v. New Hampshire, 315 U.S. 568, 571–572 (1942). Or, where the “evil” is “cre-
ated by the medium of expression itself,” as in the case of signs posted on utility poles, as dis-
tinct from leaflets handed to individuals. See City Council v. Taxpayers for Vincent, 466 U.S.
789, 810 (1984).
121
contribute to an environment of racial or national antagonism or to one

in which dominance and submission are erotized. Here the harm is not
the expression by itself, since in the absence of other similar sayings, the
environment would not be degraded, hypercommercialized, or hostile;
nor can we trace particular harmful or injurious consequences to partic-
ular acts of expression that help to constitute the unfavorable environ-
ment.68 Instead, the price of the expression lies in its contribution to mak-
ing an environment hostile to, for example, achieving such fundamental
values as racial or sexual equality.
Finally there are straightforwardly indirect costs. Here the injury re-
sults from the expression causing (by persuading, suggesting, or provid-
ing information) someone to do something harmful, as when someone
persuades others to purchase too much of a scarce resource or to join
the Ku Klux Klan or to support a war that results in massive death and
destruction.
ba ckground fa c t s
To complete the picture of the bases of stringent protections, I now come
to the background facts.69 These facts are the set of sociological and an-
thropological claims that plays a central role in arguments about freedom
of expression, though often only as an implicit, half-articulated, and so
easily manipulable background.70 Whatever their common treatment,
their importance will eventually become clear. My aim here is simply to
make them explicit.
I will group the facts into three broad categories, which I label the
Facts of Reasonableness, the Bare Facts of Life, and the Unhappy Facts of
68. A work environment, for example, may be actionably hostile under Title VII of the Civil
Rights Act because of either the severity or the pervasiveness of conduct. See Meritor Savings
Bank v. Vinson, 477 U.S. 57, 67 (1986). For discussion of severity and pervasiveness, see Ellison
v. Brady, 924 F.2d 872 (9th Circuit, 1991).
69. Scanlon discusses the importance of “linking empirical beliefs” in arguments about
right to expressive liberty. His discussion overstates, I believe, the importance of beliefs spe-
cifically about the role of government in suppressing expression. See “Freedom of Expression
and Categories of Expression,” 534.
70. For a general discussion of the tendency of legal-doctrinal argument to suppress refer-
ence to background factual assumptions, see Roberto Unger, The Critical Legal Studies Move-
ment (Cambridge, MA: Harvard University Press, 1982). In the case of freedom of expression,
Bollinger claims that the “fortress model” of speech protection presumes a set of irrational ten-
dencies to suppress speech that conflicts with the assumption of rational competence that
drives the ideal of an open market of ideas. So the conjunction of these two doctrines in a jus-
tification of free expression is untenable because it requires inconsistent background beliefs.
See The Tolerant Society, 92–93.
122
Life. Intuitively, the difference among the three categories is that the
facts of reasonableness are considerations that would favor the protection
of speech even under fully ideal conditions; the bare facts favor protec-
tion and are unalterable; the unhappy facts of life are considerations that
now favor protection but that we might hope are alterable features of our
circumstances.
Among the Facts of Reasonableness are:
(1) The Fact of Reasonable Pluralism: Under conditions of expressive

liberty, people will arrive at conflicting, reasonable evaluative con-
victions.
(2) The Fact of Reasonable Persuasion: People have the capacity to
change their minds when they hear reasons presented, and some-
times they exercise that capacity. This is the assumption of
Brandeis’s remark that “if there be time to expose through discus-
sion the falsehood and fallacies, to avert the evil by the process of
education, the remedy to be applied is more speech, not enforced si-
lence.”71 But for the Fact of Reasonable Persuasion, more speech
would be a diversion rather than a remedy.
As Bare Facts of Life, we have:
(1) The Fact of Resource Dependence: Expression depends on re-

sources, and the access to those resources is commonly unequally
distributed.
(2) The Fact of Innocent Abuse: If expression is relatively uninhib-
ited, people will sometimes—even without malign intent—say
things that are false, offensive, insulting, psychically injurious,
emotionally distressing, and reputationally damaging. As James
Madison put it, “Some degree of abuse is inseparable from the
proper use of everything.”72
(3) The Cold (Chilling) Facts: If sanctions are attached to expression
for being false, offensive, insulting, psychically injurious, for exam-
ple, then people will be reticent to express themselves (chilled),
even if they think their expression is true, inoffensive, not insult-
ing, or the like. Moreover, if the regulation of expression proceeds
in ways that are highly uncertain—because standards are vague
71. Whitney v. California, 274 U.S. 357, 375–376, 377 (1927) (concurring).
72. Cited in New York Times v. Sullivan, 376 U.S. 254, 271 (1964).
123
(e.g., if sanctions attach to remarks that are offensive, deeply dis-

turbing, “outrageously” insulting)73 or their application depends
on a weighing of competing considerations in each case—then
many people will be reticent to express themselves, even if their
views deserve protection.
Finally, I count among the Unhappy Facts of Life:
(1) The Fact of Power: Most people—and those with power in partic-
ular—do not like to be criticized or disagreed with, and are
tempted to use the means at their disposal to avoid criticism or dis-
agreement.74
(2) The Fact of Bias: We tend, as a general matter, to confuse what we
would prefer other people to do with what would be best for them
to do or with what they must do on pain of immorality.75
(3) The Fact of Disadvantage: In a society with relatively poor and
powerless groups, members of those groups are especially likely to
do badly when the regulation of expression proceeds on the basis
of vague standards whose implementation depends on the discre-
tion of powerful actors.
(4) The Fact of Easy Offense: Putting sociopaths to the side, everyone
is offended by something.76
(5) The Fact of Abuse: Against a background of sharp disagreement,
efforts at persuasion sometimes proceed through exaggeration,
vilification, and distortion.77
4. An Alternative Strategy: Implications

I want now to bring the different pieces together into a case for a scheme
of stringent protections. I will proceed through the four themes discussed
73. On the problems of regulating outrageous insults, see Hustler v. Falwell, 485 U.S. 46
(1988).
74. This is, of course, true of public officials but hardly unique to them. It is commonly said
that they are especially untrustworthy, but I know of no evidence for the claim that political
power breeds arrogance more surely than economic power.
75. As Mill put it: “No one acknowledges to himself that his standard of judgment is his own
liking.” See On Liberty, chap. 1, par. 6. The point is not that standards of judgment are, in gen-
eral, simply matters of liking and disliking, but rather that even when they are, we do not see
them that way.
76. Bollinger in effect argues that there is a fundamental conflict between appealing to the
Facts of Reasonableness and to the Unhappy Facts of Life. See The Tolerant Society, 92–93. I
do not see the conflict.
77. Paraphrasing Justice Roberts in Cantwell v. Connecticut, 310 U.S. 296, 310 (1940).
124
in Section 1, showing how each can be explained by reference to the ele-

ments I have just sketched. In my explanation I place principal emphasis
on the expressive and the deliberative interests and on the Facts of Rea-
sonableness. The aim is to show that stringent protections are driven
principally by the substantive value of expression and the possibilities of
using speech to combat the harms of speech; such protections are only
secondarily remedial, only secondarily driven by fear and mistrust under-
written by our tendency—or the tendency of government—to under-
value or suppress expression.
cont ent regul a t i on

Take first the presumption against content regulation. This presumption
is driven in part by the fundamental expressive and deliberative interests.
Content regulation presents the possibility that regulation may effectively
exclude certain views from the marketplace, not only driving them into
another market niche but also driving them out altogether. Content-
neutral regulation may present that possibility as well, but the threat
from content-discriminatory regulations is greater because the targeting
is more precise. Because of this threat, content regulations pose a more
substantial danger that people will be prevented from expressing views
despite, as they see it, the existence of substantial reasons for such expres-
sion. In short, they represent a direct threat to the expressive interest.
Moreover, the limits imposed by content regulations on the range of
messages threaten the deliberative interest. By directly reducing the di-
versity of expression, they distort, as Meiklejohn said, the “thinking pro-
cess of the community.”78 More immediately, by restricting the range of
views and establishing official dogma, they limit reflection on alternative
views and so on the reasons for holding one’s own views. The problem is
not that content regulation keeps people from being persuaded to change
their minds; rather, it prevents us from figuring out just what our minds
are on some subject and what the reasons are for not changing them.
The Fact of Power points in the same direction. Those with power of-
ten wish to insulate themselves from criticism, and the power to regulate
content is an especially refined instrument of such insulation. This is par-
ticularly true of viewpoint regulation. By contrast, content-neutral regula-
tions are more blunt and so less desirable instruments of insulation. To
78. Alexander Meiklejohn, Political Freedom (New York: Harper & Brothers, 1960), 27.
125
be sure, blunt instruments are still instruments. And if someone expects

the distribution of messages to be unfavorable, that someone will want to
reduce the level of expression. Moreover, content-neutral regulations can
have more or less transparently discriminating effects as to classes of
speaker. So content-neutral regulations, too, raise serious concerns. But
the point suggested by the Fact of Power remains: there is typically no
motivation to reduce the quantity of expression of the same kind and in-
tensity as the motivation to target certain topics, or more particularly, cer-
tain messages. So content-neutral regulations are often less troublesome.
These considerations about the interests and the Fact of Power indicate
why content regulation is especially troubling. Given those troubles, the
Fact of Reasonable Persuasion helps to secure the case for a presumption
against such regulations. It suggests that the damaging consequences of
expression with objectionable content can—apart from the case of direct
costs—be addressed with more expression. Because such address is pref-
erable to imposing sanctions, we ought to establish a general presump-
tion in favor of relying on it.
In the case of political speech, for example, these pressures for protec-
tion exercised by the basic interests, and the Facts of Power and Reason-
able Persuasion are very strong. So some rule of the sort advanced in
Brandenburg v. Ohio is naturally suggested: that advocacy of violent polit-
ical change can legitimately be restricted only when “such advocacy is
directed to inciting or producing imminent lawless action and is likely to
incite or produce such action.”79 Advocacy of the kind recorded in this
rule is not the only kind that threatens harm, nor is the expected value of
the harm necessarily the greatest. But it is the only case in which circum-
stances prevent the pursuit of the preferred remedy.
ca t egoriz a t ion
We come next to categorization as an approach to handling content regu-
lations. Recall that the idea of categorization is to confine exceptions to a
general presumption against content regulation by singling out a small
set of categories of expression—for example, child pornography, com-
mercial speech, obscenity, fighting words, and express incitement—for
lesser protection, specifying conditions for permissible regulation of ex-
pression in each category. The rationale for this strategy divides naturally
into two parts.
79. Brandenburg v. Ohio, 395 U.S. 444, 447 (1968).
126
We need first to account for the distinctions between more and less im-
portant kinds of expression. Judgments of importance proceed princi-
pally by considering the connection of the expression to the fundamental
interests and secondarily by considering the prospects of addressing the
harms through more expression and the fragility of expression given the
Bare and the Unhappy Facts of Life.
Political expression, for example, is especially important because it is
so closely connected to each of the basic interests and because of its fra-
gility in light of the Fact of Power. Because it is commonly a form of po-
litical speech, expression that libels a group is more strongly connected
to expressive and deliberative interests than expression that threatens in-
dividual libel is; the injuries are also more easily remedied with group li-
bel than with individual libel. For these reasons it is important to confine
reduced protection to a category of individual libel, even though people
may be harmed by libeling groups to which they belong or with which
they identify. The idea that group libel ought to be more strongly pro-
tected than individual libel is not contingent on a liberal individualist
failure to acknowledge the possibility of harm through group libel, any
more than the protection of the libel of public figures requires a denial of
its harm.80
Commercial speech can be a source of information. But it is less im-
portant than political expression, because it is not so closely connected to
the expressive or the deliberative interests.81 Moreover, the Cold Facts
and the Fact of Limited Understanding have much less force in the case
of commercial speech.82 The economic interests that fuel commercial
speech ensure that it is less susceptible to a regulatory chill, while the fact
that commercial advertisers are best situated to know the accuracy of
80. Here I disagree with the defense of group libel laws in Note, “A Communitarian De-
fense of Group Libel Laws,” Harvard Law Review 101 (1988): 682–701. The main reason for re-
jecting regulations of group libel is not that such libel is harmless. See, for example, Justice
Black’s dissent in Beauharnais v. Illinois, 343 U.S. 250 (1952). Black does not deny the costs but
emphasizes the extensive “inroads” on expression that would result from accepting regulations
of group libel.
81. C. Edwin Baker proposes that commercial speech ought to have no First Amendment
protection, in part because such speech reflects the coercive logic of profit-maximization
rather than the choice of the speaker. This explanation strikes me as a strained defense of the
idea that regulations of commercial speech are less burdensome, not least because it suggests
that advertising by price-setting monopolists, which is less subject to the coercive demands of
profit-maximization, ought to be more protected. See Baker, Human Liberty and Freedom of
Speech, chap. 9.
82. See Virginia State Board of Pharmacy v. Virginia Citizens Consumer Council, 425 U.S.
748, 771 (n. 24) (1976), and 777–781 (Stewart concurring).
127
their claims reduces concern about the chilling effects of requiring accu-
racy in commercial speech.
But—and here we come to the second part of the case—even if we can
provide an account of relative importance, why filter judgments of rela-
tive importance by categorization rather than proceed case by case?83
Here the main burden is carried by the Chilling Facts and the Unhappy
Facts of Power, Bias, and Disadvantage. Together they induce an expec-
tation that ad hoc regulation will err on the side of excessive interference,
on the side of underprotection of what should be protected. Moreover,
ad hoc judgments are likely to raise greater concerns about chilling ex-
pression. Categories, then, serve as a protective device—a device of self-
binding—against excessive interference in a context in which a very sub-
stantial value is at stake.
To elaborate: unless expression falls into a less protected category, we
impose very high barriers to regulating it. And before we can consider
more substantial regulation of some act of expression, we need to find a
general category into which it falls such that we are prepared to reduce
the protection for all expression in that general category. The result may
be greater protection for some expression than we are inclined to think
suitable.84 But—if the facts are right—then the alternative would be in-
sufficient protection to some expression. Of course, the claim that cate-
gorization plays this role premises that the categories are—whether for
semantical or psychological reasons—not so utterly manipulable and in-
determinate that they serve no channeling function at all. If they are not,
if the facts are as stipulated, and if the choice of regulatory form does
have the proposed consequences, then it is reasonable to pursue the strat-
egy of protection through categorization.
digression: nihil ism re dux

Earlier I mentioned free-speech nihilism: the idea that “there is no such
thing as free speech.” The pieces are now in place for a response to it. So
I will digress here to provide a reply.
What does the nihilist denial of free speech come to? Echoing Holmes’s
remark that “every idea is an incitement” and Kalven’s “speech has a
price,” Fish explains it this way: “There is no such thing as ‘speech alone’
or speech separable from harmful conduct, no such thing as ‘mere
83. My discussion here is influenced by Ely, “Flag Desecration,” 1496–1502.
84. This concern is expressed in Hustler v. Falwell, 485 U.S. 46 (1988).
128
speech’ or the simple nonconsequential expression of ideas.”85 Beginning

from these familiar observations, Fish arrives at the conclusion that deci-
sions about the permissibility of speech always require a balancing of
benefits and costs in particular cases by reference to “some particular par-
tisan vision.”
I have two disagreements with this conclusion: first, with the idea that
decisions about cases are always a matter of ad hoc balancing, and sec-
ond, with the idea that such balancing must proceed by reference to a
particular partisan vision.
As to the first: Fish in fact acknowledges the importance of general
categories and principles in deciding how to handle particular cases, and
for roughly the reasons I just sketched in my remarks on categorization.
He says that “free speech principles function to protect society against
over-hasty outcomes; they serve as channels through which an argument
must pass on its way to ratification.”86 This acknowledgement of the role
of “free speech principles” in protecting against “over-hasty outcomes”
shows that Fish is not really—as it might have seemed—offering balanc-
ing as the mandatory way to resolve particular cases. Neither metaphysics
nor politics condemns the resolution of cases by reference to general,
“free speech principles” that serve (as we see it) to tie our hands against
“over-hasty outcomes.” So the mere fact that speech is consequential car-
ries no implications at all about the proper—much less the necessary—
forms for the regulation of expression.
If free-speech nihilism is not nihilism about principles and a corre-
sponding embrace of ad hoc balancing as the proper form of regulation,
then perhaps it registers a point about the justification of the princi-
ples used to decide cases: that because speech is “never free of conse-
quences,”87 any justification of principles for resolving cases must take
into account the values that a scheme of restrictions and permissions pro-
motes and the costs it imposes.
This thesis is indisputable, but it is also uncontested. Justice Black,
for example, urged free-speech absolutism as a doctrine about decision-
making under the First Amendment—“no law” means “no law”—not as a
theory about the justification of that amendment. He did not deny the
85. Stanley Fish, “There’s No Such Thing as Free Speech and It’s a Good Thing Too,”
Boston Review 17, 1 (1992): 3–4, 23–26.
86. See ibid., 26. I am indebted to Duncan Kennedy for discussion of this point.
87. Ibid.
129
importance of a “balancing of conflicting interests” at the level of the jus-

tification of the First Amendment prohibition on laws restricting free-
dom of speech; he just thought that the authors of the First Amendment
did all the balancing necessary when they settled on the phrase “shall
make no law.”88
Perhaps, then, free-speech nihilism consists neither in the rejection of
principles as guides to decision-making, nor simply in the claim that a
justification of such principles must take the consequences of speech
into account. Perhaps it is the claim that justification must always pro-
ceed in terms of the aims, interests, and aspirations of particular groups,
in terms of “some particular partisan vision”:89 i.e., there are no common
or shared interests that can serve as a basis for justification. Thus under-
stood, nihilism suggests a pair of practical precepts: If you are weak, ar-
gue as forcefully as you can for an encompassing protection of speech in
the hope of gaining some political space for your vision; if you are strong,
“refashion” principles “in line with your purposes,” and then “urge them
with a vengeance.”90
But—here I come to my second disagreement—expressive, delibera-
tive, and informational interests do, I claim, provide common ground
among a range of genuinely different views and “particular partisan vi-
sion[s].” Of course neither those interests nor any other general scheme
of values resolves all controversy about specific cases. But if nihilism
amounts only to the thesis that judgments in this area are controversial
and contestable, then it wins a quick and uninteresting victory.
Some views, to be sure, do deny the importance of expressive and de-
liberative interests, so it might be said that the endorsement of those in-
terests is itself partisan. But partisanship in this sense—not being ac-
cepted by all—is consistent with holding that these interests provide
common ground for a wide range of distinct moral-political views; they
are not the exclusive possession of one particular partisan vision. As to the
views that deny these interests, we need to consider actual cases in order
to see whether the positions have any serious claim to be reasonable, and
whether the partisanship they embrace is not still more narrow and par-
ticular. In short, we need to consider cases to decide whether the parti-
sanship is really troubling.
88. Black, “The Bill of Rights,” New York University Law Review 35 (1960): 879.
89. Fish, “There’s No Such Thing,” 26.
90. Ibid.
130
Take, for example, the case of the “rationalist fundamentalist.”91 This is

the person who denies the idea of reasonable pluralism, affirming instead
that it lies within the competence of reason to know that salvation is the
supreme value, that there is a single path to salvation, that there is no sal-
vation among the damned, that there are no expressive and deliberative
interests, and that free expression is to be condemned along with liberty
of conscience. This is not a common view, if only because it claims for
reason territory more commonly reserved for faith.92 But if someone ad-
vances it, then one ought to say that she is simply mistaken about the
powers of reason.93 Even if the views of the rationalist fundamentalist are
all rationally permissible, reason surely does not mandate them, and in
insisting that it does, the fundamentalist is not acknowledging the facts.
So the fact that expressive and deliberative interests are not recognized by
the rationalist fundamentalist does not seem very troubling.94 To be sure,
other cases may present greater difficulties. But that needs to be shown. It
is not enough simply to point to the fact of disagreement and conclude
that there are only particular partisan visions.
Let us now return to the case for stringent protections.
cost l y p rot ect io ns

What, then, about protecting expression despite its costs? Why is the fact
that expression imposes conditions that are reasonable to want to avoid
not sufficient to remove the presumption of protection from it?
To address this question, let us start with the special case of offensive
expression—in particular, expression that shocks the sensibilities. There
is no way to ensure fair opportunities for expression while at the same
time protecting people generally from offensive expression. Given the
Fact of Easy Offense and the associated ubiquity of offense, such protec-
tion would have to take the form of substantially restricting expression.
But the weight of the interests that underlie expression—the expressive
91. I take this example from “Moral Pluralism and Political Consensus,” 286.
92. It is an analog to “creation science,” operating in the domain of salvation. The proper re-
sponse is the same in both cases.
93. Most fundamentalists are not rationalist fundamentalists, and would I think agree with
this response.
94. I offer the rationalist fundamentalist simply as one illustration. The case of the
nonrationalist fundamentalist—who affirms that the basis of religious conviction lies in faith—
is more complicated. The latter may wish to distinguish truths delivered by faith from the bases
of political justification, and so may be prepared to acknowledge expressive and deliberative
interests, at least in the context of political argument.
131
and deliberative interests—is much greater than the weight of the interest
in not being offended, so those restrictions would be intolerable. More-
over, it will not help to confine efforts to regulate to the “grossly offen-
sive”—then we will run into troubles because of the chilling effects of
vagueness. The likely vagueness in regulations of the “grossly offensive,”
that is, threatens to chill acceptable expression.95
I do not deny that there is a cost or “price” to offensive expression; in-
deed the cost is direct. Instead, I claim that the costs of avoiding offense
are to be borne by those subject to it—they must, for example, “avert
their eyes.”96
Offensive expression is, as I said, a special case. Moving beyond it,
then, the general strategy in deciding whether to protect expression de-
spite its price is to consider the importance of the expression (with atten-
tion to the role of categories), the directness and seriousness of the harm,
and the vulnerability of the expression to underprotection, given the
background facts. Let me illustrate with three kinds of case.
In cases of the first type, expression belongs to an important category, it
is vulnerable, and the costs are either environmental or indirect. Then
the reasons against restriction are especially strong, even if the cost is sub-
stantial.
Consider, for example, the pornography ordinances adopted in Min-
neapolis and Indianapolis in the 1980s. According to the Indianapolis or-
dinance, pornography is the “graphic, sexually explicit subordination of
women, whether in pictures or in words,” which also meets one of the
following conditions:
(i) women are presented as sexual objects who enjoy pain or

humiliation; or (ii) women are presented as sexual objects who
experience sexual pleasure in being raped; or (iii) women are
presented as sexual objects tied up or cut up or mutilated or
bruised or physically hurt; or (iv) women are presented being
penetrated by objects or animals; or (v) women are presented in
scenarios of degradation, injury, torture, shown as filthy or infe-
rior, bleeding, bruised, or hurt in a context that makes these
conditions sexual; or (vi) women are presented as sexual objects
for domination, conquest, violation, exploitation, possession, or
95. See Hustler v. Falwell, 485 U.S. 46 (1988).
96. See Cohen v. California, 403 U.S. 15 (1971).
132
use, or through postures or positions of servility or submission

or display.97
As this language indicates, those ordinances—by contrast with obscenity

regulations—included no provision for the artistic, literary, scientific, or
political value of the expression they sought to regulate. So they were in-
attentive to the importance to the expressive, deliberative, and informa-
tional interests associated with expression on issues of sexuality. But ex-
pressive interests are important in this area because advancing views
about human sexuality is supported by substantial reasons from the point
of view of the expresser. I noted this in my earlier discussion of expressive
interests. Moreover, deliberative and informational interests are at stake
as well:
[The existence of pornography] serves some social functions

which benefit women. Pornographic speech has many, often
anomalous, characteristics. One is certainly that it magnifies
the misogyny present in the culture and exaggerates the fan-
tasy of male power. Another, however, is that the existence of
pornography has served to flout conventional sexual mores, to
ridicule sexual hypocrisy and to underscore the importance
of sexual needs. Pornography carries many messages other
than woman-hating: it advocates sexual adventure, sex outside
of marriage, sex for no other reason than pleasure, casual sex,
anonymous sex, group sex, voyeuristic sex, illegal sex, public
sex.98
Apart from their inattention to basic interests, the ordinances were

vaguely drawn, suggesting inattention to the historical vulnerability of
sexual expression to overregulation for moralistic reasons.99 And they did
not consider alternative ways to address the injuries that they associated
with pornography. For example, if the problem with pornography is that
it sexualizes and thereby legitimates abuse, then one natural step would
97. Indianapolis, Ind., City-Council General Ordinance No. 35 (June 11, 1984), cited in
MacKinnon, Feminism Unmodified, 274, n. 1.
98. Lisa Duggan, Nan Hunter, and Carole Vance, “False Promises: Feminist Antipornog-
raphy Legislation,” in Caught Looking: Feminism, Pornography, and Censorship (East Haven,
CT: Long Rivers Books, 1992), 82. The authors were members of the Feminist Anti-Pornogra-
phy Task Force.
99. Though the rationale for the regulations was emphatically not moral. See Catharine
MacKinnon, “Not a Moral Issue,” in Feminism Unmodified.
133
be to target sexual abuse—the abuse of women as women—directly and

seriously. Such targeting might, for example, include a tort of domestic
sexual harassment modeled on workplace sexual harassment—including
elements of quid pro quo and hostile environment harassment.100 If the
injury of pornography is that it silences women, then—taking seriously
Brandeis’s idea of combating the harms of speech with more speech—
there could be regular public hearings on sexual abuse, perhaps subsidies
for women’s organizations to hold such hearings,101 or easier access of
women to broadcast licenses.
To be sure, the regulations of pornography did claim to be addressed to
the harms of pornography. But in the face of the breadth of coverage of
the regulations, and given the importance of what they were target-
ing, the claims about costs seem too speculative to sustain the case for
regulation.
But these criticisms of the speculative character of the connections be-
tween the widespread consumption of pornography and its alleged costs
derive their force in part from the broad sweep of the regulations and so
from the importance of the expression they sought to regulate. The case
does not rest entirely on freestanding doubts about the speculative quality
of the connections between the expression and the costs. Less sweeping
regulations, drafted with more attention to the value of sexual expression,
ought to trigger correspondingly less concern about the need for a con-
clusive showing of injury and so demand less exacting scrutiny.
Consider, for example, a regulation targeted on the “pornographi-
cally obscene”: the subset of the constitutionally obscene (prurient, of-
fensive, and minimally valuable expression) that erotizes violence. The
case against this regulation would be weaker, because of the weak rela-
tion of obscenity to the fundamental interests. Given that weak relation,
it is less important that the costs are not direct and the arguments in sup-
port of the costs are speculative.102 (I will provide a more detailed case for
this conclusion later on.)
100. I take the proposal from Duncan Kennedy, “Sexual Abuse, Sexy Dressing and the
Eroticization of Domination,” New England Law Review 26 (1992), 1318.
101. For a more general discussion of associative approaches to reconciling egalitarian and
liberal commitment, see Joshua Cohen and Joel Rogers, “Secondary Associations and Demo-
cratic Governance,” Politics and Society 20 (1992): 393–472.
102. See, for example, Cass Sunstein’s proposal in “Pornography and the First Amendment,”
Duke Law Journal 4 (1986): 589–627. He sharply narrows the class of pornographic expression,
defining the class in a way that aims to make it low value. It amounts, more or less (and implic-
itly), to substituting “erotizes violence and subordination” for “offensive” in the definition of
obscenity.
134
I come now to a second type of case: here, expression belongs to an im-

portant category and is vulnerable, but the costs are direct and unavoid-
able. It is still to be protected. Paradigms here are expressions that cause
emotional distress or reputational injury to public figures.
Consider, for example, the case of Hustler v. Falwell. In a Hustler
parody of a Campari ad, Reverend Jerry Falwell was represented as hav-
ing had his first sexual encounter while drunk in an outhouse with his
mother. Falwell won a substantial settlement for the intentional inflic-
tion of emotional distress. The Supreme Court overturned the settle-
ment, rejecting the idea that tort law protections should define the scope
of expressive liberty. Without denying the reality of Falwell’s distress, or
dismissing it as merely “mental” or emotional, or disputing Hustler’s re-
sponsibility for it,103 the Court nevertheless argued that the parody was
protected, absent a showing of actual malice. The decision did not sim-
ply protect offensive expression; emotional distress is not a matter of be-
ing offended. Nor did it reflect the view that the liberty to inflict emo-
tional distress is in general of greater weight than the injury of such
distress. The decision turned instead on Falwell’s standing as a public fig-
ure and the importance of freewheeling, sharp criticism of public figures.
In a world in which carefully crafted personal images play a central role
in politics, and in which fundamental interests are dependent on the
operation of the political arena, equally well-targeted efforts at deflation
deserve strong protection. By requiring actual malice, the Court in effect
licensed increased emotional distress in order to protect the values associ-
ated with expressive liberty.
In a third type of case, importance and vulnerability diminish, and
there are direct costs. Here, restriction is permitted. Take, for example,
the case of libel of private figures. The vulnerability of reputations, the
difficulty of repairing them through more speech, and the fact that such
libel is typically not supported by weighty expressive or deliberative inter-
ests combine to reduce the appropriate level of protection.
fa ir a ccess
Finally, we come to the requirement of ensuring fair access to expres-
sive opportunities. Three main lines of argument converge on this con-
clusion.
The first begins by underscoring the central role played in the ac-
103. There was, for example, no suggestion that Falwell was really responsible for the distress
because of a hypersensitivity to accusations of sin.
135
count of stringent protections by the Fact of Reasonable Persuasion and

Brandeis’s associated counsel that we remedy the harms of speech with
more speech. By holding out the hopeful prospect of reconciling strin-
gent protection of expressive liberties with other substantial political val-
ues (including the value of equality), Brandeis’s point helps to remove
the sectarian edge from freedom of expression. Instead of winning argu-
ments by always insisting that the “danger has been exaggerated,” we take
the costs seriously and embrace expression as the preferred strategy for
addressing them.
But if we help ourselves to Brandeis’s thesis, then we must also take its
implications on board. When Brandeis urged more speech in the case of
Whitney v. California, the context was subversive advocacy.104 But his re-
marks were not addressed to the advocates: Anna Whitney was using
speech; the state was shutting her up. Brandeis was reminding political
elites of the vast means at their disposal for addressing arguments for revo-
lutionary change: they might, for example, try to cure the social ills that
prompt them or to argue the case against a revolutionary solution.
Addressed to less powerful groups, with restricted access to means of
expression, the easy injunction “More speech!” loses its edge. If we insist
that “more speech” is the preferred remedy for combating the harms of
speech, and appeal to the Brandeisian injunction as part of a case against
content regulation, then we also have an obligation to ensure fair access
to facilities of expression where the additional speech might plausibly
help the “deliberative forces” to “prevail over the arbitrary.”105 Put other-
wise, any argument in which Brandeis’s thesis figures as a premise must
count assurance of fair access among its conclusions. It is simply unac-
ceptable to impose a high burden on justifying restrictions on expression,
to justify that burden in part in terms of the possibilities of combating the
harms of speech with more speech, and then not to endorse the require-
ment of ensuring such facilities.
A second line of argument for fair access is rooted in the expressive in-
terest. The argument follows a generic egalitarian strategy of argument
for substantively egalitarian norms. Described abstractly, the strategy
begins with a more formal and less controversial political norm—for ex-
ample, the norm of formal equality of opportunity—and then argues
that the best justification for that norm actually provides a rationale for
104. Whitney v. California, 274 U.S. 357 (1927) (Brandeis concurring).
105. Ibid., 375.
136
a more egalitarian norm—for example, substantive equality of opportu-

nity.106
To put the point less abstractly and apply it to the issue at hand: the ra-
tionale for the more formal requirement of an equal right to expressive
liberties rests centrally on a conception of the human interests served by
that guarantee. More specifically, the reason for protecting expressive lib-
erties against content regulation or other forms of undue restriction lies
in part in the importance of assuring favorable conditions for the pursuit
of the expressive interest. But once we acknowledge the need for favor-
able conditions for realizing this basic interest, we are naturally led from
a more formal to a more substantively egalitarian requirement, since the
latter more fully elaborates the range of favorable conditions. In particu-
lar, given the Fact of Resource Dependence, favorable conditions for re-
alizing the expressive interest will include some assurance of the re-
sources required for expression and some guarantee that efforts to express
views on matters of common concern will not be drowned out by mes-
sages of better endowed citizens.
The deliberative interest provides the foundation for a third, and more
instrumental, rationale for fair access. The cornerstone of this delibera-
tive case is provided by the Millian thesis that favorable deliberative con-
ditions require a diversity of messages. Such diversity might be encour-
aged in a variety of ways, but one natural means to diversity is to ensure
that all citizens have fair opportunities for expression, with the expecta-
tion that the breadth of messages will increase if the extent of expressive
opportunity is not determined by economic or social position.107
I have already indicated some ways to achieve fair access in a world of
unequal resources (supra, 106–108). One requirement is to endorse a
more “functional” conception of a public forum,108 rejecting the concep-
tion of such forums as places that are by tradition or explicit designation
open to communicative activity, and instead accepting a presumption
106. As, for example, in Rawls’s argument that reflection on the ideal of natural liberty leads
to the ideal of democratic equality. See Theory of Justice, 65–74; and my discussion of the boot-
strapping strategy in “Moral Pluralism and Political Consensus,” 278–279.
107. Reasoning of broadly this kind can be found in Metro Broadcasting, Inc. v. FCC, 100 S.
Ct. 2997 (1990), where the Court upholds an FCC program aimed at increasing broadcast di-
versity, by increasing the number of minorities with broadcast licenses. For criticisms, see
Charles Fried, “Metro Broadcasting, Inc. v. FCC: Two Concepts of Equality,” Harvard Law Re-
view 104 (1990): 107–127.
108. See Owen Fiss, “Silence on the Street Corner,” Suffolk University Law Review 26 1
(1992): 13–14.
137
that any location with dense public interaction ought to be treated as a

public forum that must be kept open to the public.109 Another condition
of fair access is a heightened presumption against content-neutral regula-
tions that have substantially disparate distributive implications—when, as
with regulations on the distribution of handbills, or on the use of parks
and sidewalks, they work to impose disproportionate burdens on those
who otherwise lack the resources to get their message out.
Furthermore, fair access recommends financing political campaigns
through public resources—at least to ensure reasonable floors—and reg-
ulating private political contributions and expenditures.110 In Buckley v.
Valeo, the Supreme Court drew a sharp distinction between regulations
of contributions—which are acceptable because they help to prevent the
appearance and reality of corruption—and regulations of expenditures,
which are an unacceptable burden of expressive liberty.111 In arguing
against expenditure limits, the Court appealed in part to the greater
burdensomeness of such regulations. More fundamentally, however, the
majority condemned restrictions (even if content-neutral) on expressive
liberty imposed in the name of “enhanc[ing] the relative voice of others”
and thereby “equaliz[ing] access to the political arena.”112 The Court did
not deny that expenditure limits would work to “equalize access,” but in-
stead held that regulations of expression aimed at such equalization were
“wholly foreign to the First Amendment.”113
Whatever their connection to the First Amendment, it is difficult to
understand how any plausible account of expressive liberty would regard
content-neutral regulations enacted in the name of fair access as foreign
to its concerns. In any case, I have suggested that requirements of fair ac-
109. Current tendencies in doctrine are, more or less, opposite to the suggestion here. See
U.S. v. Kokinda, 110 S. Ct. 3115 (1990); and the discussion in Fiss, “Silence on the Street Cor-
ner.”
110. For an argument—close to the perspective in this essay—that the current private
scheme of campaign financing violates requirements of equal protection, and a sketch of alter-
native directions of reform, see Jamin Raskin and John Bonifaz, “Equal Protection and the
Wealth Primary,” Yale Law and Policy Review 11 (Winter 1993): 273–332. For an instructive dis-
cussion of campaign finance that focuses more or less exclusively on the deliberative interest,
see Charles Beitz, Political Equality (Princeton: Princeton University Press, 1989), chap. 9.
111. 424 U.S. 1 (1976). It is consistent with Buckley to move to a system of voluntary public
financing, with matching funds for candidates whose opponents opt to spend private contribu-
tions, or to spend their own money. For a sketch of such a system, see Ellen S. Miller, “Money,
Politics, and Democracy,” Boston Review 18 (March/April 1993): 5–8.
112. 424 U.S. 1, 48–49 (1976).
113. Ibid.
138
cess share a common justification with other stringent protections of ex-

pressive liberty; rather than being “wholly foreign,” they are on a par.
Thus far I have focused on measures for ensuring fair access that are
content-neutral and concerned to remedy the effects of inequalities of
material resources on access to expressive opportunities. But as I indi-
cated in the discussion of fair access in Section 1, it is not clear that con-
tent-neutral regulations suffice when it comes to addressing problems
of fair access that do not reflect the distribution of material resources.
In the case of pornography, for example, the mechanisms of exclusion
have been tied directly to what is said. Consider the argument that por-
nography works by silencing women. Responding to the Brandeisian
“more speech” argument, MacKinnon explains the problem of silencing
and the consequent tension between content-neutrality and fair access
this way:
The situation in which women presently find ourselves with re-

spect to the pornography is one in which more pornography is
inconsistent with rectifying or even counterbalancing its dam-
age through speech, because so long as the pornography exists
in the way it does there will not be more speech by women. Por-
nography strips and devastates women of credibility, from our
accounts of sexual assault to our everyday reality of sexual sub-
ordination. We are stripped of authority and devalidated and
silenced. Silenced here means that the purposes of the First
Amendment, premised upon conditions presumed and pro-
moted by protecting free speech, do not pertain to women
because they are not our conditions. . . . Any system of freedom
of expression that does not address a problem where the free
speech of men silences the free speech of women, a real con-
flict between speech interests as well as between people, is not
serious about securing freedom of expression in this country.114
I agree with the last claim about the implications of a serious commit-
ment to freedom of expression, and later I will present a style of pornog-
raphy regulation that is less encompassing than MacKinnon’s proposals
but consistent with the perspective I have advanced in this essay. I do
wish, however, to resist jumping too quickly to the conclusion that con-
114. MacKinnon, “Francis Biddle’s Sister,” 193.
139
tent regulation is the only way to ensure fair access. Other measures of
empowerment that are more affirmative than regulations of expression
may show real promise in addressing silencing and exclusion—at least as
much promise as restricting pornography. In particular, alongside efforts
to address the general unjust inequalities of men and women—to over-
come the division of household labor and the labor-market segregation of
women115—alternative ways to meet the problems of silencing directly
should be explored. Earlier, for example, I mentioned a tort of domestic
sexual harassment, regular public hearings on sexual abuse, perhaps sub-
sidies for women’s organizations to hold such hearings, or easier access of
women to broadcast licenses.
Indeed, it is not clear that MacKinnon would disagree with the plausi-
bility of these remedies. Responding to the Brandeisian idea of address-
ing the harms of speech with more speech, she asks: “Would more
speech remedy the harm [of pornography]?” Her response is instruc-
tive: “In the end, the answer may be yes, but not under the abstract
system [my emphasis] of free speech, which only enhances the power of
pornographers while doing nothing to guarantee the free speech of
women, for which we need civil equality.”116 MacKinnon is right in say-
ing that a serious commitment to freedom of expression cannot be
sharply distinguished from a program of civil equality. For that reason,
the proposals I have mentioned are not exclusively about “the abstract
system of free speech”; they aim directly to enhance the speech of wo-
men and are part of a program of “civil equality.” So it is unclear why
they should not be expected to do as well as a more restrictive strategy for
addressing the harms at issue.
5. Hate Speech, Pornography, and Subcategorization

At several points in the discussion—for example, in my remarks on regu-
lating the pornographically obscene—I have suggested that a commit-
ment to stringent protections of expressive liberty is consistent with a cer-
tain style of restriction on expression. Other examples of the style—apart
from regulations of pornographic obscenity—are regulations of racist
fighting words or “sexually derogatory fighting words.”117 The idea of
115. For discussion, see Susan Moller Okin, Justice, Gender, and the Family (New York: Ba-
sic Books, 1989).
116. “Francis Biddle’s Sister,” 193.
117. I take the term from R. A. V. v. St. Paul, 112 S. Ct. 2538, 2546 (1992).
140
such regulations is to restrict expression within a less important class (ob-

scenity, fighting words) by targeting a particular subcategory (porno-
graphic, racist, sexually derogatory) of the broader class on grounds of the
special harmfulness of that subclass. For example, rather than targeting
fighting words generally, regulations focus on racially insulting fighting
words; rather than targeting obscenity generally, they focus on obscenity
that erotizes violence. Subcategorization is a distinctive and controversial
style of regulation because—to put the point abstractly—the defining fea-
tures of the subcategory would not provide a permissible basis for regula-
tion outside the less protected category. To be a little less abstract, the
strategy raises the following question: Why is it permissible to regulate
hateful fighting words or pornographic obscenity, while acknowledging
that a general regulation of hate speech or pornography would not be ac-
ceptable?
The acceptability of subcategorization will be important to my con-
cluding comments on the Stanford regulation. But it has recently been
the target of sharp criticism by Justice Scalia, writing for the Court in
R. A. V. v. St. Paul. While I am not concerned here with the constitu-
tional issue as such, Scalia’s objection raises important issues about regu-
lating expression that are not narrowly constitutional. I propose here to
address those issues.
ba ckground
The facts in R. A. V. v. St. Paul are straightforward and uncontested.
R. A. V. (Robert A. Viktora, a juvenile at the time of prosecution) and his
friends burned a cross in the yard of a black family; he was arrested, and
charged under a St. Paul Bias-Motivated Crime Ordinance. The ordi-
nance provides that:
Whoever places on public or private property a symbol, object,
appellation, characterization or graffiti, including, but not lim-
ited to, a burning cross or Nazi swastika, which one knows or
has reasonable grounds to know arouses anger, alarm or resent-
ment in others on the basis of race, color, creed, religion or
gender commits disorderly conduct and shall be guilty of a mis-
demeanor.118
118. Minnesota Legislative Code §292.02 (1990), cited in ibid., 2541.
141
R. A. V. challenged the ordinance, arguing that it was overbroad and

impermissibly content-based. The Minnesota Supreme Court rejected
the challenge. Central to the Court’s holding was its construction of
the phrase “arouses anger, alarm or resentment in others” as restricted
to “fighting words.” As defined in Chaplinksy v. New Hampshire, such
words are directed to individuals, form “no essential part of any exposi-
tion of ideas,” and their “very utterance inflicts injury” or “tends to incite
an immediate breach of the peace.”119 Premising that First Amendment
protection does not extend to fighting words,120 the Minnesota Court
held that the ordinance was neither overbroad nor an impermissible form
of content regulation.
The Supreme Court rejected the conclusions of the Minnesota Court
and agreed unanimously on the infirmity of the St. Paul ordinance. This
consensus, however, emerged from a convergence of two distinct lines of
argument about the sources of that infirmity. Writing for the Court, Jus-
tice Scalia maintained that the regulation, understood to be restricted to
fighting words, was an impermissible form of content discrimination; re-
jecting this contention, the separate concurrences by Justices White and
Stevens held that it was not really restricted to fighting words and so was
objectionable because overbroad.121 I am concerned here with the major-
ity’s claim: that, even as restricted to fighting words, the regulation is
impermissibly content-discriminatory. To state the problem more exactly:
assuming, as the majority does, that fighting words are a proscribable cat-
egory of expression, is it permissible to focus a regulation on the particu-
lar subcategory of fighting words mentioned in the ordinance? Let us call
the subcategory “hateful fighting words.” We have two competing propos-
als: (1) a regulation of hateful fighting words represents an impermissible
regulation of subject matter (and perhaps viewpoint), and (2) a regulation
of hateful fighting words represents a permissible targeting of a subcate-
gory of concededly low value and regulable expression on grounds of the
special injuriousness of that subcategory.
119. 315 U.S. 568, 571–572 (1942). For doubts about the fighting words doctrine, see Note,
“The Demise of the Chaplinsky Fighting Words Doctrine: An Argument for its Interment,”
Harvard Law Review 106 (1993): 1129–1146.
120. This is the basis of the reasoning by the Minnesota Court. Scalia’s opinion emphati-
cally rejects this claim. See R. A. V., 112 S. Ct. 2543.
121. The Minnesota Court said it was restricted to fighting words. But the concurrences re-
jected that Court’s construal of the fighting words test. Ibid., 2558–2560.
142
t hree p oint s of a gre e me nt

To locate the disagreement between these two proposals more precisely,
we need first to clarify three points of common ground:
Proscribable Expression Is Not without Protection. From the fact that the
government could proscribe a whole category of expression—say, child
pornography—it does not follow that every less inclusive regulation pro-
scribing a subclass and permitting the rest is also acceptable: think of a
child pornography statute restricted to kiddie porn in which at least one
actor wears an “I like Dan Quayle” button, or a regulation of obscenity
produced after supper. Regulations targeted on those subcategories are
unacceptable. So the argument for restricting hateful fighting words can-
not count among its premises the claim that every subcategory of a
proscribable category can be targeted permissibly.
Subcategories Can Sometimes Be Restricted on the Basis of Their Content.

Agreeing that fighting words (along with obscenity and defamation) are
proscribable because of their content, the majority accepts further that
regulations can target certain subcategories of proscribable expression in
virtue of the distinctive content of those subcategories. The federal gov-
ernment, for example, can “criminalize only those threats of violence
that are directed against the President.”122 So the argument against regu-
lating hateful fighting words cannot count among its premises the claim
that all content-based regulations of subcategories of fighting words are
impermissible.
Taking these first two points together: the disagreement is about the
specific subcategory singled out by the St. Paul ordinance. That disagree-
ment, in turn, is sharpened by a third point of agreement between the
two positions:
It Is Impermissible to Proscribe All Speech That Arouses Anger, Alarm, or

Resentment on the Basis of Race, Color, Creed, Religion, or Gender.123
Such a regulation would aim at and almost certainly produce an unac-
122. Ibid., 2546.

123. See ibid., 2558–2560 (White dissenting). This point is not very controversial, even
among defenders of hate-speech regulations. See, for example, Charles Lawrence, “If He Hol-
lers Let Him Go.”
143
ceptable “suppression of ideas.” The issue, then, is whether a regulation

targeted specifically at fighting words that “arouse anger, alarm. . . .” is
acceptable.
regul a t ing ha t ef ul f igh t i ng w ords

With these three points in place, we can fix the precise disagreement and
assess the alternative positions.
The first view is that a regulation of hateful fighting words will trigger
exactly the same suspicion about the suppression of ideas as a general
hate-speech regulation will, directed to all speech that arouses anger,
alarm, or resentment in others on the basis of race, color, creed, religion,
or gender. The underlying principle that bars a general regulation of hate
speech (the third point of agreement) is that hateful messages are not
proscribable because of their content. They do not forfeit that immunity
because they travel in a vehicle that is, for reasons other than the hate
message, dangerous. Thus, immediate provocative speech can be regu-
lated. But the fact that a hateful message is conveyed, for example, in an
immediately provocative way does not make it permissible to target it as
distinct from other messages conveyed in an equally (or more) provoca-
tive way.
Content regulation threatens the official suppression of ideas; so the
question is always whether the “official suppression of ideas is afoot.”124
And that question—according to the first view—loses none of its force
when a regulation is targeted on a proscribable category of speech; the
fact that expression falls into a less-protected category does not make it
permissible to use a regulation of such expression as a device to restrict
concededly protected messages.125
The alternative view is that there is indeed less concern about content
discrimination—less concern about the suppression of ideas—when reg-
ulated speech falls into a proscribable category. Why? The neatest answer
would be this: “How could there be any concern about the suppression
124. R. A. V., 112 S. Ct. 2547.
125. Even this misses the full subtlety of the majority view. They suggest that the Title VII
ban on hostile work environment sex discrimination “may” permissibly regulate “sexually de-
rogatory ‘fighting words,’ among other words” (ibid., 2546). But, it is permissible to regulate sex-
ually derogatory fighting words in the workplace only as an “incidental” effect of a general pro-
tection against hostile work environments. The fact that the regulation of speech was an
incidental part of a general code of conduct would immediately answer the concern that the
speech itself was being regulated because of its message rather than because of its harmful ef-
fects.
144
of ideas? Expression in proscribable categories conveys no ideas.” But

that will not do; different obscene movies, for example, can convey com-
peting ideas about the pleasures of different sorts of sex.126 More to the
point, if hateful fighting words did not communicate anything, there
would be no point targeting them. Nor would it be right simply to insist
that if a category is proscribable, then we should be less concerned about
protecting it. That is, of course, true in some way. But it does not indicate
any reason for reduced concern about content discrimination, and it
threatens to fly in the face of the first point of agreement noted earlier:
that proscribable expression has some protection. The explanation for
the reduced concern about content regulation cannot lie in the bare fact
that expression belongs to a proscribable category but must, instead, be
provided by the reason for treating it as proscribable in the first place.
Consider, then, the category of fighting words. Provocations directed
to individuals, such words comprise “no essential part of the exposition of
ideas.” For that reason, concerns about the official suppression of ideas
are naturally reduced when regulations are targeted on them; it is, intu-
itively, difficult to see how a regulation targeted on expression that is not
an essential part of the exposition of ideas could seriously threaten to
drive certain ideas, topics, or viewpoints from the marketplace of ideas or
the forum of political debate.127
More specifically, recall the reasons for being especially troubled about
regulations targeted on content (125–126 above): They represent espe-
cially serious threats to the deliberative and expressive interests; the rela-
tive precision of their targeting raises the specter of the abuse of power in
an especially acute way; and, even if such regulations are targeted on real
evils, the Fact of Reasonable Persuasion should lead us to trust more
speech to address those evils.
Because fighting words are insults or provocations directed to individu-
als, they do not make a significant contribution to discussion. The threat
to deliberative interests seems, then, relatively small. Moreover, insofar as
they serve as vehicles for expression—for advancing the expressive inter-
est—proscribing them leaves a wide range of alternative vehicles. And
126. Thus Scalia’s firm distinction between “no part” of the exposition of ideas and “no es-
sential part” of that exposition, ibid., 2544. Implicit in these remarks is the suggestion that the
concurrences endorse the tempting but implausible view that I note in the text.
127. There is certainly no “prohibition of public discussion of an entire topic.” Boos v. Barry,
485 U.S. 312, 319 (1988).
145
the reasons for expression in the form of fighting words do not seem espe-
cially substantial. Taking these points together, it seems much less plausi-
ble that a regulation targeted on hateful fighting words would severely
suppress ideas or would be motivated by a desire to suppress them than
that a regulation targeted on hate speech generally would have that unac-
ceptable effect or illegitimate motivation. So there does appear to be a
substantial difference in the fears about suppression that would reason-
ably be triggered by a general regulation of hate speech and a regulation
targeted specifically on the hateful subset of fighting words.
Of course, given the Facts of Power, Easy Offense, and Abuse, con-
cerns about suppression could be revived if a regulation were focused on
a relatively insignificant harm. But racial subordination, for example,128 is
a serious evil; and it is at least plausible that racist fighting words play
some role—plausibly a significant role—in maintaining racial inequality.
They contribute to an environment of fear, suspicion, hostility, and mis-
trust that makes racial division so resistant to remedy. So the regulation
picks out not an arbitrary class of fighting words but a class that is espe-
cially damaging to fundamental political values, for example the value of
racial equality. Finally, it seems especially implausible that the injuries
produced by hateful fighting words can be remedied with more speech.
The anger, the fear, and the suspicion that they produce are not kinds
that can be easily addressed by verbal reassurances.
The regulation, then, is targeted on a category with only a minimal
connection with the fundamental deliberative and expressive interests;
and within that category it focuses on a subcategory that is plausibly more
injurious than other elements of the category and whose effects are plau-
sibly more recalcitrant to expressive cure.
There are three responses to this argument, each of which aims to rein-
state suspicions about the suppression of ideas in a hateful fighting words
regulation.
The first is that there is a straightforward basis for the suspicion: it is
agreed, as I indicated earlier, that a general hate-speech regulation would
128. I say “for example” because the St. Paul ordinance was not simply addressed to racial
hate speech nor, more particularly, to racial hate speech targeted on African Americans or
other groups subordinated on the basis of race. For the suggestion that such a narrow and
“openly asymmetric regulation” might have been less constitutionally suspect—in light of the
Thirteenth Amendment ban on badges of servitude—see Akhil Reed Amar, “Comment: The
Case of the Missing Amendments: R. A. V. v. City of St. Paul,” Harvard Law Review 106 (1992):
155–161. See also my discussion of the asymmetry in the Stanford regulation, 152–153 below.
146
threaten the suppression of ideas. But if it is unacceptable to single out

hateful words for special regulation, why isn’t it also unacceptable to re-
strict the hateful subset of fighting words? Contrast that restriction with
one that singles out the fighting words that are especially likely to incite
breach of the peace. Fighting words are low value in part because they
tend to incite breach, so there could be no objection to singling out those
fighting words that threaten, in more extreme form, the very evil that
prompts the reduced protection in the first place.129 But a regulation of
hateful fighting words (arguably) does not pick out the especially provoc-
ative, so it is objectionable.
The problem with this objection is that it fails to take into consider-
ation the bases for reduced protection for fighting words and the reasons
for special concerns about content regulation. The category of fighting
words is such that the restriction of expression within the category does
not present a substantial threat of the suppression of ideas. But the sup-
pression of ideas is the main threat posed by content regulation. So the
explanation of the reduced protection for fighting words also explains
why a regulation of hateful fighting words does not threaten the suppres-
sion of ideas and so accounts for the legitimacy of a form of content dis-
crimination that would be unacceptable outside the limited context of
fighting words.
A second reason for the concern might be a familiar “camel’s nose”
concern: that once we allow the suppression of some subcategory of hate
speech, we will then be tempted to regulate hate speech generally. But
that regulation is not legitimate.
The problem here is that the argument proves too much: it provides a
case for an absolute ban on content regulation, a position that no one in
the debate occupies (see the second of the three points of agreement
noted above). Moreover, while the concern about excessive regulation is
real, the point of carving out such less protected categories as fighting
words, obscenity, and commercial speech is precisely to address that con-
cern. To endorse the strategy of categorization as a device against tempta-
tions to overregulate and then to revisit concerns about those temptations
in the context of regulating subcategories, strikes me as an exaggerated
129. See R. A. V., 112 S. Ct. 2545: “When the basis for the content discrimination consists en-
tirely of the very reason the entire class of speech is proscribable, no significant danger of idea
or viewpoint discrimination exists.”
147
form of distrust, and one that runs up against the premise—accepted by

the Court majority—that fighting words themselves are low value.
The third reason is that, while hateful fighting words are certainly of-
fensive and insulting, even gross offensiveness and insult cannot provide
a basis for regulation.130 But to say that the “price” is offensiveness, repre-
sents a tendentious misstatement of the harms. The harms of hateful
fighting words are several and include the role of such words in sustain-
ing racial division and preserving racial inequality.131 This is a very great
harm. Of course, not every restriction of expression that contributes to
avoiding it is, for that reason, acceptable. But a regulation that may con-
tribute, and do so without threatening the suppression of ideas (for exam-
ple, a regulation of hateful fighting words), is acceptable.
p ornogra p hic obsc e ni t y

In introducing this discussion of regulations of hateful fighting words, I
presented such regulations as one example of the more general strategy
of regulation by subcategorization. I have now indicated why the strategy
is, as a general matter, unobjectionable. To clarify the basis of this view, I
want now to say more about an example I discussed earlier—the case of
pornographic obscenity.
I will assume the Miller test for obscenity.132 According to that test, a
work is obscene just in case the average person, applying community
standards, finds that the work, taken as a whole, appeals to the prurient
interest; moreover, the work must present an offensive depiction of sex-
ual conduct; and, finally, it must lack serious literary, artistic, political,
or scientific value. The intuition is that sexually preoccupied, offensive
junk does not merit stringent constitutional protection. It is an interest-
ing question, which I will not pursue here, why the sexual preoccupation
makes a difference.133 It does not appear to diminish the value of the ex-
130. This appears to be the force of Scalia’s remark that “what makes the anger, fear, sense of
dishonor, etc. produced by violation of this ordinance distinct from the honor fear, sense of dis-
honor, etc. produced by other fighting words is nothing other than the fact that it is conveyed
by a distinctive idea, conveyed by a distinctive message.” Ibid., 2548.
131. Here I agree with Amar that a cleaner focus on the nature of the harms and a more dis-
criminating discussion of the differences among the categories mentioned in the ordinance—
“race, color, creed, religion, or gender”—would have sharpened both the regulation and the
Court’s assessment of it. See Amar, “The Case of the Missing Amendments,” 155–160.
132. Miller v. California, 413 U.S. 15 (1973).
133. See Roth v. United States, 354 U.S. 476, 512 (Douglas dissenting), and Harry Kalven,
“The Metaphysics of the Law of Obscenity,” The Supreme Court Review: 1960, ed. Philip B.
Kurland (Chicago: University of Chicago Press, 1960), 18–19.
148
pression, which by stipulation is not very great; furthermore, since the

costs lie in offensiveness, and expression can be offensive without being
sexually preoccupied, the sexual content is not required for the costs.
So why is it not permissible to regulate violence-preoccupied, offensive
junk? Or offensive junk preoccupied with frightening people? Or preoc-
cupied with money making? Or cruelty? Leaving these questions for an-
other occasion, I will assume for the sake of argument that obscenity
merits reduced protection. I want to ask about the implications of that as-
sumption for the regulation of subcategories of the obscene.
Consider, then, three obscenity regulations. The first targets all ob-
scene forms of expression. The second targets obscene expression in
which women are subjected to violence—what I referred to earlier as the
“pornographically obscene.” I stipulate a regulation covering all obscen-
ity in which women are subjected to violence, rather than obscenity
in which that violence is applauded, because I want the regulation to
be content-based but viewpoint neutral.134 The third regulation targets
“grossly” obscene expression, by which I mean expression that is obscene
and grossly offensive by the lights of the community—perhaps golden-
shower movies and movies featuring oral sex with animals fall into this
class.
Paralleling the earlier discussion of hateful fighting words, let us dis-
tinguish two natural responses to these regulations. The first is con-
structed on analogy with a view that accepts a regulation targeted on
extremely provocative fighting words but not one targeted on hateful
fighting words. So it would accept regulations of all the obscene or of
only the grossly obscene, but not of only the pornographically obscene.
Why would a regulation focused on the grossly offensive subcategory
be acceptable? Although the determination of gross offensiveness is a
matter of content (prohibitions on golden-shower movies and movies dis-
playing oral sex with animals are subject-matter restrictions), and content
regulations are generally objectionable, offensiveness is precisely the rea-
son for reducing the protection of the obscene in the first place. So if it is
permissible to target all offensive, prurient junk without engendering
concern about the suppression of ideas, then surely it is permissible to
target the grossly offensive, prurient junk without engendering such sus-
picion.
134. I am not sure that an obscene movie could present violence against or humiliation of
women in an unfavorable light, because by so doing, it would plausibly have serious political
value, thus defeating the categorization as obscene.
149
A regulation of the pornographically obscene is, like a regulation of the

grossly obscene, not subject matter neutral. But, this first line of response
emphasizes, it singles out for regulation, a subcategory of the obscene on
the basis of considerations other than those that render it obscene in the
first place; the subclass of the violent may not correspond to the subcate-
gory that is either especially prurient or grossly offensive in its prurience.
Moreover, a general regulation of pornography seems unacceptable, for
the reasons I indicated earlier. Because the feature that defines the sub-
category is unrelated to obscenity, and because that feature would trigger
concern about the suppression of ideas if applied outside the context of
obscenity, it might be thought to trigger that concern here. Defenders of
the regulation will, to be sure, argue that the regulation is justified by ref-
erence to the distinctive harms of the pornographically obscene. But if
those alleged harms cannot provide the rationale for a general regulation
of pornography, then why should they provide an acceptable rationale for
a regulation of the pornographically obscene?
Here again we meet the central concern: the fact that a whole category
of expression is proscribable does not imply a reduced concern about the
evil of a kind of content discrimination that would be unacceptable if ap-
plied to a wider category of expression.
Once more, however, an alternative view—constructed on analogy
with the position that approves a regulation of hateful fighting words—
seems more plausible. This alternative would accept the regulation of
the pornographically obscene.135 The contention fueling this second line
of argument is that if obscenity is low value in the first place, then it
is permissible to restrict pornographically obscene representations on
grounds that such representations are injurious (though not on grounds
of viewpoint), even though the alleged injuries would be insufficient to
sustain the regulation of pornography generally. The reason is this: a reg-
ulation of pornography generally does, for reasons I discussed earlier,
present a substantial threat to fundamental expressive and deliberative
interests. (This might be conceded even by those who argue that the
threat is overpowered by injuries reasonably attributed to pornography.)
But the basis for treating obscenity as low value is that it contributes little
to the fundamental interests. Because it does, a regulation of it would not
135. Indeed, I suspect that many who hold this second view would be more inclined to regu-
late the pornographically obscene than the grossly obscene. But I will put this matter to the
side.
150
present a substantial threat to those interests. Because it would not, the

concerns that provide the basis for opposing content discrimination are
diminished. Because they are diminished, the injuries associated with
the fusing of sex and violence by the pornographically obscene provide
sufficient basis for regulation. Indeed there is a better case for this regula-
tion—which is focused on genuine harms—than either for the regu-
lation of obscenity generally or for the regulation of grossly offensive
obscenity. For those regulations aim to prevent the uncertain evil of of-
fensiveness rather than the genuine evil of injuries to women.
6. Reflections on the Stanford Case

Finally, I come back to the Stanford regulation. At the beginning of this
essay, I promised to fit a pallid endorsement of it into the conception of
freedom of expression I have outlined here. Everything I have said thus
far should suffice to explain the lack of enthusiasm (though I will add a
few more considerations). What are the bases for the endorsement?
Recall that the regulation restricts “speech or other expression” that: (1)
is intended to insult or stigmatize individuals on the basis of their sex,
race, color, handicap, religion, sexual orientation, or national and ethnic
origin; and (2) is addressed directly to the individual or individuals whom
it insults or stigmatizes; and (3) makes use of insulting or “fighting words”
or nonverbal symbols that are “commonly understood to convey direct
and visceral hatred or contempt for human beings on the basis of their
sex, race, etc.”
My endorsement reflects three features of the regulation, each of
which indicates sensitivity to the case for stringent protection that I have
presented here.
(1) The regulation is directed to remarks that are intended to insult,

and the insult must be directed to an individual or small group. So
the regulated expression bears at most a loose connection to the
fundamental expressive and deliberative interests.
(2) The insult must be conveyed through fighting words and in partic-
ular, words that insult or stigmatize on the basis of sex, race, color,
and the like. Because of the requirement of immediate provoca-
tion and injury associated with fighting words, some of the costs
are direct, and there is no deflecting them with “more speech.”
(3) The rule singles out an exceptional category and does not repre-
151
sent an open-ended invitation to balancing the benefits and costs

of expression. So it is attentive to concerns about vulnerability.
Given the minimal interests, direct costs, and attention to potential

abuse, the supports for protection are substantially reduced, and it seems
appropriate (or at least permissible) to shift the burden of restraint to the
speaker.
To be sure, the regulation is not without its troubles, principally be-
cause (as interpreted) it would be viewpoint-discriminatory; for example,
racist remarks addressed to black students may (depending on condi-
tions) count as a form of discriminatory harassment; racist remarks to
white students do not.136 Is the general presumption against such discrim-
ination rebuttable in this case?
In assessing the troublesomeness of the viewpoint discrimination, we
need to keep in focus the requirements of intent and fighting words, and
the stipulation that the words be directed to an individual or group with
the intent to insult or stigmatize. Expression meeting these conditions
has only a marginal claim to protection in the first place. So—as I indi-
cated in the discussion of hateful fighting words—it seems permissible to
deny protection to a subcategory of it in order to promote the substantial
value of ensuring equality of educational opportunity for the groups sin-
gled out in the regulation.
Of course “permissible” does not imply “recommended.” Other con-
siderations are relevant to deciding that issue. How much injurious ex-
pression would actually be avoided? Would the regulation be at all effec-
tive in combating the underlying problems reflected in hate speech?
Furthermore, apart from addressing these questions about the regulation
itself, we need to consider the wisdom of focusing energy and attention
on regulating hate speech (or pornography) rather than on taking more
affirmative measures to combat the harms that the regulation aims to
avoid. The focus of regulating expression has at least three defects: it can
serve as a distraction of energy from other measures; it divides people
who are allied in their commitment to equality; and it suggests a depress-
ingly profound loss of constructive, egalitarian, political, and social imag-
ination.
136. According to a clarification offered in a debate in the Stanford faculty senate. See the
discussion in Nadine Strossen, “Regulating Racist Speech,” Duke Law Journal (1990): 494 n.
110; and the less measured discussion in Charles Fried, “The New First Amendment Jurispru-
dence: A Threat to Liberty,” University of Chicago Law Review 59 (1992): 22ff.
152
Together these considerations strike me as good grounds for skepti-

cism. To be sure, such skepticism is not costly for those of us who are not
now targets of hate speech.137 This point has some force, and so I do not
treat my skepticism about effectiveness as a basis for rejecting the regula-
tions as inconsistent with a commitment to stringent protections of ex-
pressive liberty.
But ineffectiveness may in turn lead to pressure for more stringent reg-
ulations in the name of equality. And this could represent a serious chal-
lenge to the conception of freedom of expression I have sketched here.
For if the harms of subordination cannot be fought with more speech and
other nonrestrictive remedies, then—the world being as it is—a commit-
ment to substantive equality simply cannot be reconciled with a strong
affirmation of expressive liberties. If my account of the basis of freedom
of expression is correct, then that conclusion will not show that we ought
to give up on the value of equality; for as I indicated early on, nothing in
the defense turns on a freestanding preference for liberty over all compet-
ing values, and in particular a freestanding preference for liberty over
equality. Nor would it show that we ought to give up on the value of lib-
erty. Instead, we would face a grim standoff between concerns about ex-
pressive liberty and concerns about equality.
So those of us who celebrate the values of equality, toleration, and ex-
pressive liberty—and the remedial powers of speech in reconciling these
values—ought to conduct our celebration by getting to work.
137. I say “not now” because I was frequently called “kike,” “bagel-bender,” and the like
when I was growing up.
153
5
P R O C E D U R E A N D S U B S TA N C E
I N D E L I B E R AT I V E D E M O C R A C Y
T he fundamental idea of democratic legitimacy is that the authoriza-

tion to exercise state power must arise from the collective decisions
of the members of a society who are governed by that power.1 More pre-
cisely—and stated with attention to democracy’s institutional charac-
ter—it arises from the discussions and decisions of members, as made
within and expressed through social and political institutions designed to
acknowledge their collective authority. That is an abstract statement of
the concept of democracy, and deliberately so. Democracy comes in
many forms, and more determinate conceptions of it depend on an ac-
count of membership in the people and, correspondingly, what it takes
for a decision to be collective, to be made by citizens “as a body.”
Take a political community in which adherence to a comprehensive
moral or religious doctrine,2 perhaps rooted in national tradition, is a
I would like to thank John Rawls, Charles Sabel, T. M. Scanlon, Cass Sunstein, and Iris
Marion Young for illuminating comments on earlier drafts of this essay. The “Deliberative De-
mocracy” section draws on my “Deliberation and Democratic Legitimacy,” in Alan Hamlin
and Phillip Petit, eds. The Good Polity (Oxford Blackwell, 1989), 17–34 [reprinted as essay 1 in
this volume]. The “Three Principles” section draws on my review of Robert Dahl’s Democracy
and Its Critics (New Haven: Yale University Press, 1989) in Journal of Politics 53, 1 (1991): 221–
225; and on my “Pluralism and Proceduralism,” Chicago-Kent Law Review 69, 3 (1994): 589–
618. The “Realizing Democracy” section draws on Joshua Cohen and Joel Rogers, Democracy
and Association (London: Verso, 1995).
1. “Governed by” rather than “affected by.” Democracy is about justifying authority not
about justifying influence. See Michael Walzer, Spheres of Justice (New York: Basic Books,
1983); and Christopher McMahon, Authority and Democracy (Princeton: Princeton University
Press, 1994). Alternatively stated, authorization must come from the popular will, where “pop-
ular will” is understood as indicating the ultimate authority and responsibility of citizens as a
body not as implying a collective ranking of alternatives that preexists institutions and seeks au-
thentic expressions through them. See William Riker, Liberalism Against Populism (San Fran-
cisco: W. H. Freeman, 1992).
2. On the notion of a comprehensive doctrine, see John Rawls, Political Liberalism (New
York: Columbia University Press, 1996 [1993]), 13.
154
P R O C E D U R E A N D S U B S T A N C E
condition of full membership. Authorization, then, will require congru-

ence with that view, and only decisions exhibiting such congruence can
properly be deemed “collective.” For that reason, the test for democratic
legitimacy will be, in part, substantive—dependent on the content of
outcomes, not simply on the processes through which they are reached.
What happens, though, when the idea of collective authorization is set
against a different background: where there is no shared comprehensive
moral or religious view, members are understood as free and equal, and
the national project, such as it is, embraces a commitment to expressing
that freedom and equality in the design of institutions and collective
choices?3 Does this shift in background drive us to an entirely procedural
view of democracy and collective decision? I think not. But before ex-
plaining why, I want to say something about the interest of the question
and the terms in which it is stated.
My question about the effects of a shift in background is prompted by
the aim of formulating a conception of democracy suited to the kind of
human difference captured in the “fact of reasonable pluralism”4—the
fact that there are distinct, incompatible understandings of value, each
one reasonable, to which people are drawn under favorable conditions
for the exercise of their practical reason. The good-faith exercise of practi-
cal reason, by people who are reasonable in being concerned to live with
others on terms that those others can accept, does not lead to conver-
gence on one particular philosophy of life.
The claim about reasonable pluralism is suggested by persistent
disagreements about, for example, the values of choice and self-
determination, happiness and welfare, and self-actualization; disputes
3. American national identity is commonly tied to such a conception, as in Lincoln’s claim

that the nation was conceived in liberty and dedicated to the proposition that all men are cre-
ated equal. Some regard this abstract national self-definition as exceptionally American. Con-
sidering the conflictual conditions under which modern nationalism evolved, I doubt that this
claim can be sustained without substantial qualification. Claims about the content of national
identity—like all claims about identity—are endlessly contested: they are as much moves in so-
cial and political conflicts aimed at establishing the authority of a particular nationalist under-
standing as they are intellectual discoveries. For every person who will claim that the concep-
tion of people as free and equal is foreign to their particular national identity, we can always
find someone who shares the national self-definition and will deny that foreignness.
4. For discussion of this fact, see my “Moral Pluralism and Political Consensus,” in The Idea
of Democracy, ed. David Copp, Jean Hampton, and John Roemer (Cambridge: Cambridge
University Press, 1993), 270–291 [reprinted as essay 2 in this volume]; John Rawls, Political Lib-
eralism; and my “A More Democratic Liberalism,” Michigan Law Review 92, 6 (May 1994):
1502–1546.
155
about the relative merits of contemplative and practical lives and the im-
portance of personal and political engagement; and disagreements about
the religious and philosophical backgrounds of these evaluative views.
Apart from the sheer fact of disagreement, there is, moreover, no appar-
ent tendency to convergence generated by the exercise of practical rea-
son; furthermore, we have no theory of the operations of practical reason
that would lead us to predict convergence on comprehensive moralities,
nor can I think of any marginally attractive social or political mecha-
nisms that might generate such agreement.
This fact of reasonable pluralism gives shape to the conception of citi-
zens as free and equal that constitutes part of the conception of democ-
racy I want to explore here. To say that citizens are free is to say, inter alia,
that no comprehensive moral or religious view provides a defining condi-
tion of membership or the foundation of the authorization to exercise po-
litical power. To say that they are equal is to say that each is recognized as
having the capacities required for participating in discussion aimed at au-
thorizing the exercise of power.
What, then, are the implications of reasonable pluralism for a concep-
tion of democracy? It is natural to suppose that by excluding a compre-
hensive consensus on values, the fact of reasonable pluralism leads to a
procedural conception of democracy. According to such a conception,
the democratic pedigree that lies at the source of legitimacy can be set-
tled by looking exclusively to the processes through which collective de-
cisions are made and to the values associated with fair processes: for ex-
ample, values of openness, equal chances to present alternatives, and full
and impartial consideration of those alternatives. The fact of reasonable
pluralism appears to require a procedural conception because it deprives
us of a background of shared moral or religious premises that could give
determinate content to the idea of popular authorization or constrain the
substance of genuinely collective choices. Without that background, we
are left, it may seem, with no basis for agreement on anything more than
fair procedures—and perhaps not even that.
I think this conclusion is not right and will sketch a view that combines
an assumption of reasonable pluralism with a more substantive concep-
tion of democracy. Moreover, I will argue that this combination is a natu-
ral result of a particular way of thinking about democracy—a “delibera-
tive” understanding of the collective decisions that constitute democratic
governance. Before discussing the deliberative conception, though, I
156
need first to fix the concerns about procedure and substance more pre-
cisely, distinguish a deliberative from an aggregative conception of de-
mocracy, and show how aggregative conceptions lead to proceduralism.
Liberties, Ancient and Modern

Consider a familiar dilemma associated with the idea of tracing legiti-
macy to popular authorization.5 On the one hand, democracy may seem
too much a matter of procedure to provide a basis for an account of legiti-
macy; some democratic collective choices are too execrable to be legiti-
mate, however attractive the procedures that generate them. On the
other hand, the idea of democracy appears to exclude any competing ba-
sis of legitimacy. Democracy appears to be the form of collective choice
mandated by the fundamental idea that citizens are to be treated as
equals. So democracy is commonly thought to be the way we must de-
cide how other political values are to be ordered; it is not simply one po-
litical value that is to be combined with others.
This dilemma is familiar from discussions of democracy and the “lib-
erties of the moderns”—religious liberty, liberty of conscience more gen-
erally, liberty of thought and expression, rights of person and personal
property. Lacking any evident connection to conditions of democratic
procedure, such liberties are commonly understood as constraints on
democratic process. Not so with political liberties. A constitution dis-
abling government from restricting political participation or regulating
the content of political speech can be interpreted as safeguarding, rather
than constraining, democratic process. Assurances of such political liber-
ties help to preserve the connection between popular authorization and
political outcome—to preserve the continuing authority of the people,
and not simply the majority of them.6 These liberties—the liberties of the
ancients—are constitutive elements of democratic process.
Things are different when it comes to abridgments of religious liberty
or restrictions on expression whose content can be construed as political
only on a uselessly capacious construal of “political.” In these cases, dis-
abling provisions in a constitution appear simply to limit democracy, not
to be among its preconditions, either implicit or explicit.
5. By “tracing legitimacy to popular authorization,” I mean treating such authorization as a

sufficient condition for the exercise of political power.
6. See John Hart Ely, Democracy and Distrust (Cambridge, MA: Harvard University Press,
1980); and Robert Dahl, Democracy and Its Critics.
157
The liberties of the moderns appear, then, to be founded on values en-

tirely independent from the values of democracy. And that appearance
may prompt one of two undesirable conclusions. The first is that the po-
litical liberties are merely instrumental, of value just insofar as they pro-
tect the liberties of the moderns; when they fail to ensure such protec-
tion, an authority external to the people ought to do so. Here, a conflict
between democracy and other political values is easily translated into a
conflict between democratic and nondemocratic procedures of political
decision making.7
A second view holds that the liberties of the moderns have no standing
deeper than contingent popular consensus. Although abridgments of
nonpolitical liberties that emerge from a fair democratic process may be
unjust, then, they face no problems of democratic legitimacy.8
We are pushed into this dilemma by a particular understanding of de-
mocracy, which I will call “aggregative”—as distinct from deliberative.9
According to an aggregative conception, democracy institutionalizes a
principle requiring equal consideration for the interests of each member;
or, more precisely, equal consideration along with a “presumption of per-
sonal autonomy”—the understanding that adult members are the best
judges and most vigilant defenders of their own interests.10 To criticize
processes as undemocratic, then, is to claim that those processes failed to
give equal consideration to the interests of each member. The natural
method for giving such consideration is to establish a scheme of collec-
tive choice—majority or plurality rule, or group bargaining—that gives
equal weight to the interests of citizens, in part by enabling them to pres-
ent and advance their interests. And that requires a framework of rights of
participation, association, and expression.
7. See Dahl’s concerns about judicial review in Democracy and Its Critics, 183.
8. It is, of course, open to a democratic pluralist to hold that such infringements are unjust
and that the people ought to reject them.
9. On the distinction between aggregative and deliberative views, and its bearing on the pos-
sibility of reconciling commitments to values of liberty and equality within a conception of de-
mocracy, see my review of Dahl’s Democracy and Its Critics. For discussion of the related dis-
tinction between strategic and deliberative conceptions, see David Estlund, “Who’s Afraid of
Deliberative Democracy? On the Strategic/Deliberative Dichotomy in Recent Constitutional
Jurisprudence,” Texas Law Review 7, 7 (June 1993): 1437–1477. Estlund identifies strategic theo-
ries with views that make use of the idea of utility-maximization. I think that the crucial issue is
whether a conception of democracy emphasizes the idea of providing reasons acceptable to
others.
10. In Democracy and Its Critics, chaps. 6–8, Dahl derives conditions on democratic proce-
dure from a principle of equal consideration and a presumption of personal autonomy.
158
Arguably, the aggregative view can be extended beyond such straight-

forwardly procedural rights to some concerns about outcomes. For it
might be said that collective choices that depend on discriminatory
views—on hostility or stereotyping—do not give equal weight to the in-
terests of each who is governed by them. And when we face outcomes
that disadvantage people who are the likely targets of such views, we have
strong evidence of a failure of the process to give equal consideration to
the interests of each.11
This procedural reinterpretation of important political values can,
however, go only so far. Religious liberty, for example, has no apparent
procedural basis. To be sure, abridgments of freedom of worship are
sometimes troubling because they result from discriminatory (anti-Cath-
olic, anti-Semitic) attitudes. When they do, protections of religious liber-
ties will emerge from the requirement of equal consideration. But the
failure to give appropriate weight to religious convictions need not reflect
hatred, discrimination, or stereotyping—nor must it depend on any other
of the conventional ways of demeaning a person or failing to treat her as
an equal. The problem may have a different source: it may trace to a fail-
ure to take seriously the stringency or the weight of the demands placed
on the person by her reasonable moral or religious convictions—not the
intensity with which she holds those convictions, which does figure in
aggregative views, but the stringency or the weight of the demands im-
posed by the convictions, given their content.12 It is precisely this strin-
gency that compels reasons of especially great magnitude for overriding
those demands. But such considerations about the relative stringency of
demands are absent from the aggregative conception; so, therefore, is the
11. When, for example, legislation relies on racial classifications—or at least on malign ra-
cial classification—we have reason to suspect that discriminatory preferences prompted the
legislation. And if they did, then the procedural-democratic pedigree of the regulation is argu-
ably corrupt. See Ely, Democracy and Distrust, chap. 6; and Ronald Dworkin, Law’s Empire
(Cambridge, MA: Harvard University Press, 1986), chap. 10. For a less social-psychological
view of unacceptable procedural pedigree, see Bruce Ackerman, “Beyond Carolene Products,”
Harvard Law Review 98 (1985): 713–746. Unfortunately, the Supreme Court has endorsed the
view that “malign racial classification” is a pleonasm, and “benign racial classification” a con-
tradiction in terms. See Richmond v. Croson, 488 U.S. 469 (1989); Shaw v. Reno, 113 S. Ct.
2816 (1993); and Miller v. Johnson, slip op. (1995). For an alternative view, see Metro Broadcast-
ing v. FCC, 497 U.S. 547 (1990).
12. The distinction between rights required to prevent discrimination and rights required to
protect fundamental interests plays a central role in equal protection doctrine. See Laurence
Tribe, American Constitutional Law (Mineola, NY: Foundation Press, 1988), chap. 16. On the
importance of paying attention to the content of views in an account of free exercise, see Ron-
ald Dworkin, Life’s Dominion (New York: Knopf, 1993), 162–166.
159
need to find reasons of great weight before overriding those demands.

That is a fundamental deficiency, and it lies at the source of the dilemma
I sketched earlier.
A deliberative conception of democracy does not face the same trou-
bles about reconciling democracy with nonpolitical liberties and other
substantive, nonprocedural requirements. While accepting the fact of
reasonable pluralism, it is attentive to the stringency of demands to
which agents are subject, and therefore does not present its conception of
democracy or collective decision in an exclusively procedural way. To
make this case, I will first sketch the main ideas of a deliberative view;
then I will show how, on the deliberative conception, we can accommo-
date the fact of reasonable pluralism without endorsing a wholly pro-
cedural conception of democracy. In particular, I will show how the lib-
erties of the moderns and other substantive conditions are themselves
elements in an institutional ideal of deliberative democracy.
Deliberative Democracy
The deliberative conception of democracy is organized around an ideal
of political justification. According to this ideal, justification of the exer-
cise of collective political power is to proceed on the basis of a free public
reasoning among equals. A deliberative democracy institutionalizes this
ideal. Not simply a form of politics, democracy, on the deliberative view,
is a framework of social and institutional conditions that facilitates free
discussion among equal citizens—by providing favorable conditions for
participation, association, and expression—and ties the authorization to
exercise public power (and the exercise itself) to such discussion—by es-
tablishing a framework ensuring the responsiveness and accountability of
political power to it through regular competitive elections, conditions of
publicity, legislative oversight, and so on.13
I will come back later to the conditions for institutionalizing delibera-
tion in greater detail. First, though, I want to say more about the idea of
deliberative justification itself.
A deliberative conception puts public reasoning at the center of politi-
cal justification. I say “public reasoning” rather than “public discussion”
because a deliberative view cannot be distinguished simply by its empha-
sis on discussion rather than on bargaining or voting. Any view of democ-
13. On the role of the idea of democracy as more than a political idea, see Gordon Wood,
The Radicalism of the American Revolution (New York: Knopf, 1992), esp. 232.
160
racy—indeed any view of intelligent political decision making—will see

discussion as important, if only because of its essential role in pooling in-
formation against a background of asymmetries in its distribution. Nor is
it marked by the assumption that political discussion aims to change the
preferences of other citizens. Though a deliberative view must assume
that citizens are prepared to be moved by reasons that may conflict with
their antecedent preferences and interests, and that being so moved may
change those antecedent preferences and interests,14 it does not suppose
that political deliberation takes as its goal the alteration of preferences.
Nor is it distinguished by its endorsement of an epistemic conception of
voting, according to which votes are interpreted as expressions of beliefs
about the correct answer to a political question, rather than as prefer-
ences about what policy is to be implemented.15
The conception of justification that provides the core of the ideal of
deliberative democracy can be captured in an ideal procedure of politi-
cal deliberation. In such a procedure participants regard one another as
equals; they aim to defend and criticize institutions and programs in
terms of considerations that others have reason to accept, given the fact
of reasonable pluralism and the assumption that those others are reason-
able; and they are prepared to cooperate in accordance with the results of
such discussion, treating those results as authoritative.
Which considerations count as reasons? A suitable answer will take
the form not of a generic account of reasons but of a statement of which
considerations count in favor of proposals in a deliberative setting suited
to free association among equals, where that setting is assumed to in-
clude an acknowledgment of reasonable pluralism. This background is
reflected in the kinds of reasons that will be acceptable. In an idealized
deliberative setting, it will not do simply to advance reasons that one
takes to be true or compelling: such considerations may be rejected by
others who are themselves reasonable. One must instead find reasons
that are compelling to others, acknowledging those others as equals,
aware that they have alternative reasonable commitments and knowing
something about the kinds of commitments that they are likely to have—
for example, that they may have moral or religious commitments that im-
14. See Cohen, “Deliberation and Democratic Legitimacy,” 24.

15. On the idea of an epistemic conception, see Jules Coleman and John Ferejohn, “De-
mocracy and Social Choice,” Ethics 97, 1 (October 1986): 6–25; and Joshua Cohen, “An
Epistemic Conception of Democracy,” Ethics 97, 1 (October 1986): 26–38.
161
pose what they take to be overriding obligations. If a consideration does

not meet these tests, that will suffice for rejecting it as a reason. If it does,
then it counts as an acceptable political reason.
To be sure, the precise characterization of the acceptable reasons, and
of their appropriate weight, will vary across views. For that reason, even
an ideal deliberative procedure will not, in general, produce consensus.
But even if there is disagreement, and the decision is made by majority
rule, participants may appeal to considerations that are quite generally
recognized as having considerable weight, and as a suitable basis for col-
lective choice, even among people who disagree about the right result:
when participants confine their argument to such reasons, majority sup-
port itself will commonly count as reason for accepting the decision as le-
gitimate.
To underscore this point about the importance of background context
in the account of acceptable political reasons, I want to highlight a differ-
ence between the idea of reasonable acceptance at work here and the
idea of reasonable rejection in Scanlon’s contractualism.16 Scanlon char-
acterizes the wrongness of conduct in terms of the idea of a rule that “no
one could reasonably reject,” and he advances this characterization as
part of a general account of the subject matter of morality and the nature
of moral motivation. So his account of reasonableness—of reasonable
grounds for rejecting principles—is required to work quite generally,
even in settings with no ongoing cooperation, institutional ties, or back-
ground of equal standing as citizens.
My concern is not with reasons generally, or morality generally, or with
political deliberation generally, or with the reasons that are suited to
democratic discussion quite generally, but with a view about the implica-
tions of democracy given a specific background. And that background
constrains what can count as an acceptable reason within a process of de-
liberation. For if one accepts the democratic process, agreeing that adults
are, more or less without exception, to have access to it, then one cannot
accept as a reason within that process that some are worth less than oth-
ers or that the interests of one group are to count for less than those of
16. T. M. Scanlon, “Contractualism and Utilitarianism,” in Amartya Sen and Bernard Wil-
liams, eds., Utilitarianism and Beyond (Cambridge: Cambridge University Press, 1982). The
point of contrast in the text is prompted by Scanlon’s discussion of the role of maximin reason-
ing in moral contractualism in What We Owe to Each Other (Cambridge, MA: Harvard Uni-
versity Press, 1997), 223–229.
162
others. And these constraints on reasons will limit the substantive out-
comes of the process; they supplement the limits set by the generic idea
of a fair procedure of reason giving.
I am not here raising an objection to Scanlon’s view. He has a different
topic—morality generally, as distinct from democratic legitimacy. In-
stead, I am urging that this difference in background makes a difference
to the kinds of reasons that are suited to the two cases.
To conclude these general remarks about the deliberative view, I want
to emphasize that its virtues are allied closely with its conception of bind-
ing collective choice, in particular with the role in that conception of the
idea of reasons acceptable to others who are governed by those choices
and who, themselves, have reasonable views. By requiring reasons ac-
ceptable to others, the deliberative view suggests an especially compel-
ling picture of the possible relations among people within a democratic
order.
To see the character of those relations, notice first that the deliberative
conception offers a more forceful rendering than the aggregative view of
the fundamental democratic idea—the idea that decisions about the ex-
ercise of state power are collective. It requires that we offer consider-
ations that others (whose conduct will be governed by the decisions) can
accept, not simply that we count their interests in deciding what to do,
while keeping our fingers crossed that those interests are outweighed.
Thus the idea of popular authorization is reflected not only in the pro-
cesses of decision making but also in the form—and, we will see later,
the content—of political reason itself.
This point about the force of the deliberative view and its conception
of collective decisions can be stated in terms of the idea of political com-
munity. If political community depends on sharing a comprehensive
moral or religious view, or a substantive national identity defined in
terms of such a view, then reasonable pluralism ruins the possibility of
political community. But an alternative conception of political commu-
nity connects the deliberative view to the value of community. In particu-
lar, by requiring justification on terms acceptable to others, deliberative
democracy provides for a form of political autonomy: that all who are
governed by collective decisions—who are expected to govern their own
conduct by those decisions—must find the bases of those decisions ac-
ceptable. And in this assurance of political autonomy, deliberative de-
mocracy achieves one important element of the ideal of community.
163
Not because collective decisions crystallize a shared ethical outlook that

informs all social life, nor because the collective good takes precedence
over the liberties of members, but because the requirement of providing
acceptable reasons for the exercise of political power to those who are
governed by it—a requirement absent from the aggregative view—ex-
presses the equal membership of all in the sovereign body responsible for
authorizing the exercise of that power.
To explain the deliberative ideal more fully, I want now to explore
some of its implications: the conditions that need to be met by social and
political arrangements that, within the setting of a modern state, institu-
tionalize deliberative justification. What conditions will such arrange-
ments need to satisfy, if they are to sustain the claim that they establish
the conditions for free reasoning among equals, and they root the autho-
rization to exercise state power in those conditions?
As a partial answer, I will indicate why deliberative democracy needs to
ensure the liberties of the moderns. Then I will connect the deliberative
view to conceptions of the common good and political equality.
Three Principles
The aggregative conception of democracy promises the protections re-
quired for a fair process of binding collective choice, including pro-
tections against discrimination that would undermine the claim of the
process to ensure equal consideration. I said earlier that the deliberative
view will provide a basis for wider guarantees of basic liberties. It is time
to make good on that claim. The main idea is that the deliberative con-
ception requires more than that the interests of others be given equal
consideration; it demands, too, that we find politically acceptable rea-
sons—reasons that are acceptable to others, given a background of differ-
ences of conscientious conviction. I will call this requirement the “prin-
ciple of deliberative inclusion.”
Consider, for example, the case of religious liberty. Religious views
set demands of an especially high order—perhaps transcendent obli-
gations—on their adherents; moreover, if we see these requirements
from the believer’s point of view, then we cannot think of them as self-
imposed. Instead, the requirements are fixed by the content of the con-
victions, which the agent takes to be true. Reasonable adherents, then,
cannot accept as sufficient, reasons in support of a law or a system of pol-
icy, considerations that would preclude their compliance with those de-
164
mands. What, then, about people who do not share those views? (I will
describe the issue from the point of view of citizens who have fundamen-
tal moral convictions but no religious convictions. Broadly parallel re-
marks could be made from the standpoint of citizens with different reli-
gious convictions.) They might regard all religious views that impose
such stringent demands, whatever their content and foundation, as un-
reasonable. I see no rationale for this view. Or they might treat the reli-
gious demands as intense preferences, to be given equal consideration
along with other preferences of equal intensity. This reductive response
indicates an unwillingness to see the special role of religious convictions
from the point of view of the person who has them, an unwillingness to
see how the religious view, in virtue of its content, states or implies that
the requirements provide especially compelling reasons.
Alternatively, they might take seriously that the demands impose what
the adherent reasonably regards as fundamental obligations, accept the
requirement of finding reasons that might override these obligations, and
acknowledge that such reasons cannot normally be found. The result is
religious liberty, understood to include freedom of conscience and wor-
ship. It emerges as the product of the demanding character of religious
requirements—which are seen, from the point of view of those who are
subject to them, as matters of fundamental obligation—together with the
requirement of finding reasons that those who are subject to those re-
quirements can reasonably be expected to acknowledge, and the fact that
citizens who are not religious have fundamental convictions that they
take to impose especially compelling obligations.
Suppose, then, that we prevent others from fulfilling such demands for
reasons that they are compelled—by the lights of a view that commands
their conviction—to regard as insufficient. This is to deny them standing
as equal citizens—full membership in the people whose collective ac-
tions authorize the exercise of power. And that, according to the delibera-
tive conception, is a failure of democracy. We have failed to provide a
justification for the exercise of power by reference to considerations that
all who are subject to that power, and prepared to cooperate on reason-
able terms, can accept. There are many ways to exclude individuals and
groups from the people, and this surely is one.
These points about religious liberty—essentially about its free exer-
cise—do not say anything about how to handle claims for religious ex-
emption from general obligations with a strong secular justification (in-
165
cluding obligations to educate children), or about whether special

provision is to be made for specifically religious convictions, as distinct
from conscientious ethical convictions with no religious roots.17 My aim
here is not to resolve or even address these issues: any view that recog-
nizes rights of free exercise will need to face those hard questions. My
aim is to show only that a deliberative conception of democracy is not
barred—by its structure—from acknowledging a fundamental role for
rights of religious liberty; indeed, it must provide a place for such rights.18
Finally, I emphasize that the point of guarantees of religious liberty,
which fall under the requirement of deliberative inclusion, is not nar-
rowly political: it is not to enable people to participate in politics—or to
participate without fear—nor is the aim to improve public discussion by
adding more diverse voices to it.19 The idea instead is that abridgment of
such liberties would constitute denials to citizens of standing as equal
members of sovereign people, by imposing in ways that deny the force of
reasons that are, by the lights of their own views, compelling. The reasons
for abridgment are unacceptably exclusionary because they are unsuited
to the ideal of guiding the exercise of power by a process of reason giving
suited to a system of free and equal citizens.
The principle of deliberative inclusion extends naturally from religious
liberty to a wide guarantee of expressive liberty.20 In this respect, it con-
17. On this last point: the key to the case for religious liberty is that the content of a view as-
signs stringent obligations to a person who holds it. But specifically religious content is not es-
sential.
18. This account of religious liberty may seem to rest on the idea of a natural right to reli-
gious liberty: to say, in effect, that reasons will count as acceptable in a deliberative process
only if they accept this right. If the idea of a natural right to religious liberty simply comes to
the claim that there is a right that can be abridged only on pain of illegitimacy, then the delib-
erative view includes natural rights. But natural rights views have claimed more than this: they
offer an explanation of the basis of fundamental rights in human nature, in natural law, or in a
prepolitical normative order to which political society must conform. The idea of democratic
legitimacy does not depend on that explanation, though it asserts nothing inconsistent with it.
It suffices that religious liberties have an explanation tied to the idea of democratic legitimacy.
For the purpose of political argument, nothing more needs to be said, positively or negatively.
19. Roberto Unger argues that a system of immunity rights is one component of a demo-
cratic order, because “freedom as participation presupposes freedom as immunity.” Rejecting
the view of “critics of traditional democratic theory,” who hold that “participatory opportunities
[are] a more than satisfactory substitute for immunity guarantees,” Unger sees immunity rights
as necessary if a citizen is to have the “safety that encourages him to participate actively and in-
dependently in collective decision-making.” In False Necessity (Cambridge: Cambridge Uni-
versity Press, 1987), 525. I agree with Unger’s observations, but I think that a conception of de-
mocracy can make a less instrumental place for certain liberties, even when those liberties are
not procedural.
20. This discussion draws on my “Freedom of Expression,” Philosophy and Public Affairs 22,
3 (Summer 1993): 207–263 [reprinted as essay 4 in this collection].
166
trasts with a more familiar strand of free speech theory that traces the
foundations of stringent guarantees of expressive liberty to the need to as-
sure a democratic framework of collective choice but guarantees strin-
gent protection only for political speech.21 This limit is in tension with
the requirement of deliberative inclusion.
Confining stringent protection to political speech seems natural, once
one has decided to base rights to free expression on the importance of re-
quiring government accountability and responsiveness to citizens as a
body. But as my remarks on the religion case suggest, a deliberative con-
ception of democracy cannot accept such a limit. To be sure, the idea of
discussion aimed at reaching reasonable agreement is fundamental to
the deliberative view. But it does not follow that the protection of expres-
sion is to be confined to speech that contributes to such discussion.
Consider expression that is not part of any process of discussion or per-
suasion—that is not “intended and received as a contribution to public
deliberation about some issue”22—but that, nevertheless, reflects what a
citizen takes, for quite understandable reasons, to be compelling reasons
for expression.23 This might be so in cases of bearing witness, with no ex-
pectation or intention of persuading others, or giving professional advice,
with no expectation or intention of shaping broader processes of collec-
tive decision making. The deliberative view extends stringent protection
to such expression, as a way to acknowledge the weight of those reasons.
Given the background of reasonable pluralism, the failure to do so—to
give due weight to an expressive interest that does not serve as input to
political discussion—will constitute a denial of equal standing, and de-
cisions that fail to ensure those stringent protections are not suitably col-
lective.
The tradition that traces protections of expressive liberty to democratic
ideals and then restricts stringent protection to contributions to debate in
the public forum conflates the general strategy of providing a case for
freedom of expression rooted in the idea of democracy with one ele-
21. See Alexander Meiklejohn, Free Speech and Its Relation to Self-Government (New York:
Harper and Row, 1948); and Cass R. Sunstein, Democracy and the Problem of Free Speech
(New York: Free Press, 1993). Also, Robert Bork, “Neutral Principles and Some First Amend-
ment Problems,” Indiana Law Journal 47, 1 (Fall 1971): 1–35; Ely, Democracy and Distrust; and
Owen Fiss, “Why the State?” Harvard Law Review 100 (1987): 781–794.
22. This is Sunstein’s account of political speech, in Democracy and the Problem of Free
Speech, 130.
23. I do not mean to suggest that stringent protection ought to be confined to expression ani-
mated by such compelling reasons. The conventional democratic defense of rights of expres-
sion also provides a basis for stringent protection. My aim is to supplement that rationale.
167
ment of that strategy: the need to protect inputs to a process of discussion.

But as with religious liberty, so, too, with expressive liberty: the delibera-
tive view also ties protections to acceptable outcomes of a deliberative
process, outcomes, that is, that can be justified given the requirement
on finding reasons acceptable to others under conditions of reasonable
pluralism.
Earlier I suggested a connection between the deliberative conception
and the value of community. That suggestion may now seem strained in
light of the connections between the requirement of acceptable reasons
and the protection of nonpolitical liberties. For such liberties are com-
monly represented as—for better or worse—the solvent of community.
But the deliberative view suggests a need for caution about that repre-
sentation. Given conditions of reasonable pluralism, the protection of
the liberties of the moderns is not a solvent of community. Reasonable
pluralism itself may be such a solvent: at least if we define community in
terms of a shared comprehensive moral or religious view. But once we as-
sume reasonable pluralism, the protection of the liberties of the moderns
turns out to be a necessary, though insufficient, condition for the only
plausible form of political community. As the phrase “principle of in-
clusion” indicates, those liberties express the equal standing of citizens as
members of the collective body whose authorization is required for the
legitimate exercise of public power.
Turning now to the common good: aggregative views of democracy are
conventionally skeptical about conceptions of the common good. Robert
Dahl, for example, has suggested that in pluralistic societies, conceptions
of the common good are too indeterminate to provide guidance, determi-
nate but unacceptable because they lead us to “appalling results” in con-
ditions that “are by no means improbable,”24 or determinate and accept-
able because purely procedural—because they define the common good
as a democratic process.25 On the deliberative conception, this skeptical
outlook is unwarranted, yet it is another reflection of the absence of con-
straints beyond the requirement of fair aggregation.
A deliberative account of the principle of the common good begins by
observing that citizens have good reason to reject a system of public
policy that fails to advance their interests at all. (I say a “system of policy”
because I do not wish to exclude the possibility that particular laws, regu-
24. Democracy and Its Critics, 283.
25. Ibid., 306–308.
168
lations, or policies that are not attentive to the interests of some citi-
zens may be justifiable as part of an overall package of laws and policies
that is.)26 This minimal constraint—of advancing the interests of each—
comes out of the generic conception of a deliberative process and suf-
fices to establish a pareto-efficiency requirement, as one element of a
conception of democracy.
But as I have emphasized, the deliberation that plays a role in the con-
ception of deliberative democracy is not simply a matter of reason giving,
generically understood. The background conception of citizens as equals
sets limits on permissible reasons that can figure within the deliberative
process. For suppose one accepts the democratic process of binding col-
lective choice, agreeing that adults are, more or less without exception, to
have access to it. One can then reject, as a reason within that process,
that some are worth less than others or that the interests of one group are
to count for less than the interests of other groups. That constraint on rea-
sons will, in turn, limit the outcomes of the process, adding to the condi-
tions set by the generic idea of deliberation. In particular, it provides a
case for a public understanding about the distribution of resources that
severs the fates of citizens from the differences of social position, natural
endowment, and good fortune that distinguish citizens.
John Rawls’s difference principle provides one illustration of such an
understanding.27 Treating equality as a baseline, it requires that inequali-
ties established or sanctioned by state action must work to the maximal
advantage of the least advantaged. That baseline is a natural expression
of the constraints on reasons that emerge from the background equal
standing of citizens: it will not count as a reason for a system of policy
that that system benefits the members of a particular group singled out by
social class or native talent or any other feature that distinguishes among
equal citizens. I do not wish to suggest here that Rawls’s difference princi-
ple is the uniquely acceptable conception of the common good. But
there is an especially strong case for it, both because it accepts the pre-
26. The vices of a sales tax, for example, depend on the nature and level of exemptions, the
presence (or not) of tax credits, and the nature of the policies that the revenue pays for.
27. See John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971),
513. For discussion of the connections between the difference principle and an ideal of democ-
racy, see my “Democratic Equality,” Ethics 99, 4 (July 1989): 736–743. Another view that might
be used to illustrate the points in the text is Dworkin’s equality of resources. See Ronald
Dworkin, “What Is Equality? Part 2: Equality of Resources,” Philosophy and Public Affairs 10, 4
(Autumn 1981): 283–345.
169
sumption of equality that emerges from the special constraints on reasons

within the deliberative democratic view and because it insists, roughly
speaking, that no one be left less well off than anyone needs to be—
which is itself a natural expression of the deliberative conception.
I want finally to connect the deliberative view with the rights of partici-
pation—the liberties of the ancients. More particularly, I want to show
how the deliberative view accommodates a “principle of participation.”28
According to that principle, democratic collective choice—institutional-
izing the tie between deliberative justification and the exercise of public
power—must ensure equal rights of participation, including rights of
voting, association, and political expression, with a strong presumption
against restrictions on the content or viewpoint of expression; rights to
hold office; a strong presumption in favor of equally weighted votes; and
a more general requirement of equal opportunities for effective influ-
ence.29 This last requirement condemns inequalities in opportunities for
office holding and political influence that result from the design of ar-
rangements of collective decision making.30
Notice first that the mere fact that decisions are to be made in a generi-
cally deliberative way does not go very far toward establishing a case for
the principle of participation.31 Perhaps an ideal deliberative procedure is
best institutionalized by ensuring well-conducted political debate among
28. See Rawls, A Theory of Justice, 36–37.

29. On the requirement of opportunities for effective influence, see Rawls, Political Liberal-
ism, 327–330. For a discussion of the constitutional dimension of the problem, see Davis v.
Bandemer, 478 U.S. 109, 132 (1986). The Court here acknowledges equal protection troubles
when the “electoral system is arranged in a manner that will consistently degrade a voter’s or
group of voters’ influence on political process as a whole.” Low-Beer distinguishes a require-
ment of equally weighted votes, at stake in apportionment issues, from equally meaningful
votes, at stake in gerrymandering cases. The value threatened by gerrymandering is better un-
derstood, I believe, as political influence more generally, not simply as voting strength. See
John Low-Beer, “The Constitutional Imperative of Proportional Representation,” Yale Law
Journal 94, 1 (November 1984): 163–188.
30. Among the concerns that fall under this requirement are vote dilution due to racial and
political gerrymandering and unequal influence due to campaign finance arrangements, re-
strictive rules on ballot access, and regulations of political parties.
31. Historically, the deliberative conception of politics was associated with highly exclusivist
forms of parliamentarism. Moreover, according to one influential line of thought, mass de-
mocracy destroyed the possibility of deliberative political decision-making. According to Carl
Schmitt, “The belief in parliamentarism, in government by discussion, belongs to the intellec-
tual world of liberalism. It does not belong to democracy.” Moreover, “the development of
modern mass democracy has made argumentative public discussion an empty formality.” See
The Crisis of Parliamentary Democracy, trans. Ellen Kennedy (Cambridge, MA: MIT Press,
1985), 6, 8.
170
elites, thus enabling people to make informed choices among them and
the views they represent, without any special provision for more substan-
tive political equality, understood as requiring equally weighted votes and
equal opportunities for effective influence.32 How, then, does the deliber-
ative view connect to concerns about participation and political equality?
Three considerations are important.
First, given the principles of deliberative inclusion and of the common
good, the deliberative view can avail itself of conventional instrumental
reasons in support of equal political rights. Such rights provide the means
for protecting other basic rights and for advancing interests in ways that
might plausibly promote the common good. Moreover, absent assur-
ances of effective influence, such promotion seems an unlikely result.
And it would be especially unlikely if inequalities in effectiveness corre-
sponded to underlying social or economic inequalities in the society.33
In making this instrumental case, I may appear to be shifting to a bar-
gaining conception of politics, with assurances of equal power working to
ensure a political equilibrium with fair outcomes. But that gets the in-
strumental rationale and the mechanism wrong. The idea instead is that
ensuring that all citizens have effective political rights serves as a re-
minder that citizens are to be treated as equals in political deliberation
and, by reducing inequalities of power, reduces the incentives to shift
from deliberative politics to a politics of bargaining.
A second consideration is that many of the conventional, historical jus-
tifications for exclusions from or inequalities of political rights—justifica-
tions based on race and gender, for example—will not provide accept-
able reasons in public deliberation. This consideration will not exclude
all reasons for inequality—for example, if votes are of unequal weight be-
cause the political system relies, as in the case of the U.S. Senate, on a
32. Thus Beitz’s account of political equality connects the interests in recognition and equi-
table treatment with assurances of equally weighted votes and fair access. What he calls the
“deliberative interest,” by contrast, simply requires well-conducted political debate. See Politi-
cal Equality (Princeton: Princeton University Press, 1989).
33. See the discussion of the interest in equitable treatment in Beitz, Political Equality, 110–
114. This interest plays an important role in the apportionment cases decided by the Supreme
Court in the early 1960s. “No right is more precious in a free country than that of having a
voice in the election of those who make the laws under which, as good citizens, we must live.
Other rights, even the most basic, are illusory if the right to vote is undermined.” Wesberry v.
Sanders, 376 U.S. 1 (1964), cited in Reynolds v. Sims, 377 U.S. 533, 558 (1964). Or again: “Espe-
cially since the right to exercise the franchise in a free and unimpaired manner is preservative
of other basic civil and political rights, any alleged infringement of the right of citizens to vote
must be carefully and meticulously scrutinized.” Reynolds v. Sims, 562.
171
scheme of territorial representation, in which districts correspond to po-

litical subdivisions. But it establishes a further presumption in favor of
the principle of participation.
Finally considerations analogous to those we met with in the case of re-
ligion and expression strengthen the case for equal political rights, with
assurances of equal opportunities for effective influence. A characteristic
feature of moral and religious convictions is that they give us strong rea-
sons for seeking to shape our political-social environment. The compre-
hensive views underlying those reasons range from Aristotelian views
about the central role of civic engagement in a good life, to Rousseauean
claims about the connection between personal autonomy and participa-
tion, to views, founded on religious convictions, about the commanding
personal responsibility to ensure social justice and the corresponding per-
sonal sin of failing in that responsibility. It is common ground, however,
that citizens have substantial, sometimes compelling, reasons for address-
ing public affairs. Because they do, the failure to acknowledge the weight
of those reasons for the agent and to acknowledge the claims to opportu-
nities for effective influence that emerge from them reflects a failure to
endorse the background idea of citizens as equals.
Realizing Democracy
The deliberative conception of democracy captures the role of “undemo-
cratic” as a term of criticism applying to results as well as processes: it
provides common roots for the “by the people” and “for the people” as-
pects of the ideal of democracy. But this incorporation of important sub-
stantive requirements into the conception of democracy gives rise to a
problem of its own. The concern is that if we offer an interpretation of
democracy that treats all good things as ingredient in the idea of democ-
racy—requirements of political equality, considerations of common
good, and liberties of the moderns—then we may appear to integrate pro-
cedural and substantive values at the cost of practical guidance. What are
we to do when the many elements of deliberative democracy come into
conflict? Common foundations in deliberative democracy do not pro-
vide any insurance against conflict in practice. For example, the liberties
mandated by the requirement of deliberative inclusion may conflict with
the equal political liberties that fall under the requirement of participa-
tion. Why does it help to have all these elements ingredient within the
ideal of democracy, given conflicts among them?
172
The answer is that by underscoring common foundations, we highlight

the need to find ways to accommodate the different requirements, so far
as accommodation is possible. That may be more often than we are in-
clined to think, though how often is a function of politics. To make this
point less telegraphic, I will sketch some examples. I want to focus the
discussion on two cases in which the various requirements arguably con-
flict and see what might be said about their reconciliation in these cases.
My first case is campaign finance. The central problem arises from
a familiar dilemma: on the one hand, restrictions on political expendi-
tures by candidates, individual citizens, and organizations appear to bur-
den expressive liberty, particularly given a background expectation that
such expenditures are permissible; arguably, burdens also result from
very stringent limits on contributions to political campaigns. Moreover,
restrictions on candidate and party expenditures, even when they are
accepted as a condition for receiving public financing, may reinforce in-
cumbency advantages, resulting in a less competitive electoral system,
less capable of holding elected officials accountable and so of ensur-
ing public authorization of the exercise of power.34 On the other hand,
a regime of unrestricted expenditures is a regime in which political in-
fluence—chances to hold office and to affect the outcomes of political
contests—reflect economic position, and that means inequalities in op-
portunities for effective influence.35
Thus the familiar conflict about restrictions on political spending.
Some reject restrictions, even if they are content-neutral and motivated
by a sincere desire to ensure greater equality of political influence. In an
infamous sentence in the majority opinion in Buckley v. Valeo, the Su-
preme Court said that “the concept that government may restrict the
34. This may seem puzzling. Making the safe assumption that incumbents have advantages
in raising funds, it might seem clear that challengers would fare better under a system of spend-
ing restrictions. But, according to one influential line of argument, background incumbency
advantages make challengers more dependent on money. Thus a challenger is better off run-
ning with $300,000 against an incumbent with $500,000 than running with $250,000 against
an incumbent with $250,000. See Gary Jacobson, “Enough Is Too Much: Money and Compe-
tition in House Elections,” in Kay Lehman Schlozman, ed., Elections in America (Boston: Al-
len and Unwin, 1987), 173–195. For criticisms of Jacobson’s view, see Donald Philip Green and
Jonathan S. Krasno, “Salvation for the Spendthrift Incumbent: Reestimating the Effects of
Campaign Spending in House Elections,” American Journal of Political Science 32, 4 (Novem-
ber 1988): 884–907.
35. I say a “regime” of unrestricted expenditures because the choice among systems of
financing is a choice among alternative schemes of permissions and restrictions, not a choice
between regulation and nonregulation.
173
speech of some elements of our society in order to enhance the relative

voice of others is wholly foreign to the First Amendment”;36 as a result,
they were unwilling to find any basis beyond concerns about quid pro
quo corruption for regulating political spending.37 Others, concerned to
insist on the importance of fair political equality, argue that limits are es-
sential.
The first idea—that it is impermissible to restrict the voice of some in
order to enhance the relative voice of others—seems bizarre. My earlier
account of the bases of rights of expression and political participation
suggested a common foundation for both; so there is no basis for the sub-
ordinate role of political equality. Moreover, once we have accepted a
presumption in favor of equally weighted votes—one person/one vote—
we are already committed to precisely such restrictions and enhance-
ments.38
Still, focusing on the permissibility of restrictions may be putting the
emphasis in the wrong place. Given the bases of rights of expression in
the principles of participation and deliberative inclusion, it would be de-
sirable to promote equality of opportunity for effective influence through
less restrictive means than expenditure limits, should such means be
available.39 And the natural route to such reconciliation is to establish
a scheme of public financing. The idea of such a system is to rely prin-
cipally on “floors” rather than “ceilings”—subsidies rather than limits—
to remedy violations of the principle of participation.40 By establishing
36. 424 U.S. 1 (1976), 48–49.
37. Buckley, 26–27.
38. See Gray v. Sanders, 372 U.S. 368 (1963); Wesberry v. Sanders; and Reynolds v. Sims.
The tension between the apportionment decisions and Buckley is noted in Rawls, Political Lib-
eralism, 361; and David A. Strauss, “Corruption, Equality, and Campaign Finance Reform,”
Columbia Law Review 94, 4 (May 1994): 1382–1383. The Court itself has retreated from the
Buckley position, acknowledging possibilities of corruption involving unfair influence without
quid pro quo and the permissibility of regulating expenditures—at least in the case of for-profit
corporations—in order to avoid such corruption. See Austin v. Michigan Chamber of Com-
merce, 494 U.S. 652, 660 (1990).
39. A problem with relying principally on spending restrictions is the capacity of contribu-
tors and candidates to maneuver around restrictions. See Frank Sorauf, Inside Campaign Fi-
nance: Myths and Realities (New Haven: Yale University Press, 1992). Increase the level of pub-
lic subsidy, and you reduce the incentives to such maneuvering.
40. The United States is one of four OECD countries with contributions limits. All the
other political systems rely more substantially than the United States does on public financing;
the Scandinavian countries have no contribution or expenditure limits and rely entirely on
public funding. See Ellen S. Miller and Joel Rogers, The World of Campaign Finance (Madi-
son and Washington, D.C.: Center for a New Democracy and Center for Responsive Politics,
1992).
174
floors, a suitable scheme of public financing helps to make office holding

more widely available; by reducing dependence of parties and candidates
on private resources, it assures greater equality of opportunity for influ-
ence.41 The effectiveness of floors in providing such assurance may de-
pend on making the availability of support conditional on accepting
spending limits. But limits of this kind may be unnecessary, given a re-
gime with substantial public financing.
Of course a wide range of public financing schemes are possible: sup-
port can be provided to candidates or to parties42 or to individual voters
(as citizen vouchers)43 or, in the case of initiatives and referenda, to
nonparty organizations; funds can be made available for electoral activ-
ity or for more general party support; and support can be provided in
the form of free media access. And in deciding among such schemes, it is
important to consider their effects on deliberation as well as opportuni-
ties for effective influence. Citizen vouchers are especially promising, I
think. But I do not propose to go into such details here. The point is to
state the main principles, emphasize the importance of finding some ac-
commodation of them in view of their common basis in the value of de-
mocracy, and indicate that the strategy of accommodation is, roughly
stated, a strategy of empowerment, not of restriction.
My second case concerns possible tensions between a deliberative pol-
itics and the principles of participation and the common good—and the
role of a strategy of “associative democracy” in blunting those tensions.44
41. For a description of a scheme of public financing animated by concerns about equality
and deliberation, see Jamin Raskin and John Bonifaz, “The Constitutional Imperative and
Practical Superiority of Democratically Financed Elections,” Columbia Law Review 94, 4
(May 1994): 1160–1203.
42. For an interesting public financing proposal, built around support for parties that would
be distributed by congressional leadership, see Daniel Hays Lowenstein, “The Root of Evil Is
Deeply Rooted,” Hofstra Law Review 18, 2 (Fall 1989): 351–355.
43. On voucher systems, see Bruce Ackerman, “Crediting the Voters: A New Beginning for
Campaign Finance,” American Prospect (Spring 1993); and Edward Foley, “Equal Dollars Per
Voters: A Constitutional Principle of Campaign Finance,” Columbia Law Review 94, 4 (May
1994): 1204–1257.
44. A broadly parallel concern arises in connection with the role of race-conscious measures
in drawing lines around electoral districts. Given a background of racial bloc voting, the princi-
ple of participation may suggest a need for race-conscious districting to ensure opportunities
for effective influence. But race-conscious districting arguably works against deliberative poli-
tics. According to Lani Guinier, cumulative voting would address this tension. Like other forms
of proportional representation, cumulative voting combines increased chances of effective mi-
nority influences with voluntary constituencies that may encourage deliberation. See her “Sec-
ond Proms and Second Primaries: The Limits of Majority Rule,” Boston Review 17, 5 (Septem-
ber-October 1992): 32–34; and The Tyranny of the Majority (New York: Basic Books, 1994).
175
The problem here is less straightforward, as is the proposed solution. So I

first need to set some background.45
Begin, then, with two familiar premises. First, any well-functioning
democratic order satisfying the principles of participation and the com-
mon good requires a social base. Beyond the world of voters and parties,
secondary associations—organized groups intermediate between market
and state—are needed both to represent otherwise underrepresented in-
terests (as in the case of trade unions or other independent worker organi-
zations) and to add to public competence in advancing the common
good (think of the role played by unions and employer associations in es-
tablishing standards on worker training in any well-functioning training
system). Representing underrepresented interests helps to ensure politi-
cal equality; adding to public competence helps to promote the common
good.
Second, the right kinds of association do not naturally arise, for the
purposes of either addressing problems of underrepresentation or per-
forming more functional tasks: there is, for example, no natural tendency
for an emergence of secondary associations to correct for inequalities of
political opportunity due to underlying economic inequalities or to en-
sure the regulatory competence needed to advance the common good.
Now put together the need for a favorable associative environment
with the fact that such an environment is not naturally provided. This
conjunction suggests a strategy for addressing the associative deficit: a
strategy of associative democracy that would use public powers to encour-
age the development of the right kind of secondary association. For ex-
ample, where manifest inequalities in political representation exist, the
associative strategy recommends promoting the organized representation
of presently excluded interests. Where associations have greater compe-
tence than public authorities have for advancing the common good, it
would recommend encouraging a more direct and formal governance
role for groups. So trade unions and employer associations that took on
responsibility for the joint development of training curricula, for exam-
ple, might be encouraged by public grants contingent on their assump-
tion of such responsibilities.
But here we arrive at the tension. In seeking to meet the principles of
45. This section of the essay draws on Joshua Cohen and Joel Rogers, “Solidarity, Democ-
racy, Association,” in Wolfgang Streeck, ed., Staat und Verbaende, special issue of Politischen
Vierteljahresschrift (Wiesbaden: Westdeutscher Verlag, 1994), 136–159.
176
participation and the common good by fostering governance roles for

groups, we may heighten the role of group affiliation in defining political
identity. And that may encourage a factionalized politics of group bar-
gaining—albeit under more fair conditions—rather than a more deliber-
ative politics.46
Standard responses to this problem are to encourage greater insulation
of the state from groups or to give up on egalitarian political values be-
cause no agent has the capacity to advance them. The idea of associative
democracy suggests a different line of response. It begins by rejecting the
implicit assumption that solidarities formed outside formal political are-
nas must be narrowly focused on particular groups, and it proposes some
institutional invention guided by that rejection. To explain the bases for
rejecting that assumption and the relevant kinds of invention, I will make
some very sketchy remarks about the idea of a deliberate use of associa-
tions in regulation.
Generally speaking, the idea of a regulatory role for associations re-
flects a sense of the limited capacity of the state to regulate for the com-
mon good. Those limits appear in four kinds of cases:
1. Where government has the competence to set specific regulatory
terms, but the objects of regulation are sufficiently numerous, dispersed,
or diverse to preclude serious government monitoring of compliance.
Many workplace regulations—on appropriate wages and hours, compen-
sation, and especially the appropriate organization of work, pertaining for
example to occupational health and safety—provide instances of this
monitoring problem.
2. Where government has the competence to set general standards of
performance, but the objects of regulation are sufficiently diverse or
unstable to preclude government specification of the most appropriate
means of achieving them at particular regulated sites. Much environ-
mental regulation is of this kind.
3. Where government may (or may not) be able to enforce standards
once set but cannot set appropriate ends itself.47 Often, an appropriate
46. This concern emerges naturally from criticisms of modern pluralism. See, for example,
Theodore Lowi, The End of Liberalism: The Second Republic of the United States, 2d ed. (New
York: Norton, 1979). For discussion of associative democracy as a response to the problem of
faction, see Joshua Cohen and Joel Rogers, “Secondary Associations in Democratic Gover-
nance,” Politics and Society 20, 4 (December 1992): 393–472.
47. Or it can set them only in very abstract terms, for example, as requirements of “reason-
ableness” or “due care.”
177
standard can be determined only by those with local knowledge not

readily available to government, or it can be specified only as the out-
come or in the context of prolonged cooperation among nongovernment
actors. Industry standards on product or process uniformity and perfor-
mance are often of this kind, as are standards on training. The appropri-
ate norm shifts constantly; the content of the norm derives from coopera-
tion in the process of establishing it.48
4. Where problems are substantially the product of multiple causes
and are connected with other problems, crossing conventional policy
domains and processes. In such cases, the appropriate strategy requires
coordination across those domains as well as cooperation from private ac-
tors within them. Urban poverty, local economic development, and ef-
fective social service delivery are among the familiar problems in this
class. None can be solved without cooperation across quite different insti-
tutions and groups—lending institutions, health care providers, technol-
ogy diffusers, education and training establishments, housing authorities,
community development corporations, neighborhood associations—op-
erating wholly or substantially outside the state itself. These and other
parties involved in the problem and its proposed solution, however, typi-
cally have distinct, if not competing, agendas and different identities and
interests.
To address such problems, the associative approach recommends ex-
plicitly relying on the distinctive capacity of associations to gather local
information, monitor compliance, and promote cooperation among pri-
vate actors. When problems are more or less functionally specific—corre-
sponding roughly to the first three classes of cases described earlier—as-
sociative governance is not uncommon. As a general matter, it is best
developed in the areas of workplace regulation and training, and it relies
on institutions controlled by the traditional “social partners” of labor and
capital. The use of plant committees to enforce occupational safety and
health regulations, for example, or groupings of trade unions and em-
ployers to facilitate technology diffusion, or employer and union associa-
tions to set standards on training, is familiar. The lessons of practice in
these areas might be more explicitly generalized to include nontradi-
tional parties.
48. For discussion of the problem of shifting standards as it applies to the more general prob-
lem of measures of business performance, see Charles Sabel, “A Measure of Federalism: As-
sessing Manufacturing Technology Centers,” in Research Policy 25, 2 (March 1996): 281–307.
178
As the scope of associative efforts moves beyond functionally specific

problems to issues that are decidedly more sprawling and open-ended—
as in the urban poverty or regional economic development examples—
models are less clear. Here the associative strategy recommends the
construction of new arenas for public deliberation that lie outside con-
ventional political arenas49 and that aim to establish the desired coordina-
tion.
Notice, however, that both the inclusion of nontraditional stakeholders
and the development of deliberative arenas suggest a new possibility: that
of constructing new bases of social solidarity through a process of defin-
ing and addressing common concerns. It is one thing for a well-funded
union to be asked to participate in the design of training standards of ob-
vious concern to it as well as to the rest of society. It is quite another for a
nascent or an underfunded community environmental organization to
gain significant resources (and thus greater organizational life) if it assists
in designing an environmental early warning system that is expected to
take notice of emerging environmental problems before they become
unmanageable. In this case, support is tied to public service. Or for a
neighborhood association and an economic development corporation in
a poor community to receive assistance conditional on their jointly orga-
nizing a training program for parents and a child care program for train-
ees as part of a broader job-training effort: once more, participation and
support are tied to a project of public advantage.
The solidarities characteristic of such efforts will be the bonds of peo-
ple with concerns—say, a concern to address persistent urban poverty—
who treat one another as equal partners in addressing those shared con-
cerns.50 In short, these efforts—which could have very wide scope—have
the potential to create new “deliberative arenas” outside formal politics
that might work as “schools of deliberative democracy” in a special way.
Deliberative arenas established for such coordination bring together peo-
ple with shared concrete concerns, very different identities, and consider-
able uncertainty about how to address their common aims. Successful
cooperation within them, fostered by the antecedent common concerns
of participants, should encourage a willingness to treat others with re-
49. Though to the extent that they receive public support, they are to be subject to constitu-
tional constraints, in particular to guarantees of equal protection.
50. This claim depends, of course, on the background assumption of a democratic state pro-
tecting basic liberties and ensuring equal protection.
179
spect as equals, precisely because discussion in these arenas requires fash-

ioning arguments acceptable to those others. Assuming fair conditions of
discussion and an expectation that the results of deliberation would regu-
late subsequent action, the participants would tend to be more other-re-
garding in their outlook. The structure of discussion, aimed at solving
problems rather than pressuring the state for solutions, would encourage
people to find terms to which others could agree. And that would plausi-
bly drive argument and proposed action in directions that would respect
and advance more general interests. Moreover, pursuing discussion in
the context of enduring differences among participants would incline
parties to be more reflective in their definition of problems and proposed
strategies for solution and would tend to free discussion from the precon-
ceptions that commonly limit the consideration of options within more
narrowly defined groups.
If this is right, then a social world in which solidarities are formed in
part by reference to such arenas is different from a social world whose
associational life is narrower and factionalized. And that means that it
may be possible to use the associative strategy to advance the principles
of participation and the common good without thereby encouraging
particularistic group identities that turn politics from deliberation to bar-
gaining.
Conclusion
The fact of reasonable pluralism does not, I have argued, mandate a pro-
cedural account of democracy and collective choice. Conjoined with a
deliberative conception of justification, it is compatible with a substan-
tive account of democracy whose substance—captured in principles of
deliberative inclusion, the common good, and participation—includes
values of equality and liberty. Moreover, such a deliberative conception
offers an attractive rendering of the idea of collective choice, tying that
idea to a view of political community. Finally, we are not without re-
sources for addressing possible tensions between and among the values
of liberty, equality, and community built into the deliberative concep-
tion. But whether those resources are exploited is, of course, a matter of
politics.
180
6
D I R E C T LY D E L I B E R AT I V E P O LYA R C H Y
with Charles Sabel
1. Introduction
In this essay we defend a form of democracy that we will call “directly de-
liberative polyarchy.” We argue that it is an attractive kind of radical,
participatory democracy with problem-solving capacities useful under
current conditions and unavailable to representative systems. In directly
deliberative polyarchy, collective decisions are made through public de-
liberation in arenas open to citizens who use public services or who are
otherwise regulated by public decisions. But in deciding, those citizens
must examine their own choices in the light of the relevant delibera-
tions and experiences of others facing similar problems in comparable
jurisdictions or subdivisions of government. Ideally, then, directly delib-
erative polyarchy combines the advantages of local learning and self-
government with the advantages (and discipline) of wider social learning
and heightened political accountability that result when the outcomes of
many concurrent experiments are pooled to permit public scrutiny of the
effectiveness of strategies and leaders.
One starting point for our argument is a commonplace of contempo-
rary political debate: that current economic and political institutions are
not solving problems they are supposed to solve, in areas of employment,
economic growth, income security, education, training, environmental
regulation, poverty, housing, social service delivery, or even basic per-
sonal safety. A second point of departure is the intrinsic appeal of collec-
We presented earlier versions of this essay at the University of Lausanne, the European Uni-
versity Institute, and the University of Chicago. We thank participants for instructive com-
ments. We also wish to thank Frank Michelman, Norman Daniels, Michael Dorf, Mark
Barenberg, and Sam Bowles for helpful criticisms and suggestions.
181
tive decision-making that proceeds through direct participation by and

reason-giving between and among free and equal citizens. Directly delib-
erative polyarchy is the natural consequence of both beginnings: desir-
able both in itself and as a problem-solver. That is what we hope to show,
or at least make plausible.
But obstacles lie along both paths. However commonplace the recog-
nition of institutional failures in problem-solving, the conventional cate-
gories used to explain those failures and defend strategies of repair ob-
scure important developments that suggest the plausibility of a directly
deliberative alternative. Moreover, gestures at radical democracy invite
skeptical observations about the “dark side” of localism or the scarcity of
evenings. And the force of such observations will only be deepened by
adding improbable claims about the problem-solving powers of participa-
tory self-government in vast, heterogeneous societies. To take the chill of
manifest implausibility from our project, therefore, we start by discussing
the limits of current debate as revealed in the promising developments it
overlooks (or misrepresents) and by specifying the criticisms of radical de-
mocracy to which we must respond if we are to offer more than a consol-
ing prospect for democrats in hard times.1
Consider first the conventional interpretations of institutional failure
and the projects of reconstruction associated with them. On one interpre-
tation, these failures reveal the limits of state regulation and suggest pos-
sibilities for a more comprehensive commodification of social life that
will finally lift the political fetters from the free exchange of individuals.
To fulfill this promise, we need only remove the detritus of twentieth-
century political failure. And that means constraining government from
doing anything wrong by constraining it from doing much at all: by frac-
turing political power both vertically and horizontally, setting stricter
1. For background on diagnosis and remedies, see Joshua Cohen and Joel Rogers, Associa-
tions and Democracy (London: Verso, 1995), especially the concluding chapter; Joshua Cohen,
“Procedure and Substance in Deliberative Democracy,” Seyla Benhabib, ed., Democracy and
Difference: Contesting the Boundaries of the Political (Princeton: Princeton University Press,
1996), 95–119 [reprinted as essay 5 of this volume]; and Charles Sabel, “Learning by Moni-
toring: The Institutions of Economic Development,” in Neil Smelser and Richard Swedberg,
eds., Handbook of Economic Sociology (New York and Princeton: Russell Sage and Princeton
University Press, 1995). For a companion essay on the constitutional and institutional implica-
tions of directly deliberative polyarchy, see Michael Dorf and Charles Sabel, “A Constitution
of Democratic Experimentalism,” Columbia Law Review 98, 2 (1998): 267–473.
182
D I R E C T L Y D E L I B E R A T I V E P O L Y A R C H Y
constitutional limits on government, and interpreting the rule of law as a

law of rigid rules.2
A counter-interpretation sees comprehensive commodification as a threat
to political arrangements carefully crafted earlier in this century to pro-
vide goods collectively that will not be provided individually, protect the
weak from the strong, and ensure that our destinies in life are not deter-
mined by the vicissitudes of market success. The correlative political
project is to protect the increasingly fugitive state from attack and hope
that a turn in the political cycle will restore public confidence in collec-
tive political action.3
Yet a third interpretation condemns the false dichotomy of state and
market. Well-functioning markets and well-ordered political institutions
can, it observes, be mutually reinforcing. Both, however, require prior
bonds of trust that can be undermined, but not created or sustained, by
self-interested market exchange or selfishly exercised political influence.
Those bonds depend, rather, on protecting family, church, and volun-
tary association—the pre-contractual, pre-political background responsi-
ble for accumulating the social capital we need to preserve our economic
and political artifice.4 But because such social solidarities are understood
as anterior to both economy and state—preconditions for the proper
functioning of both (on any conception of such proper functioning)—
the implications of such rebuilding for economic or political institutions
are entirely indeterminate.
We are skeptical about these contending diagnoses and remedies. A
2. See Friedrich Hayek, The Constitution of Liberty (Chicago: University of Chicago Press,
1960); James Buchanan, The Limits of Liberty (Chicago: University of Chicago Press, 1975);
Antonin Scalia, “The Rule of Law Is a Law of Rules,” University of Chicago Law Review 56, 4
(Fall 1989): 1175–1188; William Riker, Liberalism Against Populism (San Francisco: Freeman,
1982); William Riker and Barry Weingast, “Constitutional Regulation of Legislative Choice:
The Political Consequences of Judicial Deference to Legislatures,” Virginia Law Review 72, 2
(1988): 373–401.
3. See Robert Kuttner, Everything for Sale: The Virtues and Limits of Markets (New York:
Knopf, 1997).
4. See Robert Putnam, Democracy and the Civic Community: Tradition and Change in
an Italian Experiment (Princeton: Princeton University Press, 1992); idem, “Bowling Alone:
Democracy in America at the End of the Twentieth Century,” Journal of Democracy 1 (1995):
35–50. Michael Sandel’s remarks on a new public philosophy mix the sociological anxiety
characteristic of much communitarianism with the concerns about the political economy of
citizenship that are closer to our own focus on new arrangements of democratic governance:
see Michael Sandel, Democracy’s Discontent: America in Search of a Public Philosophy (Cam-
bridge, MA: Harvard University Press, 1995), 324–328.
183
number of emergent solutions to problems as varied as public safety

and public education seem not to result from either a shift in the bal-
ance between “state” and “market” forms of coordination or a shift in the
balance between these taken together and civil society. Instead of the
state’s retreat, or the market’s resurgence, or even the transfer of func-
tions from government to non-governmental organizations, secondary as-
sociations, civil society more broadly, or some other third something
alongside state and market, these phenomena suggest a set of changes
that disrupts those categories, the social-political boundaries they express,
and the associated idea that an effective polity is one that balances re-
sponsibilities optimally among the arrangements that fall within those
boundaries.
Consider, for example, community policing: a strategy for enhancing
public security that features a return of police officers to particular beats,
regular discussions between them and organized bodies in the communi-
ties they are policing, and regular coordination between those bodies and
the agencies providing other services that bear on controlling crime.5 Or
consider forms of school decentralization that—while shrinking school
size and permitting parents to choose schools—replace close controls by
central bureaucracies with governance mechanisms in which teachers
and parents play a central role. Or arrangements for local and regional
economic development, which include strong components of training
and service provision, and whose governance includes local community
interests, service providers, representatives of more encompassing organi-
zations, as well as local representatives of regional or national govern-
ment. Or, closely related to these arrangements, consider firm-supplier
relations that transcend episodic exchange to establish long-term collabo-
ration coordinated through regular discussions, disciplined by reference
to officially recognized standards—standards that commonly emerge in
regular discussions between and among groups of firms and suppliers
5. The Chicago experiment in community policing—the Chicago Alternative Policing

Strategy (CAPS)—involves assignments of police officers to single beats (30 square block areas)
for an entire year; neighborhood-based organizations called “problem-solving groups” that
work in partnership with police in each beat; and open meetings with community and police
each month. The emphasis on community participation distinguishes the Chicago scheme
from other strategies that share the label “community policing.” See Archon Fung, Empowered
Participation (Princeton: Princeton University Press, 2004).
184
and that may include public research, technical assistance, or training fa-
cilities as well.6
These new arrangements suggest troubles for the standard categories of
analysis and remedy. The arrangements are not conventionally public
because, in solving problems, they operate autonomously from the dic-
tates of legislatures or public agencies; they are not conventionally pri-
vate in that they do exercise problem-solving powers and their gover-
nance works through discussion among citizens rather than assignment
of ownership rights. At the same time, they do not presuppose a success-
ful, densely organized, trust-inspiring network of associations. Indeed,
they often emerge precisely against a background of associative distress.
Nor are these new arrangements mere intellectual curiosities. They are
attractive because they appear to foster two fundamental democratic val-
ues—deliberation and direct citizen participation—while potentially of-
fering advantages as problem-solvers that programs conceived within the
limits of conventional representative democracies do not. Indeed, if the
same properties made them both democratically and pragmatically at-
tractive, we would have a compelling case for the novel form of public
governance that we call directly deliberative polyarchy.
Because these new governance arrangements resonate so strongly with
the (often implicit) programmatic suggestions associated with radical
democratic criticisms of the modern state, a straightforward and appeal-
ing generalization of them seems at hand. Congenitally hostile to the
market inequalities and economic subordination, but always suspicious
of an overweening state as the best defense against them, radical democ-
racy emphasized the deficiencies of centralized power, the virtues of de-
centralization, the expressive and instrumental values of participation,
and the values of citizen discussion both as an intrinsically attractive form
of politics and as a good method of problem-solving.7
But evoking the core features of the radical democratic tradition—its
emphasis on direct participation and deliberation—immediately suggests
three lines of criticism. First, that in a large-scale political system, wide-
6. For discussion, see Charles Sabel, Local Development in Ireland: Partnership, Innovation,
and Social Justice (OECD, 1996); idem, “Milwaukee Jobs Initiative Consortia Employment
Project Description,” (unpublished, on file with author, 1996).
7. See, for example, Hannah Arendt, On Revolution (New York: Penguin, 1973); Jürgen
Habermas, Between Facts and Norms, trans. William Rehg (Cambridge, MA: MIT Press,
1996); we discuss our differences with Habermas and Arendt in section 5.
185
spread participation in decision-making is organizationally or administra-

tively impossible, so the ideal of radical democracy is vacuous. Second,
if participation could be ensured, the mutual reason-giving that consti-
tutes deliberation depends on a higher degree of homogeneity among cit-
izens than can reasonably be assumed in a large-scale, pluralistic democ-
racy. And third, direct decision-making requires a localism incompatible
with the constitutional safeguards needed to ensure equal treatment for
citizens.
Here, the threads of our argument come together: guided by the expe-
rience of emerging problem-solving institutions and mindful of the val-
ues associated with radical democracy, our aim is to sketch the alternative
social-political world of directly deliberative polyarchy in sufficient detail
to meet these objections. We start (section 2) by presenting an account of
the ideal of democracy and explaining why the properties of directness
and deliberativeness make highly participatory forms of direct democracy
especially compelling realizations of that ideal. To be sure, the classical
institutions of direct, assembly democracy are unavailable as realizations
of directness and deliberativeness. But by separating those properties
from their familiar institutional expressions, we suggest that they might
still guide current institutional reform. In section 3, we describe the cur-
rent practical impasse in problem-solving and propose that the roots of
that impasse lie in part in the mismatch between current arrangements of
constitutional democracy and fundamental properties of unsolved prob-
lems. In section 4, we describe the new form of state that would result
from the generalization of deliberative problem-solving arrangements
and foster their successful operation. We conclude with some reflections
on the idea of the public, indicating contrasts with the radical-demo-
cratic views of Habermas and Arendt.
Our approach is conjectural. We are guided by political values, a view
of current failures, and some hunches about promising developments.
But our aim is neither to articulate a set of normative principles and de-
duce institutional conclusions from them nor to predict the course of
current institutional evolution. Still less is it to explain fully the causes of
the failures of representative democracy or the origins of the new ar-
rangements. Instead, we take the very existence of these arrangements as
a sign of the insufficiency of theories that explain what democracy can do
and try to imagine what democracy could be from the vantage point of
the possibilities suggested by their presence.
186
2. What’s Good about Democracy?

Democracy is a political ideal that applies in the first instance to arrange-
ments for making binding collective decisions.8 Generally speaking, such
arrangements are democratic just in case they ensure that the authoriza-
tion to exercise public power—and that exercise itself—arises from col-
lective decisions by the citizens over whom that power is exercised.
The ideal of democracy comes in several variants, which are associated
with different interpretations of “authorization” and “collective decision.”
Our principal aim in this section is to sketch and defend a directly delib-
erative interpretation of the democratic ideal. We begin by exploring the
virtues associated with democracy quite generally, and we then consider
the special advantages of directly deliberative democracy, as against rep-
resentative-aggregative democracy. We conclude by returning to the con-
ventional criticisms of directly deliberative democracy, thus setting the
stage for our later efforts to describe a form of radical democracy that can
answer these criticisms.
Before pursuing these competing interpretations, however, we want to
clarify the relationship between those democratic ideals and the conven-
tional institutions of electoral democracy. Following Robert Dahl, we use
the term “polyarchy” to cover political systems in which virtually all
adults have rights of suffrage, political expression, association, and office-
holding, as well as access to diverse sources of information; in which
elected officials control public policy, and citizens choose those officials
through free and fair elections.9 Continuing to follow Dahl (and subse-
quent writers), we note that polyarchy has considerable value, both for its
intrinsic fairness and its instrumental success in keeping the peace and
protecting certain basic rights. It is not of value simply because it estab-
lishes the conditions required for achieving some greater ideal. Further-
more, under the modern circumstances of political scale and social plu-
ralism, polyarchal institutions are necessary for realizing fully an ideal
of democracy, however that ideal is specified. Though polyarchies can
be more or less democratic, making them more so does not require ne-
8. The ideal of democracy also has considerable force for organizations whose collective de-
cisions are not binding. But the rationale for democratic decision-making is most compelling
in the case of binding collective choices: that is, when members of the collectivity are expected
to regulate their own conduct in accordance with its decisions.
9. Robert Dahl, Democracy and Its Critics (New Haven: Yale University Press, 1989), 221–
222.
187
gating, sublating, or otherwise transcending the political institutions

definitive of polyarchy. This said, however, polyarchy is insufficient for
full democracy—or full political equality—because, for example, it is
compatible with inequalities in opportunities for effective political influ-
ence that would be condemned by any plausible statement of the ideal.
Building on these three considerations, then, we use the term “di-
rectly deliberative polyarchy” for a form of polyarchy distinguished by the
presence of a substantial degree of directly deliberative problem-solving.
(As we will see later, this presence transforms the role and the function
of conventional polyarchal institutions.) And we use the term “directly
deliberative democracy” for our account of the democratic ideal: fully
democratic arrangements that feature a substantial degree of directly de-
liberative problem-solving. Directly deliberative polyarchies, then, more
closely approximate the ideal of directly deliberative democracy than ex-
isting forms of polyarchy do, but—like polyarchies sans phrase—need
not have the entire range of qualities necessary for full democracy.
t hree v irt ues

Consider an ideal society whose members are free and equal and treat
one another as such. Very roughly, they are equal in that they all have, to
a minimally sufficient degree, a set of capacities whose possession makes
persons free.10 These freedom-making capacities include: the capacity to
regulate their conduct by reference to a conception of justice and a set of
ends with which they identify, to use practical reason to bring both to
bear on individual and collective conduct, to reflect on the plausibility of
both, and to adjust their aims to the requirements of justice. Though ac-
tual societies do not fully achieve this social ideal, modern democracies
impute these capacities to their citizens and arguably aspire to the ideal.
Assume now that the association needs to make binding collective deci-
sions. Why should a free association of equals make such decisions dem-
ocratically?11
10. Locke says that people are naturally equal in that all have the natural right to freedom.
See Second Treatise (Cambridge: Cambridge University Press, 1988), §54. For related discus-
sion, see John Rawls, Political Liberalism (New York: Columbia University Press, 1993), lec-
tures 1 (sec. 5) and 2; Charles Beitz, Political Equality (Princeton: Princeton University Press,
1989), chap. 5.
11. In addressing the question for this case, we do not mean to suggest that democracy is im-
portant only when these background assumptions are in place, but that the answer for different
cases will vary and that important considerations are likely to get lost if we confine attention to
answering the more general question.
188
Recall that democracies, abstractly conceived, are systems in which de-

cisions to exercise collective power are made in institutions that treat
those subject to such power as the ultimate authors. To that end, democ-
racies need at least to satisfy the conditions of polyarchy—to protect con-
stitutive liberties of participation, association, and political expression, es-
tablish direct or indirect electoral control of public policy, and ensure
adequate information. That said, the reasons for democratic authoriza-
tion divide naturally into goods intrinsic to the process and goods that ar-
guably result from it.12
First, democratic arrangements have the intrinsic virtue of treating
those who are subject to binding collective decisions with respect, as free
and equal: “The person of the humblest citizen is as sacred and inviola-
ble as that of the first magistrate.”13 Thus, the judgments of citizens, who
are expected to govern their conduct in accordance with collective deci-
sions, are treated by the processes of collective decision as equally author-
itative. Though decisions will rarely, if ever, be unanimous, no one’s
judgment of the proper rules of cooperation is treated as having greater
weight. Given the background conception of citizens as free and equal,
any assignment of differential weights to the views of different citizens is
a form of disrespect (unless it can be provided with a suitable justifica-
tion).14 Furthermore, the protection of the basic expressive and associa-
tive liberties establishes favorable conditions for reflecting on the plausi-
bility of alternative views about justice and on which ends are worth
pursuing. And the assurance of adequate and diverse information con-
tributes to the exercise of practical reason, in working out the implica-
tions of conceptions of justice and of suitable ends.
Second, democratic arrangements are instrumentally important: they
help protect the basic rights of citizens and advance their interests, as de-
12. A common rationale for democracy is that it treats people as equals by giving equal con-
sideration to their interests. Dahl, Democracy and Its Critics; Thomas Christiano, The Rule of
the Many (Denver: Westview, 1996). We avoid this rationale because the idea of equal consid-
eration of interests is normatively implausible in as much as it may conflict with the equal con-
sideration owed to persons.
13. Rousseau, Social Contract, trans. Victor Gourevitch (Cambridge: Cambridge University
Press, 1997), Book 3, chap. 14.
14. In the case of the U.S. Senate, for example, votes are of unequal weight because the po-
litical system relies on a scheme of territorial representation in which districts (in this case
states) correspond to political subdivisions: in this case, the inequality seems less objectionable
because it can be provided with a rationale that does not offend against the requirement of
treating members as equals.
189
fined by the ends and projects with which they identify. Thus, democra-
cies provide mechanisms for regular, popular authorization of exercises
of public power: in a representative democracy that means (at a mini-
mum) regular elections of legislators; in a direct democracy it means reg-
ular opportunities to review past decisions and evaluate the performance
of officials responsible for implementing those decisions. Such regular
renewal serves to make the exercise of collective power accountable to
the governed in the formal sense that the governed can impose sanctions
of removal from office on government. More fundamentally, an account-
able system for the exercise of collective power, in which citizens are
treated as equals, arguably helps ensure peaceful transitions of power, re-
strain the exercise of power by protecting majorities from minority rule,
avoid at least some egregious violations of minority rights, and foster
greater responsiveness of government to the governed.15
Both arguments—intrinsic and instrumental—are strengthened when
we consider, third, the educative aspects of democracy. Thus, by estab-
lishing the position of equal citizen, with associated entitlements to
participate in determining the terms of association, democratic arrange-
ments not only respect but also provide instruction in fundamental polit-
ical values—in particular, the value of equality itself and the conception
of citizens as free and equal. By participating, citizens acquire political
ideas in the light of which democracy itself is justified. Furthermore,
by opening debate to all and addressing problems through public dis-
cussion—rather than through market exchange or bureaucratic com-
mand—democracy not only assumes adequate information but also
helps ensure it. Democracy provides a way to pool dispersed information
relevant to problem-solving and to explore the range of possible solu-
tions to practical problems: in short, a framework for collective learning.
As Rawls puts it, within a democracy: “Discussion is a way of combin-
ing information and enlarging the range of arguments. At least in the
15. Such instrumental considerations played an important role in the U.S. Supreme Court’s
classic apportionment decisions, which urged that the same instrumental reasons that support
universal political rights also support equally weighted votes. Thus, in Wesberry v. Sanders: “No
right is more precious in a free country than that of having a voice in the election of those who
make the laws under which, as good citizens, we must live. Other rights, even the most basic,
are illusory if the right to vote is undermined.” Cited in Reynolds v. Sims, 377 U.S. 533, 558. Or
in Reynolds v. Sims itself: “Especially since the right to exercise the franchise in a free and
unimpaired manner is preservative of other basic civil and political rights, any alleged infringe-
ment of the right of citizens to vote must be carefully and meticulously scrutinized.” Reynolds,
562.
190
course of time, the effects of common deliberation seem bound to im-

prove matters.”16
t wo dimensions of de moc ra c y
We said that political institutions are democratic just in case they link the
authorization to exercise public power—and that exercise itself—to col-
lective decisions of citizens, understood as free and equal. There are, of
course, very different ways to interpret this abstract ideal of democracy,
corresponding to different interpretations of the notions of collective and
authorization.
Democratic collective decision-making can be either aggregative or
deliberative, depending on how we interpret the requirement that collec-
tive decisions treat citizens as equals. Understood aggregatively, a demo-
cratic decision is collective just in case the procedure gives equal consid-
eration to the interests of each person: it treats people as equals by giving
their interests equal weight in making a binding decision. Conventional
rationales for majority-rule as a method of collective decision rest on the
idea that it gives direct expression to this requirement of equal consider-
ation.17
Understood deliberatively, democratic decisions are collective just in
case they proceed on the basis of free public reasoning among equals: in-
terests unsupported by considerations that convince others carry no
weight. Put otherwise, in deliberative decision-making, decisions are to
be supported by reasons acceptable to others in the polity of decision-
makers; the mere fact that decisions are supported by a majority of citi-
zens, deciding on the basis of their interests, does not suffice to show that
the decisions are democratically authorized. On the deliberative inter-
pretation, then, democracy is a framework of social and institutional con-
ditions that both facilitates free discussion among equal citizens by pro-
viding favorable conditions for expression, association, and discussion
and ties the authorization to exercise public power—and the exercise it-
self—to such discussion, by establishing a framework ensuring the re-
sponsiveness and accountability of political power to it.
To be sure, discussion may not—and often does not—issue in agree-
ment. So even in a deliberative democracy, collective decisions must of-
16. John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), 359.
17. According to an epistemic conception of majority rule, the rationale is that decisions
supported by a majority are more likely to be right, not simply that the process visibly assigns
equal weight to the interests of each.
191
ten be made through voting, under some form of majority rule. But it
may be argued that if collective decision-making concludes in a vote,
then participants—anticipating that final stage—will not have any incen-
tive to deliberate earlier on and instead will simply seek allies for their
position. Though we cannot resolve the issue here, this outcome hardly
seems necessary. Even if all parties know that, at the end of the day, heads
may be counted, they still may accept the importance of finding consid-
erations that others acknowledge as reasons: they only need accept that
something other than a resolution that advances their antecedent inter-
ests matters to them. They may, for example, believe that reason-giving is
an important expression of respect or that deliberation sometimes yields
solutions that could not have been achieved if discussion were purely
strategic. If they do, they will be willing to deliberate in the stages leading
up to the vote. In short, the objection supposes that, once voting is in
prospect, interaction must turn strategic. But this view is no more plausi-
ble than the claim that moral advantages and possible mutual gains from
deliberation eliminate all strategic maneuvering.
As to authorization to make collective decisions, we have again two dis-
tinct understandings: in representative democracy, popular authorization
proceeds through a choice by citizens of representatives who decide on
content of public decisions. Citizens vote as individuals for persons who
will participate in making binding collective choices in an aggregative or
a deliberative legislature. In direct democracy, citizens authorize public
action by deciding on the substance of public policy. Again, those direct
decisions can be made either aggregatively, as some argue is true in refer-
enda, because of their yes/no structure,18 or deliberatively, as in an ideal-
ized town meeting, in which decisions on policy take place after debate
on the merits. The essential distinction between direct and representative
is not the level of participation but the topic on the agenda: direct de-
mocracy requires decisions on substance, whereas representative democ-
racy involves choices on legislators, who decide on substance.19
18. See Max Weber, “Parliament and Government in a Reconstructed Germany,” in Econ-
omy and Society, vol. 3, ed. Guenther Roth and Claus Wittich (New York: Bedminster, 1968),
p. 1455; Derek Bell, “The Referendum: Democracy’s Barrier to Racial Equality,” Washington
Law Review 54, 1 (1978): 1–29; Yannis Papadopolous, “A Framework for Analysis of Functions
and Dysfunctions of Direct Democracy: Top-Down and Bottom-Up Perspectives,” Politics and
Society 23, 4 (1995): 421–448.
19. Complexities arise when we think of systems with strong parties with well-defined policy
positions and ways of disciplining members who depart from those positions. Such systems
have a direct aspect. But we abstract from these subtleties here.
192
del ibera t iv e-dir e c t

Forms of democracy that are deliberative-direct seem especially attractive
in view of the three reasons for endorsing a democracy as a way to make
binding collective decisions. While those reasons support democracy
generally, they provide especially strong support for a deliberative-direct
democracy.
Consider, for example, the idea that democratic procedures are desir-
able because they treat citizens with respect, as free and equal. The delib-
erative conception offers a particularly forceful rendering of this condi-
tion. Suppose all participants support their views with considerations that
others regard as relevant and appropriate. Nevertheless, because of differ-
ences in views about the weight of those considerations, there is disagree-
ment about the right outcome. Still, the minority can scarcely contest the
fundamental legitimacy of the decision. After all, not only the proce-
dures but also the arguments themselves treat each as well as one can
reasonably demand.20 Thus the deliberative conception of collective de-
cision extends the idea of treating people with respect from rights and
procedures to justifications themselves. A similarly strong case can be
made for directly deliberative decision-making on the basis of the argu-
ments about instrumental benefits and learning. But we postpone con-
sideration of these until we have said more about the operations of di-
rectly deliberative polyarchy.
Despite these virtues as an expression of democratic values, radical de-
mocracy—a system with high degrees of directness and deliberative-
ness—is subject, we noted earlier, to a series of closely related criticisms:
that under modern conditions of political scale, it is not feasible, except
as local pockets of direct citizen engagement; that even within those
pockets—and certainly as scale increases—cultural heterogeneity thwarts
the mutual reason-giving that defines public deliberation; and that the lo-
calism characteristic of radical-democratic schemes leaves local minori-
ties at the mercy of their locality.
The starting point of these criticisms is the identification of radical de-
mocracy with direct assembly democracy and especially with the Greek
polis as both the ideal and the practical inspiration for modern critics of
centralized, representative democracy. In a direct assembly democracy,
legislative power—and the power to review conduct of all officials—is
20. It is not reasonable to demand to win: assuming disagreement, any decision will be op-
posed by some people.
193
vested in a body that all citizens may attend. In the case of the Athenian
ecclesia, that often meant meetings of five thousand (with women and
slaves excluded from participation). In the polis, the unit of collective
decision-making was small, and the members were homogeneous in gen-
eral outlook and sufficiently disconnected from banausic activities (be-
cause sufficiently secure in their social and economic positions) to de-
vote their passions and energies to common affairs. If the combination of
directness and deliberativeness can be achieved only under these condi-
tions, then the conventional criticisms of radical democracy are individu-
ally damaging and collectively overwhelming.
To vindicate the virtues of deliberativeness and directness, then, we
must distinguish these values themselves from familiar ways of institu-
tionalizing them—for example, citizen assemblies, or such modern anal-
ogies as workers’ councils or economic parliaments—and then describe a
modern set of arrangements of collective decision-making suited to these
values and to modern conditions of scale and heterogeneity. To guide
this elaboration of a workable direct and deliberative alternative to assem-
bly democracy, we need first to establish criteria for “workable” demo-
cratic solutions by characterizing the problems democracies now face
and the limits of representative, aggregative arrangements in addressing
them.
3. Diagnosis of Current Problems

Conventional explanations of current institutional failure range, we said
earlier, from too much state (and associated rent-seeking) to too much
market (private control of investment under conditions of globalization)
to too many civic deficits (decline of trust-building associations). And we
indicated, too, that emergent problem-solving institutions suggest the
limits of those explanations. But what could an alternative be?
Our own proposal is that existing forms of constitutional democracy—
and the associated boundaries between state, market, and civil society
that inspire the limited categories of current debate—block democratic
and effective strategies of problem-solving in the current environment:
where existing political institutions favor uniform solutions throughout a
territory, the problems require locally specific ones; moreover, the envi-
ronment is volatile, so the terms of those local solutions are themselves
unstable. In short, because of high diversity and volatility, important
problem-solving possibilities are not being exploited by existing institu-
tions. To the extent that this is so—to be sure, it is not the whole story—
194
the problems of modern democracy arise quite apart from the clash of
antagonistic interests or any guileful exploitation by individuals of block-
ages created by constitutional arrangements: they are (in the game-theo-
retic sense) problems of failed coordination, in which mutual gains are
available but different parties are unable to come to terms in a way that
captures those gains. If the right arrangements of collective choice were
in place, the parties could come to terms on one of the available alterna-
tives. In contrast, recognition of the mismatch between solutions and
available structures of decision-making leads, by itself, to paralysis, as it
reasonably suggests that it is better to do nothing than to do something
that will almost certainly fail.
Put another way, we assume that for some substantial range of current
problems, citizens agree sufficiently about the urgency of the problems
and the broad desiderata on solutions that, had they the means to trans-
late this general agreement into a more concrete, practical program,
would improve their common situation and would possibly result in the
discovery of further arenas of cooperation. This is not to make the fool-
ish claim that everyone endorses the same ranking of solutions, only that
everyone prefers a wide range of alternatives to the status quo. No sur-
prise, then, that the new problem-solving institutions have begun to
emerge just in those areas—public safety, public education, economic
restructuring—where established institutions have most conspicuously
broken down and the problems are agreed to be urgent. For breakdown
opens space for new initiatives, and where, as we are assuming, actors are
urgently motivated to look for a solution and prefer many alternatives to
the status quo, that space is likely to be occupied.
But even in thus qualifying the extent of agreement, we may still be ac-
cused of an extravagant confidence in consensus. In its stronger form,
this accusation rejects the idea of deliberative problem-solving altogether
by criticizing the assumptions about consensus on which it depends. It
asserts that the fundamental problem of politics is the pervasiveness of
deep disagreement, the consequent fragility of political order, and the
immanence of its disintegration into violence. So any assumptions about
agreement—and not simply the set just noted—miss the point.21 The
criticism is right in recognizing disastrous possibilities but wrong in the
21. See, for example, Carl Schmitt, The Concept of the Political (Chicago: University of Chi-
cago Press, 1995); Adam Przeworski, “Minimalist Conception of Democracy: A Defense,” in
Democracy’s Value, ed. Ian Shapiro and Casiano Hacker-Cordón (Cambridge: Cambridge
University Press, 1999), 23–55.
195
lessons it draws from them. Assume the setting of a consolidated poly-

archy: one in which there is no organized alternative to democracy, in
which democracy is “the only game in town.”22 And assume—as is sug-
gested by such consolidation—that citizens, who know that they disagree
on moral, religious, and political issues, nevertheless accept the impor-
tance of conducting political argument on common ground. Those as-
sumptions suffice to make deliberative politics possible.
In its more limited form, the objection is straightforwardly empirical:
we assert, and the critic denies, that there is currently substantial agree-
ment on a list of public problems and on the desiderata to their solution.
We point to the diffusion of new problem-solving arrangements; the critic
points to congressional gridlock; we think our diagnosis explains the grid-
lock; the critic thinks that the new arrangements are too marginal to re-
quire explanation. We propose not to adjudicate this disagreement here
but to reconfirm that our proposal, like all others, has its empirical com-
mitments.
To return to the diagnosis: at the root of this mismatch between prob-
lems and problem-solving institutions is, we assume further, a fundamen-
tal and familiar characteristic of contemporary political problems: diver-
sity. A commonplace of discussion of regulation and administration is
that rules and services aimed at achieving any broad end—protection of
the environment or training for economic activity—must be tailored to
(constantly changing) local circumstances to be effective. Moreover, be-
cause the pursuit of such ends often requires the integration of many
means—a regime of incentives and fines may have to be combined with
monitoring and clean-up programs as well as with research and develop-
ment efforts to achieve acceptable levels of environmental protection—
local combination of locally specific solutions are required as well.
But fundamental considerations of democracy apparently favor, if they
do not mandate, uniform solutions. Thus a basic democratic idea is that
citizens are to be treated as equals, which might be thought to imply that
state regulations are to be cast in the form of general rules. Why con-
strain the free play of interest through aggregation or deliberation only to
22. On the idea of consolidation, see Alfred Stepan and Juan Linz, “Toward Consolidated
Democracies,” Journal of Democracy 7, 2 (1996): 14–33. For doubts about the importance of
consolidation, see Adam Przeworski, “What Makes Democracies Endure?,” Journal of Democ-
racy 7, 1 (1996): 39–55. But note that Przeworski and others do not consider the importance of
consolidation, as characterized in the text.
196
allow the powerful to favor themselves by writing laws that accord them
benefits directly? It might be thought, too, to imply a requirement of
precision or a lack of ambiguity in those regulations. For why prevent
directly self-serving regulations, but then permit indirect self-service
through exploiting vagueness at the stage of interpretation and applica-
tion of laws?
Other, related devices of constitutional democracy have the same ef-
fects. Thus, a basic institutional expression of the requirement of the rule
of law—in particular, of the ban on self-serving interpretation—is the
separation of powers, understood as the requirement that rule-making au-
thority be vested in a body that includes representatives of diverse partic-
ular interests but that does not itself apply the rules it makes to individual
cases. The conventional rationale for this separation of rule-making and
rule-applying is that it permits diverse interests to be incorporated into
rules, even as it decreases incentives for rule-makers to design rules that
favor themselves (either as representatives or as officials). But in obstruct-
ing corruption, the separation of powers so understood reinforces the
substantive uniformity requirement and thereby tightens the constraint
on tailoring solutions to special circumstances.
Hence a familiar and inconclusive tug of war: when problems need to
be solved, pressure mounts to violate the constitutional constraints of the
rule of law and the separation of powers—to overturn the Tudor polity—
precisely because of the restrictions these impose on problem-solving.
Then, as the dangers of violations mount, as politics threatens to de-
generate into a patchwork of particularistic deals and local privileges, as
constitutional democracy approximates pre-Tudor feudalism—pressure
mounts to reimpose a system of strict rules. Thus, in the United States,
the standard criticism leveled against administrative agencies—created
precisely to adopt law to particular circumstances—is that they pave the
road back to serfdom. And standard proposals for reform—ranging from
Lowi’s juridical democracy and Sunstein’s post-New Deal constitu-
tionalism to Hayek’s neo-liberal constitutionalism—would redeploy rule-
making authority to legislatures in order to ensure such substantive uni-
formity, thus reimposing the very constraints that had prompted earlier
constitutional reform.23
This to and fro cannot be resolved simply by cutting the Gordian
23. For discussion of these three views, see Cohen and Rogers, Associations and Democracy.
197
knot of constitutional constraint. Absent the most stringent civic sensibili-

ties, a constitutionally unconstrained representative system—in which
decisions by a representative body suffice to make the regulation legiti-
mate, irrespective of concerns about substantive uniformity—produces
the Hayekian nightmare: a pure bargaining democracy, in which legisla-
tive decision-making is under no pressure to be deliberative and hence
under no pressure to explore improved solutions or even to meet mini-
mal conditions of coherence and efficiency. Outcomes will simply reflect
the balance of political forces, with no obligation to consider how legisla-
tive choices will cumulatively solve the problem.24 (According to public
choice views, this is all that democracy is, or could be. But this supposes,
improbably, that the real purpose of democracy is to achieve political
equilibrium, not to solve problems or establish the legitimacy of solu-
tions.)
Nor can the mismatch of institutions and problems produced by cur-
rent understandings of the rule of law and the separation of powers be
finessed by a strategy of federalist decentralization that would permit lo-
cal tailoring within a regime of strict rules. Federalism, generically con-
ceived, is a system with multiple centers of decision-making, including
central and local decision-makers and separate spheres of responsibility
for different units. In such a system, problems requiring local solutions
could be delegated to local centers of decision-making, while problems
admitting general solutions could be addressed centrally. If log-rolling
was Hayek’s nightmare of democracy, a radical version of federalism, in
which the center did little more than register the generalizable results of
local units, was his democratic arcadia.
But federalism, thus understood, creates troubles of its own, precisely
because it does not require the units of decision-making to communicate
and pool their information. To underscore the force of this point, we ex-
tend our original characterization of the problem situation of modern de-
mocracies beyond the assertion that uniform solutions are not optimal to
the further proposition that particular locations, operating in isolation,
lack the capacity to explore the full range of possible solutions. For this
reason, optimal problem-solving requires a scheme with local problem-
solvers who, through institutionalized discussion, learn from the suc-
cesses and the failures of problem-solving efforts in locales like their own.
24. Moreover, if legislators can secure their own re-election by servicing constituents, then
the limits on problem-solving are not a large source of electoral instability.
198
Through such exchanges, each problem-solving unit would be better sit-

uated to capture the benefits of all relevant, locally tailored solutions,
thus transcending the limits of localism without paying the price of uni-
formity such transcendence would otherwise require.
Federalism as currently understood does not foster such mutual learn-
ing from local experience; the scheme of a “directly deliberative poly-
archy” does. Indeed, abstractly conceived, it simply marries the virtues of
deliberation and directness to an ideal of learning by explicitly pooling
experiences drawn from separate experiments. Whether this marriage
can be made to work is our next subject.
4. Radical Democracy, after the Welfare State

The intuitive idea of directly deliberative polyarchy is to foster democ-
racy in its most attractive—direct and deliberative—form and thereby in-
crease our collective capacity to address unsolved social problems by
overcoming current dilemmas of coordination. As background, to re-
mind, we assume that the institutions of polyarchy are in place. More im-
mediately, we assume that citizens—despite conflicts of interest and po-
litical outlook—agree very broadly on priorities and goals but cannot
translate this preliminary agreement into solutions fitted to the diversity
and volatility of their circumstances because of constitutional uniformity
constraints. So we look for institutions that are friendly to local experi-
mentation and are able to pool the results of those experiments in ways
that permit outsiders to monitor and learn from those efforts.
Consider first the implications for individual decision-making units.
Diversity implies that reasoned decision-making in each will need to
draw on local knowledge and values; volatility means it will need regu-
larly to update such information. As each unit is distinct, none does best
by simply copying solutions adopted by others, though it may do well to
treat those solutions as baselines from which to move; as each faces
changing conditions, practical reasoning requires a system of collective
decision-making that fosters regular readjustment of solutions to those
changes. Local problem-solving through directly deliberative participa-
tion is well-suited to bringing the relevant local knowledge and values to
bear in making decisions. Direct participation helps because participants
can be assumed to have relevant information about the local contours
of the problem and can relatively easily detect both deception by others
and unintended consequences of past decisions. Deliberative participa-
199
tion helps because it encourages the expression of differences in outlook

and the provision of information more generally: the respect expressed
through the mutual reason-giving that defines deliberation reinforces a
commitment to such conversational norms as sincerity and to solving
problems rather than to angling for advantage (perhaps by providing mis-
leading information); furthermore, if preferences over outcomes them-
selves are shaped and even formed by discussion, and mutual reason-giv-
ing reduces disagreements among such preferences, then being truthful
will also be good strategy.
But the same concern for a form of decision-making that is attentive
to unexplored possibilities and unintended consequences requires in-
stitutionalization of links among local units—in particular, the institu-
tionalization of links that require separate deliberative units to consider
their own proposals against benchmarks provided by other units. Because
practical reasoning requires a search for best solutions, decision-makers
need to explore alternatives to current practice. A natural place to look
for promising alternatives—including alternatives previously unimagined
in the local setting—is in the experience of units facing analogous prob-
lems. Thus alongside directly deliberative decision-making we need de-
liberative coordination: deliberation among units of decision-making di-
rected both to learning jointly from their several experiences and to
improving the institutional possibilities for such learning. These consid-
erations lead us to our conception of directly deliberative polyarchy—in-
tuitively, a system with both substantial local problem-solving and contin-
uous discussion among local units about current best practice and better
ways of ascertaining it.
Before filling out this intuitive idea by exploring its basic operating
principles, we underscore that directly deliberative polyarchy describes
the form of problem-solving institutions: it is an order in which problem-
solving proceeds through connected institutions and organizations that
meet a set of abstract conditions of directness and deliberativeness. But
the institutions and organizations that meet those conditions might vary
widely, from networks of private firms to public institutions working
alongside associations. In this respect, the idea of directly deliberative
polyarchy operates at a different level of analysis from the idea of associa-
tive democracy or workplace democracy. The idea of associative democ-
racy is to solve problems through means other than states or markets:
the nature of the “organizational instrument” matters. Similarly, work-
200
place democracy specifies a particular institutional arena—the work-

place. With directly deliberative polyarchy, what matters is that the con-
ditions are met, not the organizations that satisfy them.
We emphasize, too, that directly deliberative problem-solving arrange-
ments must operate within a frame of legislative, judicial, and administra-
tive institutions. The role of those institutions changes, from seeking to
solve problems to identifying problems and fostering their directly delib-
erative solution. But in this transformed role, they are essential to the le-
gitimate and successful operation of the new problem-solving arrange-
ments.
To describe the basic structure of directly deliberative polyarchy, we need

to answer three questions:
1. What are the requirements of democratic process within and

among units? More particularly, what does it mean for their deci-
sions to be made deliberatively?
2. What conditions should trigger the operation of these deliberative
mechanisms?
3. How should the circle of membership in the deliberative bodies
be drawn?
Deliberative Process Within and Among Units. At the heart of the delib-
erative conception of democracy is the view that collective decision-mak-
ing is to proceed deliberatively—by citizens advancing proposals and de-
fending them with considerations that others, who are themselves free
and equal, can acknowledge as reasons. The shared commitment of citi-
zens in a deliberative democracy is that the exercise of collective power
should be confined to cases in which such justification is presented. Citi-
zens contemplating the exercise of collective power owe one another rea-
sons and owe attention to one another’s reasons.
But not all reasons are on a par. So the kind of attention owed must be
calibrated to the kind of consideration offered. Thus, constitutional rea-
sons are considerations that command substantial weight in decision-
making. In deciding which considerations are to be assigned such
weight, we look for a close connection to the standing of citizens as free
and equal members of the political society: considerations affirming that
201
standing have substantial weight, whereas those that deny it are weight-
less. Thus, citizens must have fundamental political and civil rights be-
cause those rights are backed by reasons that affirm the standing of citi-
zens as free and equal, whereas the denial of those rights requires appeal
to considerations that throw such standing into question perhaps by de-
nying that members meet all the qualifications for citizenship. But deni-
als of qualification—assertions that some member is not to be regarded as
a free and equal citizen—do not count as reasons at all because they are
not considerations that command respect from those whose standing is
denied. So effective participation rights cannot, except perhaps in very
special circumstances (perhaps cases of extreme emergency), permissibly
be denied. Similarly, proposals backed by reasons rooted in interests fun-
damental to the standing of members as free and equal can be rejected
only upon offering alternative, more plausible projects for advancing
those interests. Thus, a requirement of ensuring a basic educational
threshold—a threshold defined relative to participation as a citizen, and
more generally, as a cooperating member of society—would be a consti-
tutional reason, and a proposal that would ensure such a threshold would
be rejected in a well-ordered deliberative body only in favor of an alterna-
tive, better designed scheme.25
The first and most fundamental requirement of a directly deliberative
polyarchy is, therefore, that it affirm its character as democratically delib-
erative by giving stringent protection to claims backed by constitutional
reasons.
Of course, not all acceptable reasons for public choices are of constitu-
tional magnitude. The class of policy reasons comprises those consider-
ations whose endorsement is neither required by nor incompatible with
a conception of citizens as free and equal and which are relevant to an
issue under consideration. A proposal framed by such considerations
may reasonably be rejected by a counter-argument that articulates an al-
ternative balancing of the reasons generally understood as relevant to al-
locating the resource in question. Consider again the case of education.
In deciding how to allocate resources, some relevant and potentially
25. Amy Gutmann argues for a democratic threshold principle in her account of the distri-
bution of primary schooling in Democratic Education (Princeton: Princeton University Press,
1987), 136ff. But she confines the threshold to “effective participation in the democratic pro-
cess”—as though there were not an equally good claim to effective participation in labor mar-
kets.
202
competing policy reasons are: helping each student to fully achieve po-
tential; ensuring that students who are performing least well are given
special attention; ensuring common educational experience for students
of diverse backgrounds. In the case of health care, the reasons include:
helping those who are worst off; helping those who would benefit most
from medical resources; assisting larger numbers of people; ensuring that
people have fair chances at receiving help, regardless of the urgency of
their situation and of expected benefits from treatment.26
As these examples suggest, the policy reasons relevant to particular do-
mains are complex and varied, and there often will be no clear, princi-
pled basis for ranking them: different, equally reasonable participants in
deliberative process (and, a fortiori, different deliberative bodies) will
weigh them differently. Reasonable people and reasonable collective de-
cision-makers reasonably disagree and recognize the results of a delibera-
tive process in which such reasons are aired as legitimate.
This distinction between constitutional and policy reasons brings us to
the second broad condition, a requirement of substantive due process on
the operation of directly deliberative polyarchy: the process is to give due
consideration to reasons of both types, suitably weighted (and allowing
for reasonable differences of weight).
Moreover, we require, third, that this consideration be explicit. It is
not sufficient to require that outcomes be rationalizable—that the delib-
erative process issues in decisions for which appropriate reasons could
be cited—and to leave it to another institution, say, a court, to deter-
mine whether that condition is met. Outcomes in directly deliberative
polyarchy are to be arrived at through discussion in which reasons of the
appropriate kind are given by participants. Five considerations lead to
this conclusion:
1. Though deliberative justifiability itself is important, it must be aimed
at being achieved; that is, it will not in general be true that results
achieved through a process of exchange or bargaining, or outcomes that
reflect a balance of power, will be defensible by reasons of an appropriate
kind. So requiring actual deliberation helps establish a presumption that
results can be defended through reasons, and thus a presumption that the
outcomes of collective decision-making are legitimate.
26. See Norman Daniels and James Sabin, “Limits to Health Care: Fair Procedures, Demo-
cratic Deliberation, and the Legitimacy Problem for Insurers,” Philosophy and Public Affairs
26, 4 (1997): 303–350.
203
2. Offering reasons to others expresses respect for them as equal mem-

bers of a deliberative body. So actual deliberation plausibly helps foster
mutual respect, which, in turn, encourages citizens to confine the exer-
cise of power, as the deliberative idea requires. No similar result can be
expected if we assign the job of assessing the justifiability of outcomes to
a separate institution.
3. Actual deliberation provides a better rationale for relying on majority
rule, should there be disagreement. With reasons openly stated, everyone
can observe that the supporting considerations were relevant reasons, de-
spite disagreements about their proper weight. It is manifest to partici-
pants, then, that people are not being asked simply to accede to the larger
number but to accept what they can see to be a reasonable alternative,
supported by others who are prepared to be reasonable.
4. In actual reason-giving, citizens are required to defend proposals by
reference to considerations that others acknowledge as reasons and not
simply by reference to their own interests. To the extent that such public
reasoning shapes preferences, conflicts over policy will be reduced, as
will inclinations to strategically misrepresent circumstances. Moreover,
actual deliberation is, by its nature, a form of information pooling: when
people take seriously the task of providing one another with reasons and
information about circumstances and outlooks, what is relevant to im-
proved policy is then brought to bear by those in possession of it. No
similar effects on preferences or on information are likely to issue from
non-deliberative processes subject to subsequent review. Indeed, under-
standing the process of review as the natural forum of principle may well
encourage strategic, as distinct from deliberative, conduct.
5. Explicit reason-giving eases the work of other decision-making units
and of outside monitors: it provides a record that other decision-makers
can consult (and perhaps learn from) in deciding how to solve problems
and deciding which monitors (legislators and courts) can refer to in judg-
ing whether solutions adopted in particular locations are appropriate and
how they might be improved in the light of experience elsewhere.
Requiring explicit reason-giving rather than rationalizable outcomes
may, however, have a downside. Critics of deliberative decision-making
fault it for being doubly exclusionary.27 Deliberation, they say, is a partic-
27. See Iris Young, “Communication and the Other: Beyond Deliberative Democracy,” in
Benhabib, ed., Democracy and Difference, 120–135; Sidney Verba, Kay Lehman Schlozman,
Henry E. Brady, Voice and Equality: Civic Voluntarism and American Politics (Cambridge,
MA: Harvard University Press, 1995), 500–508.
204
ular discursive style, with all the conventional indicia of the rational: for-
mal, deductive, and unemotional. By insisting on abstraction from the
personal and particular, deliberation excludes both people and informa-
tion. People, because it silences citizens whose discursive style is de-
tailed, narrative, and passionate; information, because it invites only con-
tributions cast in general terms. As a result, deliberation is unfair and
ineffective. Urging more of it is a reform strategy, but not an especially in-
viting one.
This objection makes two assumptions, both unwarranted. First, that
requiring an explicit statement of reasons implies that nothing other than
reasons can be stated—as though a conception of deliberative justifica-
tion supported a ban on undeliberative humor. Second, that the canoni-
cal form of deliberation is the justification of a regulation from first prin-
ciples: the argument for progressivity in the tax system on grounds of a
conception of political fairness. Deliberation may take this form, but
nothing in the concept of reason-giving requires that it do so. Nor, more
immediately, is the reason-giving that occupies us here naturally ex-
pressed in the form of deductions from general political axioms. On the
contrary, deliberative problem-solving is by its nature focused on address-
ing specific problems in local settings. Giving reasons under these condi-
tions is, generally speaking, a matter of offering considerations recog-
nized by others as pertinent to solving the problem at hand. It is simply
impossible to limit in advance the kinds of considerations that might be
relevant or the form in which those considerations are to be stated. In-
deed, deliberation will characteristically involve debating the implications
of general principles (standard operating procedures, rules of thumb) in
the light of the particulars of local experience and inviting discussion
of such experience in whatever terms suit participants—including the
ironic “yeah, yeah” that condemns the latest implausible suggestion.
Still, it might be said that requirements of deliberation unfairly bias de-
cision-making in favor of the verbal, that we may end up with a pluralis-
tic logocracy, in which the many forms of verbosity are all on display, but
the shy, the quiet, and the reserved are left out. We agree that there is a
difficulty here, but why isn’t it remediable? In settings of deliberative
problem-solving, everyone has something to contribute, so the first task
in improving the operation of deliberative arrangements is to ensure that
all participants understand that and are encouraged to contribute. The
potential for deliberative failure is no argument against efforts at such im-
provement.
205
These reasons for preferring decisions by actual deliberation—particu-

larly the last consideration—suggest a fourth requirement of democratic
process: that there be like deliberation among units as well. The advan-
tage of actual, deliberate consideration of alternatives by citizens of equal
standing but diverse experience and disposition is that the diversity of
viewpoints brings out the strengths and the weaknesses of diverse propos-
als. Moreover, the diversity of proposals reveals strengths and weaknesses
in viewpoints that make for more careful assessment in later rounds. Ex-
tending deliberation across units allows each group to see its viewpoints
and its proposals in the light of alternatives articulated by the others: in
effect, it ensures that the exercise of practical reason is both disciplined
and imaginative.
To be effective in provoking this kind of informative comparison, infor-
mation provided for this purpose must be supplied by units in a way that
both anticipates and reflects this use: in accounting for their own deci-
sions, decision-making processes, and outcomes, units must take into ac-
count information about the relevant practice elsewhere or make a case
that apparently better practice is either not genuinely better or not rele-
vant to their circumstances because of differences in population or re-
sources. A standard way of doing this is through benchmarking: evalua-
tion of one’s own activities by comparison with others, judged to be
similar, by means of metrics inherent in the choice of the comparison.
Benchmarking thus requires a survey of possible comparisons, evaluation
of possible metrics, and revision, when necessary, of initial choices of
both; and the effectiveness of such surveys, evaluations, and revisions de-
pends on the willingness of all participants to disclose information in
view of the investigations of the others. This amounts to requiring that, as
when acting alone, units actually deliberate among themselves, in the
sense of taking account of respective reasons, and not content themselves
with deliberative justifiability. This requirement implies that units that
show poorly in public comparisons will be under substantial pressure to
improve their practice to meet the standard of performance set in other
comparable units.
Responsibility for ensuring that deliberation within and among units
meets these four conditions falls ultimately to authorizing and monitor-
ing agencies—legislatures and courts. But, in contrast to the con-
ventional “division of deliberative labor,” this responsibility is to be dis-
charged by ensuring that the relevant decision-making bodies act
206
deliberatively not—so far as possible—by substituting for their deci-

sions.28 We return to these points below, in our discussion of institutions.
State/Market? With these core conditions in place, we come to the areas

of policy for which directly deliberative polyarchy is particularly well-
suited. Generally speaking, the institutions of directly deliberative poly-
archy are designed to do well where current political institutions and
market exchange do badly.
Consider first the limits on political institutions. These limits are most
severe when the following four conditions hold:
1. The sites at which a problem arises and requires address are too nu-
merous and dispersed for easy or low cost centralized monitoring of com-
pliance with regulations. Even if uniform regulations were appropriate,
these conditions would suggest a need for decentralizing the capacity to
monitor compliance. Discussions of workplace health and safety regula-
tion commonly emphasize this problem: too many workplaces for a cen-
tral inspectorate to review.
2. The diversity of sites at which similar problems arise suggests that
problem-solvers at different sites will want to employ different means to
achieve similar aims and specify their aims differently.
3. The volatility of sites suggests a need for continuous reflection on
means and ends and the importance of adjusting both in the light of new
information about the environment.
4. The complexity of problems and solutions—where problems are
substantially the product of multiple causes and connected with other
problems, crossing conventional policy domains and processes—implies
that the appropriate strategy requires coordination across those domains.
Urban poverty, local economic development, and effective social service
delivery are among the familiar problems that occupy this class. Solving
them plausibly requires cooperation across quite different institutions
and groups—for example, lending institutions, health care providers,
technology diffusers, education and training establishments, housing au-
28. Susan Sturm, “A Normative Theory of Public Law Remedies,” Georgetown Law Journal
79, 5 (1991): 1357–1446. Also, Rebecca Abers discusses the requirement that citizen-budgeters
incorporate considerations of fair distribution in the deliberations in the Porto Alegre system:
“Learning from Democratic Practice,” in Cities for Citizens: Planning and the Rise of Civil So-
ciety in a Global Age, ed. Mike Douglass and John Friedmann (Chichester: John Wiley, 1998),
39–65.
207
thorities, community development corporations, and neighborhood asso-

ciations.
When all these conditions are in force, we have a strong case for di-
rectly deliberative polyarchy, with its linked, local problem-solvers:
because of the numerosity and diversity of sites, we want a structure of de-
cision-making that does not require uniform solutions; because of volatil-
ity, we want a structure with built-in sensitivities to changing local condi-
tions; because of the complexity of problems, we want a structure that
fosters interlocal comparisons of solutions.
To be sure, departures from these four conditions imply a less strong
case for directly deliberative polyarchy and a correspondingly stronger
case for markets or regulatory solutions. But even in the face of depar-
tures from these ideal conditions, two considerations support the case for
directly deliberative polyarchy. First, as we have urged, it fits with demo-
cratic values, and that fit will tip the balance in unclear cases. Second,
our basic premise is that existing strategies of problem-solving are not
working well. So we may be aided in diagnosing the shortcomings of
those strategies if we try this alternative. Among other things, it will test
the thesis that the troubles emerge from a mismatch of problems and in-
stitutions of collective choice.
Consider next the circumstances under which problem-solving
through directly deliberative polyarchy is preferable to solution through
market exchange—here understood as a form of social coordination in
which agents need not arrive at a common decision nor defend their sep-
arate decisions by giving reasons to others.29
Thus suppose we are concerned about the production and allocation
of a good that is widely regarded as urgent—that citizens can claim as
a matter of basic right or need—and about whose proper production and/
or allocation there is disagreement.30 Because the claims for the good
are urgent, arrangements of provision should be open and accountable;
moreover, urgency and disagreement together establish a presumption
that decisions about the good’s provision should be backed by an accept-
able rationale. That presumption can be defeated in the case of goods
(for example, bread or cars) for which there are a large number of provid-
29. Bargaining and command are ways of making collective decisions without mutual rea-
son-giving; confession is the practice of giving reasons to others for individual decisions.
30. The paragraph that follows presents an account of public goods suited to the special set-
ting of a conception of deliberative democracy.
208
ers and about which it is relatively easy (either for consumers or for a cen-
tralized monitor) to acquire accurate information. Assume, then, that the
good is best supplied by a restricted range of providers, and that there are
high costs to switching among those providers: there can, then, be no
presumption of voluntarism in the choice of provider. Add, now, that in-
formation about the good is difficult to acquire or summarize because a
large number of dimensions are important to its evaluation, people dis-
agree about the relevant dimensions and their relative weights, and the
conditions of its production and allocation are volatile. Under these con-
ditions, we want goods to be provided through mechanisms in which de-
cisions are backed by reasons and based on pooled information. In short,
we have reason to favor directly deliberative polyarchy over market.
Membership. Finally, as to membership. The basic standard is that di-

rectly deliberative arenas are to be open to providers and parties affected
by the extent and the manner of provision. (In the case of schools, for
example, parents, teachers, and residents of a community served by a
school.) Very little can be said in general terms about the requisite repre-
sentational form: how many members of different groups, affected par-
ties, and so forth. Once more, there is every reason to expect at least as
much variation as we currently see in polyarchies. Still, a few consider-
ations are to frame debate about whether deliberative bodies include all
who are entitled or are instead objectionably exclusive. In general terms,
the considerations pull in two directions, reflecting the ideas of political
equality and deliberativeness that define the directly deliberative concep-
tion. The value of equality suggests a one-person/one-vote composition of
deliberative bodies, whereas the requirement of deliberativeness suggests
a constitution that assigns membership in ways that foster the provision of
relevant local information and the crisp articulation of alternative views.
More particularly, then, three considerations need to be balanced in
decisions about membership. First, citizens can object that the composi-
tion and the scope of directly deliberative bodies are objectionably dis-
criminatory—for example, that their geographic range has been gerry-
mandered on racial or ethnic lines. Second, there is a presumption in
favor of equal membership for affected parties—open meetings, with
equal rights to participate in discussion and decision-making for all af-
fected parties. Third, rights to participate might also be awarded to orga-
nizations with special knowledge that is essential to the problem area in
209
question (for example, neighborhood organizations in the area of public

safety) or which are able to articulate a point of view in ways that foster
deliberation among alternative solutions.
ef f ect iv eness
Why expect that such problem-solving will have concrete benefits? How
might it be able to overcome the problems of limited information and
diversity of sites that vex state action? Five considerations are important.
First, the parties to the discussion are presumed to have relevant local
knowledge; moreover, they can put that information to good use because
they understand the terrain better and have a more immediate stake in
the solution.
Second, assuming a shared concern to address a problem, and an ex-
pectation that the results of deliberation will regulate subsequent action,
the participants would tend to be more other-regarding in their politi-
cal practice than they would otherwise tend to be. The structure of dis-
cussion—the requirement of finding a solution that others can agree to
rather than pressuring the state for a solution—would push the debate
in directions that respect and advance more general interests. Other-
regardingness would encourage a more complete revelation of private in-
formation. And this information would permit sharper definition of prob-
lems and solutions.
Third, pursuing discussion in the context of enduring differences among
participants would incline parties to be more reflective in their definition
of problems and proposed strategies for solution; it would tend to free dis-
cussion from the preconceptions that commonly limit the consideration
of options within more narrowly defined groups, thus enabling a more
complete definition and imaginative exploration of problems and solu-
tions. The same is true for the federalism of problem-solvers that emerges
from requirements of discussion across units—here, too, comparisons of
solutions at different sites, and benchmarking of local solutions by refer-
ence to practice elsewhere, suggest a basis for improving local practice.
Here, notice that directly deliberative polyarchy—understood as a
form of problem-solving—is not thwarted by, but instead benefits from,
heterogeneity of participants. Of course, the participants must—as our
discussion of deliberation indicates—share a view about relevant rea-
sons. But this is, we think, a rather weak constraint that does not demand
substantial homogeneity—certainly not homogeneity of comprehensive
moral outlook.
210
Fourth, monitoring in the implementation of agreements would be a

natural by-product of ongoing discussion, generating a further pool of
shared information.
And, fifth, if things work, the result would be a mutual confidence that
fosters future cooperation.
In all these ways, then, deliberation about common problems with di-
verse participants might thus reasonably be thought to enhance social
learning and problem-solving capacity.
inst it ut ions
We conclude with a brief discussion of the implications of directly delib-
erative polyarchy for the design of and expectations on basic political in-
stitutions.
First, directly deliberative polyarchy is, as we have indicated, a form of
polyarchy. So we assume the continued presence of the legislatures,
courts, executives, and administrative agencies, controlled by officials
chosen through free and fair elections, in which virtually all adults have
rights to suffrage, office-holding, association, and expression, and face al-
ternative, legally protected sources of information.31 Though the opera-
tion of these institutions and arrangements changes, the institutions re-
main and continue to serve some of the political values with which they
are conventionally associated: peaceful transitions of power, restraints on
unbridled power, fair chances for effective influence over authoritative
collective decisions, opportunities to develop informed preferences, and
so forth.
But with the shift in the locus of problem-solving, the operations and
expectations of basic institutions change markedly.
Consider first the role of legislatures. Directly deliberative polyarchy is
animated by a recognition of the limits on the capacity of legislatures to
solve problems—either on their own or by delegating tasks to administra-
tive agencies—despite the importance of solutions. Rejecting the Neo-
Liberal Constitutionalist idea that the problems are essentially recalci-
trant to collective address, and the modern Civic Republican idea that
their address requires only a more vigilant exclusion of private interests
from national policy-making (and a correspondingly more acute inter-
vention by technically adept guardians of the common good), the legisla-
ture in a directly deliberative polyarchy takes on a new role: to empower
31. Dahl, Democracy and Its Critics, 221.
211
and facilitate problem-solving through directly deliberative arenas oper-

ating in closer proximity than the legislature to the problem. More partic-
ularly, the idea is for legislatures, guided by the conditions of triggering,
to declare areas of policy (education, community safety, environmental
health) as open to directly deliberative polyarchal action; state general
goals for policy in the area; assist potential deliberative arenas in organiz-
ing to achieve those goals; make resources available to deliberative prob-
lem-solving bodies that meet basic requirements on membership and
benchmarking; and review at regular intervals the assignments of re-
sources and responsibility. To be sure, legislatures can play this role only
if they are able to identify problems needing solutions and agents with
the capacity to solve those problems, even when they cannot themselves
produce the solutions. But once we acknowledge the importance of di-
versity and volatility in shaping acceptable solutions, this assumption is
entirely natural.
This changed role for legislatures does not, of course, preclude na-
tional solutions through legislative enactment when uniform solutions
are preferable (because of limited diversity among sites) or when exter-
nalities overwhelm local problem-solving. Instead, the availability of al-
ternative methods of problem-solving imposes on legislatures a greater
burden in justifying their own direct efforts: they must explicitly make the
case that the benefits of those efforts suffice to overcome the advantages
of direct-deliberative solutions.32
Administrative agencies, in turn, provide the infrastructure for infor-
mation exchange between and among units—the exchange required for
benchmarking and continuous improvement. Instead of seeking to solve
problems, the agencies see their task as reducing the costs of information
faced by different problem-solvers: helping them determine which delib-
erative bodies are similarly situated, what projects those bodies are pursu-
ing, and what modifications of those projects might be needed under lo-
cal conditions.
And the responsibility of constitutional courts is not simply to inspect
procedure for its adequacy as representative, nor to reorganize institu-
tions by reference to substantive constitutional rights, but to require that
decision-making proceed in a directly deliberative way: that is, to require
32. For related discussions of federalism, see Stephen Gardbaum, “Rethinking Constitu-
tional Federalism,” Texas Law Review 74 (1996): 795–838; and the account of the “comman-
deering problem” in Dorf and Sabel, “Constitution of Democratic Experimentalism.”
212
that problem-solvers themselves make policy with express reference to

both constitutional and relevant policy reasons. You might describe this
as a genuine fusion of constitutional and democratic ideals: a fusion, in-
asmuch as the conception of democratic process includes a requirement
that constitutional reasons be taken into account, as such. The aim is a
form of political deliberation in which citizens themselves are to give
suitable weight to constitutional considerations and not leave that re-
sponsibility to a court.
These remarks sketch, in the sparest terms, how basic political institu-
tions might shift in expectation and responsibility under conditions of di-
rectly deliberative polyarchy. Further details will vary greatly, certainly as
much as they do in existing polyarchies. Rather than outline the dimen-
sions of such variation, we propose to clarify and deepen this account of
transformed conventional institutions by addressing an objection to the
very coherence of directly deliberative polyarchy as a form of problem-
solving that conforms to basic democratic values. Generally speaking, the
objection is that directly deliberative polyarchy is an unstable combina-
tion of institutionalization of democratic values: central institutions ei-
ther will not supervise local arrangements enough to avoid local tyran-
nies or will over-supervise, thus regenerating the problems of centralized
control that directly deliberative polyarchy is supposed to avoid. More
particularly, the objection is that directly deliberative polyarchy needs to
meet two requirements that are at war with one another: deliberative
problem-solvers are supposed to satisfy various conditions (on member-
ship, deliberativeness, and external links to other problem-solvers). But
directly deliberative problem-solvers will not meet these conditions as a
matter of course, nor is their satisfaction a self-enforcing equilibrium. So
the responsibility for ensuring that they are met falls to authorizing and
monitoring agencies. If, however, problem-solvers are to achieve the vari-
ation in local solutions demanded by conditions of diversity and volatil-
ity, then authorizing and monitoring agencies must also ensure them
autonomy.
These two conditions are arguably in tension. For directly deliberative
problems-solvers can act in ways that conflict with the constitutive values
and conditions of democracy itself, either by deciding on the basis of con-
siderations that conflict with those values or by failing to take them fully
into account. If a fundamental, generic responsibility of authorizing and
monitoring bodies is to ensure that decision-making is democratic, then
213
those authorizing bodies are obligated to review and pass judgment on

the decisions of the authorized bodies. But this creates two related trou-
bles for directly deliberative problem-solving: first, reduced autonomy
in the name of ensuring democracy may substantially limit interest and
enthusiasm for participating in problem-solving bodies. Second, if ensur-
ing democracy means constantly second-guessing the solutions chosen
by directly deliberative problem-solvers, then those problem-solvers may
decide to avoid troubles by imposing uniform solutions (choosing solu-
tions that have already passed muster), disregarding the suitability of
those solutions to their circumstances. This tension, it might be argued,
is exacerbated by a tendency of decentralized systems to generate greater
inequalities, thus pressuring the center to reappropriate power and im-
pose greater uniformity of circumstance through redistribution.
We have five replies to this problem. The first is to introduce a note of
realism. The objection is entirely familiar from current discussions of
federalism and of relations between courts and legislatures. Focusing on
the latter, it is commonly agreed that courts should, whatever else they
do, uphold the democratic process, ensuring that all citizens have rights
to participate as equals in that process.33 Sometimes majorities violate
that requirement, and when they do, courts have a responsibility to over-
turn the results of those violations. The tension noted above is, gener-
ally speaking, simply an instance of this problem, which is commonly
called the “countermajoritarian dilemma”: it is not a problem created
by the proposal advanced here, but a reflection, within our proposal, of
a problem that any adequate conception of constitutional democracy
needs to face.
Second, accepting that the general structure of the problem is familiar,
it might nonetheless be argued that a deliberative conception of democ-
racy or an idea of directly deliberative polyarchy inspired by that con-
ception worsens the problem—by imposing more stringent standards of
democracy. Though a wide range of views will permit review and rejec-
tion of decisions on grounds of incompatibility with democracy, the de-
liberative view embraces an expansive conception of democracy—and a
correspondingly expansive and, therefore, invasive account of—when the
judgments of problem-solvers are properly second-guessed. The force of
this objection depends on a belief that is widely shared but simply mis-
33. The classic statement of this view is John Hart Ely, Democracy and Distrust (Cam-
bridge, MA: Harvard University Press, 1980).
214
guided: that deliberation, properly conducted, issues in consensus. We

have already explained our reasons for rejecting this claim. Deliberation
is a matter of balancing relevant considerations and arguing in the light
of such balance: competent deliberators will work out the balance differ-
ently; and, correspondingly, competent deliberative bodies will typically
arrive at different conclusions or will arrive at the same conclusions dif-
ferently. Indeed, there is no compelling a priori argument that the range
of acceptable results of deliberative processes is smaller than the range of
acceptable results of aggregative processes. So we reject the claim that
the deliberative view worsens the familiar problem.
Indeed, third, we think that the deliberative conception may reduce
the tension between democracy and autonomy. To see how, recall the
idea of a division of deliberative labor. On a conventional view of collec-
tive decision-making within a constitutional regime, the division of labor
assigns to legislatures the responsibility for devising laws that advance
the common good, and to courts the responsibility for ensuring that those
laws respect the constitution and the political values implicit in it. As
our earlier discussion indicates, the deliberative view rejects this way of
dividing deliberative labor. When objections are raised on constitutional
grounds to decisions reached by problem-solvers—when it is argued that
unacceptable reasons animated the decision, or that fundamental consti-
tutional values were neglected by it—the role of courts (and legislatures)
is not to substitute their own judgment about the proper outcome, but to
require that the deliberative body revisit the issue, taking the full range of
relevant considerations explicitly into account—and exploring the expe-
rience of similarly situated problem-solving bodies. Suppose, for exam-
ple, that a decision to impose an English-only requirement on schools is
challenged on grounds that students who are not native English speakers
will be disadvantaged by it, and consequently disadvantaged as citizens.
The response should be to require that the school committee responsible
for imposing the requirement revisit the decision, attending both to the
importance of education for equal citizenship and to the experience of
other multilingual districts in solving the problem. In short, the delibera-
tive view rejects the conventional division of deliberative labor, propos-
ing instead that all bodies making collective decisions share responsibility
for upholding the democratic constitution by treating its principles and
values as regulative in their own decisions.
Our fourth reply builds on this last point. Suppose that deliberative de-
215
cision-makers are required to arrive at decisions with explicit attention to

constitutional values and comparable experience. Still, they may make
decisions that conflict with the democratic constitution, and courts may
be required to review their decisions in this light. But when they are, they
will have a record of fact and reasoning to draw on in making their deci-
sions. Because they have imposed requirements of due consideration on
problem-solvers themselves, courts will have the information they need
to decide whether means are suitably tailored to ends, and whether ends
are specified in ways that satisfy constitutional constraints. Judgments
about whether or not to defer to problem-solvers will be backed by fact,
and not simply by a priori estimates of institutional competence.
Finally, without disputing the claim that decentralized systems, as a
rule, generate increased inequality, we dispute the extension of the rule
to the case of directly deliberative polyarchy and, therefore, do not expect
substantial pressures to recentralize in the name of equality.
5. Kicking Radical Democracy Upstairs?

Any plausible conception of democracy requires an interpretation of the
idea of the public as the arena in which free and equal citizens reflect on
and seek to advance common aims. We conclude our account of directly
deliberative democracy, then, by sketching the distinctive conception of
the public that has been implicit in our discussion thus far.
First, in directly deliberative democracy (and, by extension, in directly
deliberative polyarchy), the public arena is organizationally dispersed
and socially heterogeneous: organizationally dispersed, because public
opinion crystallizes not only in reference to the national legislature but
also in the work of the local school governance committee and the com-
munity policing beat organization, and in their analogies, in areas such as
the provision of services to firms or to distressed families; socially hetero-
geneous, because members are not presumed to share social traits, moral
outlooks, or common information. Though the public arena is in both
ways pluralistic, its pieces are nevertheless connected by: the require-
ments of reason-giving, in particular the demand to respect constitu-
tional reasons; the need for explicit comparison with other units, which
are themselves conducting similar comparisons; and the need for a wider
public debate informed by such comparisons and focused on national
projects.
Second, and more fundamentally, the public arena is the place where
216
practicality in the form of problem-solving meets political principle in

the form of deliberation through reason-giving among citizens who rec-
ognize one another as free and equal. In directly deliberative polyarchy,
public deliberation cuts across the distinction between reflection on po-
litical purposes and assessment of efforts to achieve those purposes—a
central distinction not only in familiar theories of representative democ-
racy but also, and perhaps surprisingly, in current understandings of radi-
cal democracy. To underscore this essential feature of directly delibera-
tive polyarchy and point towards the unfinished work of our project, we
look briefly at the fate of the modern radical-democratic understanding
of the public as it appears in the works of such representative figures as
Habermas and Arendt and their innovative followers, and we contrast
that fate with the idea developed above.
From this vantage point, the most striking feature of contemporary
views of radical democracy is the measure to which they have become
defensive, self-consciously chastened, typically directed more to limiting
(at times by novel means) the erosion of the institutions of nineteenth-
century parliamentary democracy than to transforming and extending
them. In part these limited ambitions are a prudent response to the tem-
per of the times, hostile since the fall of the planned economies to
any hint of collective control over life choices of individuals, and skepti-
cal, more broadly, about the very idea of public action.34 More funda-
mentally, though, this self-limitation reflects a sharp distinction, long
established in the social and political theory from which much radical
democratic theory stems, between a higher, political world of human
self-determination—through an all-encompassing exercise of theoretical
and practical reason or innovative public deeds—and a lower realm of
workaday conduct governed by calculation, technique, and organization.
The idea of the public in directly deliberative democracy questions the
underlying assumptions and institutional expression of this distinction.
Consider first the notion of the public as it appears in Habermas’s
work. According to his theory of communicative action, human inter-
actions differ fundamentally according to whether participants aim to
34. Habermas says that he has “no illusions about the problems that our situation poses and
the moods it invokes. But moods—and philosophies in a melancholic ‘mood’—do not justify
the defeatist surrender of the radical content of democratic ideals.” Between Facts and Norms,
xlii–xliii. We agree with the observation about moods and their unfortunate consequences,
but—as will emerge—think Habermas has surrendered too much.
217
achieve worldly success, in part by influencing others, or to coordinate

social action through common understandings. When action aims at
success, information is manipulated strategically to advance individual or
group interests, as in economic exchange or group bargaining. When
action aims at understanding, agents acknowledge that they are bound
by context-transcendent norms of sincerity (Wahrhaftigkeit), truth
(Wahrheit), and moral probity (Richtigkeit), and they are committed to
the view that their claims would be vindicated by an unfettered commu-
nication of equals.
Within the framework provided by the idea of communicative action,
the role of constitutional democracy is threefold: it establishes the system
of rights required for autonomous communication among equal citi-
zens (for a discursive formation of public opinion), enables discursively
formed understandings of common purpose to achieve legal expression,
and ensures that those legally crystallized understandings steer (or, in
more anguished formulations, besiege) the state’s administrative appara-
tus, understood as a system of technically constrained instrumentali-
ties for guiding the (still more constrained) activities necessary for society
to reproduce and advance.35 Democratic steering is itself divided into
stages, or phases, ordered by their distance from the apparatus of actual
decision-making, and hence their freedom from technical constraint and
organizational routine: parliamentary debate is limited by its connection
to administration, the disputations of political parties by their connection
to parliamentary debate. In the “communicatively fluid” public sphere,
democracy is most authentic because it is least constrained. Neither
hemmed in by specialized vocabularies nor confined to particular social
tasks, the public sphere is a dispersed, all-purpose, discursive network
within which citizens, connected by the means of mass communication,
form currents of opinion in seeking how best to resolve the great ques-
tions of the day.36 Because discussion within that sphere comprises all
manner of topic and question, and is guided by each of the three norms
mentioned earlier, the dispersed assembly that is the public comes as
close as can reasonably be hoped to a free community of equals, autono-
mously debating the terms of their collective life.
35. “We can interpret the idea of the constitutional state in general as the requirement that
the administrative system, which is steered through the power code, be tied to the lawmaking
communicative power and kept free of the illegitimate interventions of social power (i.e., of the
factual strength of privileged interests to assert themselves).” Habermas, Between Facts and
Norms, 150, also 176.
36. Ibid., 360.
218
But the capacity of the public’s critical contributions to steer the state
must remain, given Habermas’s fundamental partition of human action,
an open question. The freer the communication within the public—the
greater the immunities from state interference with the formation of
opinion, the more accessible the newspapers, the less venal the televi-
sion, the richer the associational life on which public discussion rests—
the greater clarification it can attain. Indeed, the call for democratization
of the public sphere, which follows naturally from Habermas’s emphasis
on the role of communicative action in social integration, is exactly the
aspect of his general theory that classes it as a type of radical democracy.37
The basic dualism of understanding and success-oriented action, how-
ever, suggests as well that even the most radical extension of the public
sphere would be of limited consequence precisely because the technical
demands, to which administration, parliament, and party must in turn
respond, set limits—but which ones?—to the direction that might is-
sue from a more encompassing, unrestricted discussion among citizens:
“Communicative power cannot supply a substitute for the systematic in-
ner logic of public bureaucracies. Rather, it achieves an impact on this
logic ‘in a siege-like manner.’”38 At its most paradoxically self-defeating,
Habermas’s view seems to be that the democratic public cannot be just
and effective, because to be just, it must be informal in the sense of con-
stituted freedom of institutions, while to be effective, it must be institu-
tionalized in forms that constrain discussion and hinder the pursuit of
justice. In the end, radical democracy serves as a series of reminders—
that human communication need not be narrowly technical, that un-
solved problems remain outside the purview of conventional institu-
tions—rather than as a program to redirect the ensemble of institutions to
ensure a controlling role for communicative power.39
As a second illustration of the self-limitation of radical democracy, con-
sider the position of Hannah Arendt. Whereas Habermas sets his account
37. Ibid, 371. Though note the immediately subsequent discussion of the “self-limiting”
quality of radical-democratic practice—in particular, the need for a communicatively gener-
ated public opinion to work its effects through conventional political institutions.
38. Habermas, “Further Reflections on the Public Sphere,” in Craig Calhoun, ed.,
Habermas and the Public Sphere (Cambridge, MA: MIT Press, 1992), 452.
39. See especially Habermas’s striking discussion of the “surprisingly active and momentous
role” that actors in civil society can play in a “perceived crisis situation.” Between Facts and
Norms, 7, 380–382. This discussion suggests that Habermas is operating with a distinction be-
tween crisis situations, in which radical democratic impulses play a central role, and normal
politics, in which they do not. For an instructive comparison, see Bruce Ackerman, We, The
People: Foundations (Cambridge, MA: Harvard University Press, 1991).
219
of democracy within a general theory of rational discourse and communi-

cative action, Arendt’s is framed by a general diagnosis of the human con-
dition and a classification of responses to that condition: thus, human
conduct counts as labor if it responds to the rhythmic necessities of bio-
logical reproduction; as work if directed to the construction of those du-
rable artifacts, from houses to highways, that provide the scaffolding and
outward signs of our “unnatural” social life; and as action if it manifests
the fundamental human capacity to begin “something new on our own
initiative.”40 Though acting “rests on initiative, it is an initiative from
which no human being can refrain and still be human.”41 The public in
this view is just the citizens in action, appearing to one another as hu-
man—as a plurality of initiators—and this citizenry in action embodies
democracy in its most radical, constitutional, aspect.42
The dilemma for this view is that democracy continues after its initia-
tion; and on Arendt’s understanding, concerns arising within the con-
stituted polity would fall from the higher, distinctively human sphere of
action and the political to the banausic social spheres of work and labor.
For practical purposes, the public would be purposeless and political de-
bate would be a matter of display: as Mary McCarthy said, “If all ques-
tions of economics, human welfare, bussing, anything that touches the
social sphere, are to be excluded from the political scene, then I am
mystified. I am left with war and speeches. But the speeches can’t just
be speeches. They have to be speeches about something.”43 Thus, if
Habermas’s view edges radical democracy to the periphery, preserving
discursive freedom at the expense of political influence, Arendt’s view
kicks radical democracy upstairs, preserving its free creativity at the ex-
pense of its content.
Recent efforts to modify the idea of the public in both views to re-
spond to these kinds of criticisms by softening the distinctions on which
they rest only underscore the constraints of the original schemes. In
40. Hannah Arendt, The Human Condition (Chicago: University of Chicago Press, 1950),
177.
41. Ibid., 176.
42. Seyla Benhabib emphasizes that “appearing” is to be taken fully literally: “When Arendt
links the public space with the space of appearances she primarily has in mind a model of face-
to-face human interactions,” within a relatively homogeneous community whose shared ethos
makes the meaning of individual action more or less transparent. The Reluctant Modernism of
Hannah Arendt (New York: Sage, 1996), 201.
43. These remarks, made by Mary McCarthy at a conference on Arendt’s work, are reported
in ibid., 155; also see 156 for Arendt’s reply and a convincing assessment of its plausibility.
220
both cases the modifications focus on the role of social movements—of

women, of racial or ethnic minorities, of citizens concerned about the
environment—as forms of the public so dispersed within society to be
acting outside of institutions, hence untainted by technical or workaday
constraints, yet directly enough engaged with changing particular social
arrangements to influence them. In those writings on social movements
that refer, critically, to Habermas and others like him, the emphasis is on
spontaneous citizen protest against the risks of (increasing) technical ma-
nipulation of the social and the natural worlds: the risk of nuclear catas-
trophe calls forth a social movement against the construction of nuclear
power stations, and the industrial use of toxins calls forth a movement for
their regulation. But this simply reminds us that a dispersed, discursive
public can play a part in limiting only the reach of a “system” whose in-
nermost mechanisms remain beyond political influence.44
Writings on social movements that take Arendt as their critical refer-
ent, but that reject her public/private, political/social distinctions, are
more ethereal still. From this perspective, social movements are seen as
the potential precursors to constitutional conventions in the small, or as
interstitial, fleeting, fugitive testimony to the human capacity for initia-
tive and, therefore, to the permanent possibility of a higher politics of
democratic refounding. In both variants, the newer views accept the de-
fensiveness of their antecedents and reduce radical democracy to an ar-
gument about the possibility of protest against the further subjugation of
freedom to necessity.
Directly deliberative polyarchy, in contrast, does not seek refuge in so-
cial movements; it holds out the promise of transforming the institu-
tions of social steering, not merely of containing their erosion.45 It claims
that the apparent limits on the applicability of democratic principles—
44. For thoughtful discussion of debates on social movements from a viewpoint close to the
one adopted here, see Hans Joas, Die Kreativität des Handelns (Frankfurt/Main: Suhrkamp,
1992), 348ff.
45. Jeffrey Isaac concludes his thoughtful discussion of Arendt’s and Camus’ radical demo-
cratic, “rebellious politics” by doubting that “such a politics, centred in civil society, can be
equally effective in fashioning stable democratic economic and political institutions.” He adds
that “if rebellious politics is to be something more than a self-actualizing and self-consuming
phenomenon, then it must challenge and seek to reshape, however cautiously and imperfectly,
existing political institutions.” But then Isaac backs off from this need to reshape—however
cautiously and imperfectly—by identifying “a more institutional kind of politics” with “a more
conventional social democratic politics.” Radical democracy thus remains at the margins. See
Jeffrey Isaac, Arendt, Camus, and Modern Rebellion (New Haven: Yale University Press, 1992),
255, 247, and 258.
221
and the background dualism of creative or freely reasoning public and

banausic life routines—can be overcome by understanding how those
principles can contribute to problem-solving, and how problem-solving
can contribute to the re-interpretation of those principles. At the core of
this mutual re-elaboration, is the idea of deliberation as reason-giving in
context—that is, relative to purposes that concern the citizens. Purpose
does not vitiate deliberation. Rather, it guides and enables deliberation
by suggesting the comparisons and contrasts that give meaning to di-
verse and mutually informative points of view. And purpose here means
purpose of all kinds: the idea of deliberation in directly deliberative
polyarchy does not distinguish between constitutional and operational
tasks. Indeed, the notion of mutual adjustment of means and ends at the
heart of the notion of deliberation in directly deliberative polyarchy—
the very feature that recommends it as a method of problem-solving in
diverse and volatile environments—undercuts the distinction between
these types of activities.
But in stating the contrast between the current, defensive ideas of radi-
cal democracy and the possibilities of directly deliberative polyarchy, we
are advancing our case by signing promissory notes. We have offered
some empirical hints of new institutional developments to warrant our
conjectures, subjected those conjectures to the preliminary tests of inter-
nal consistency, and noted their appeal as alternatives to the despairing
prospects of current debate. If we are right in thinking a new, radically
participatory form of democracy is beginning to stare us in the face, the
obvious and urgent thing to do is to stare back.
222
7
DEMOCRACY AND LIBERTY
1. Main Ideas
The fundamental idea of democratic, political legitimacy is that the au-
thorization to exercise state power must arise from the collective decisions
of the equal members of a society who are governed by that power.1 That
is a very abstract statement of the concept of democracy—as abstract as it
should be. Democracy comes in many forms, and more determinate
conceptions of it depend on an account of membership in the people
and, correspondingly, what it takes for a decision to be collective—autho-
rized by citizens as a body.
Consider two conceptions of democracy, distinguished by their inter-
pretations of the fundamental idea of collective decision: I will call them
aggregative and deliberative. Both views apply in the first instance to insti-
tutions of binding collective decision making,2 and each interprets the
fundamental ideal that such institutions are to treat people bound by
I have presented parts of this essay to the Eastern Division Meetings of the American Philo-
sophical Association, the University of Chicago Workshop on Deliberative Democracy, the
Instituto Universitário de Pesquisas do Rio de Janeiro, the North Carolina Philosophy Collo-
quium, and a seminar with the Graduate Fellows in the Harvard Program in Ethics and the
Professions; in my Wesson Lectures entitled “Liberty, Equality, and Democracy” at Stanford
University, the University of Nebraska-Lincoln, and Princeton University; and as the Wade
Memorial Lecture at St. Louis University. I also discussed these issues in my Fall 1995 political
philosophy seminar at MIT. I am grateful for all the comments and suggestions I have received
and want especially to thank David Austen-Smith, Zairo Cheibub, Alasdair MacIntyre, Carole
Pateman, Adam Przeworski, John Rawls, Elisa Reis, Charles Sabel, T. M. Scanlon, Cass
Sunstein, Iris Marion Young, and Judith Jarvis Thomson. Some of the main ideas are pre-
sented in “Procedure and Substance in Deliberative Democracy,” in Seyla Benhabib, ed., De-
mocracy and Difference: Changing Boundaries of the Political (Princeton: Princeton University
Press, 1996), 95–119 [reprinted as essay 5 in this collection].
1. “Governed by” rather than “affected by.” Democracy is about justifying authority, not
about justifying influence. See Michael Walzer, Spheres of Justice (New York: Basic Books,
1983); Christopher McMahon, Authority and Democracy (Princeton: Princeton University
Press, 1995).
2. I am grateful to T. M. Scanlon for emphasizing the importance of this parallel.
223
collective decisions as equals. According to an aggregative conception of

democracy, then, decisions are collective just in case they arise from
arrangements of binding collective choice that give equal consideration
to—more generically, are positively responsive to—the interests of each
person bound by the decisions.3 According to a deliberative conception,
a decision is collective just in case it emerges from arrangements of bind-
ing collective choice that establish conditions of free public reasoning
among equals who are governed by the decisions. In the deliberative con-
ception, then, citizens treat one another as equals not by giving equal
consideration to interests—perhaps some interests ought to be dis-
counted by arrangements of binding collective choice—but by offering
them justifications for the exercise of collective power framed in terms of
considerations that can, roughly speaking, be acknowledged by all as
reasons.
Because the requirements for free public reasoning among equals are
not narrowly political—not only a matter of the organization of the
state—democracy, on the deliberative view, is not exclusively a form of
politics; it is a framework of social and institutional arrangements that:
1. facilitate free reasoning among equal citizens by providing, for ex-
ample, favorable conditions for expression, association, and partici-
pation, while ensuring that citizens are treated as free and equal in
that discussion; and
2. tie the authorization to exercise public power—and the exercise it-
self—to such public reasoning, by establishing a framework ensur-
ing the responsiveness and the accountability of political power to
it through regular competitive elections, conditions of publicity,
legislative oversight, and so on.
Deliberative democracy, then, is not simply about ensuring a public
culture of reasoned discussion on political affairs, nor simply about fos-
tering the bare conjunction of such a culture with conventional demo-
3. See Robert Dahl, Democracy and Its Critics (New Haven: Yale University Press, 1989).
Dahl holds that the “principle of equal consideration”—which he attributes to Stanley Benn
and which states that the good or interests of each must be given equal consideration—is the
most compelling interpretation of the deeper “idea of intrinsic equality,” according to which
individuals are, for the purposes of collective decisions, to be considered equal (85–86). Dahl
justifies democracy, as a process for making collective decisions, by reference to the principle
of equal consideration, given a “presumption of personal autonomy”: the presumption that in-
dividuals are the best judges and most vigilant defenders of their own interests. See Democracy
and Its Critics, chs. 6–8.
224
D E M O C R A C Y A N D L I B E R T Y
cratic institutions of voting, parties, elections. The idea instead is mani-

festly to tie the exercise of power to conditions of public reasoning: to
establish “all those conditions of communication under which there can
come into being a discursive formation of will and opinion on the part of
a public composed of the citizens of a state”4 and to generate “communi-
cative power”5—an institutionalized impact of that will and opinion on
the exercise of political power.
In the large project of which this essay forms one part, I explore the de-
liberative conception and its implications. Assuming as background a
plurality of comprehensive philosophies of life—the fact of reasonable
pluralism, which I will explain shortly—I aim to show that democracy,
on the deliberative conception, is a substantive, not simply a procedural,
ideal and that the substance comprises egalitarian and liberal political
values. More specifically, I show the central role within a deliberative
conception of democracy of religious, expressive, and moral liberties, po-
litical equality, and an egalitarian account of the common good. Egali-
tarian and liberal political values emerge, then, as elements of democ-
racy rather than as constraints upon it.
In this essay, I focus on the liberties. After sketching the fundamental
notions of reasonable pluralism (Section 2) and deliberative democracy
(Sections 3 and 4), I discuss religious, expressive, and moral liberty, em-
phasizing their essential roles in a democracy as conditions of deliberative
inclusion. The conclusion—on political community and political legiti-
macy—explains how the deliberative conception presents a compelling
interpretation of the democratic ideal.
2. Reasonable Pluralism
I begin with the fact of reasonable pluralism:6 the fact that there are dis-
tinct, incompatible philosophies of life to which reasonable people are
drawn under favorable conditions for the exercise of practical reason. By
a “philosophy of life”—what Rawls has called a “comprehensive doc-
4. Jürgen Habermas, “Further Reflections on the Public Sphere,” in Craig Calhoun, ed.,
Habermas and the Public Sphere (Cambridge, MA: MIT Press, 1992), 446.
5. Ibid., 452.
6. For discussion of this fact, see Joshua Cohen, “Moral Pluralism and Political Consensus,”
in David Copp, Jean Hampton, and John Roemer, eds., The Idea of Democracy (Cambridge:
Cambridge University Press, 1993), 270–291; John Rawls, Political Liberalism (New York: Co-
lumbia University Press, 1996); and Joshua Cohen, “A More Democratic Liberalism,” Michi-
gan Law Review 92, 6 (May 1994): 1502–1546.
225
trine”—I mean an all-embracing view—religious or secular, liberal or

traditionalist—that includes an account of all ethical values and, cru-
cially, provides a general guide to conduct, individual as well as collec-
tive. Let us say that people are reasonable, politically speaking, only if
they are concerned to live with others on terms that those others, as free
and equal, also find acceptable. The idea of reasonable pluralism, then,
is that good-faith efforts at the exercise of practical reason, by reasonable
people thus understood, do not converge on a particular philosophy of
life. Such philosophies are matters on which reasonable people disagree.
The fact of reasonable pluralism is just one of the many forms of hu-
man difference, others being differences of preference and ability, life
chances and biological endowment, ethnicity and rhetorical style. These
differences set a generic task for an account of democracy: to explain
how people, different along so many dimensions, are to be recognized
and treated as free and equal members of a political society (however we
interpret those protean ideas). Though the task is generic, the solution
varies according to dimension, and here—as a matter of focus, and not
prejudging questions of importance—I concentrate on the dimension
captured by the fact of reasonable pluralism.
I said that philosophies of life are matters on which reasonable people
disagree, and I mean that as a platitude. But seeing it as such requires
that we distinguish the fact of reasonable pluralism itself from various
philosophical responses to it. Reflective moral divergence is, for example,
commonly taken to provide strong evidence for the conclusion that no
moral outlook is true, at least not objectively true,7 or that moral truth
transcends our cognitive powers. But persistent divergence of outlook
among reasonable people does not require a nihilist or a relativist expla-
nation. Reasonable people may disagree about a singular truth. Nor does
the absence of convergence command skepticism. For the purposes of
political argument, all we should say in response to the fact of reflective
divergence is that in matters of comprehensive morality, the truth, if
there be such, transcends the exercise of practical reason appropriate to
expect of others, as free and equal.
Five considerations speak in support of the fact of reasonable pluralism
and the force of this response to it. To start with, we observe persistent
7. On the importance of moral diversity as a source of pressures toward relativism and nihil-
ism, see Gilbert Harman, “Moral Relativism,” in Gilbert Harman and Judith Jarvis Thomson,
Moral Relativism and Moral Objectivity (Oxford: Basil Blackwell, 1996), 8–14.
226
disagreements among familiar traditions of ethical thought, each with

its own elaborate structure and complex history of internal evolution—
disagreements, for example, about the relative importance of values
of choice and self-determination, happiness and welfare, and self-
actualization, and about the religious and philosophical background of
these evaluative views. Second, in addition to the sheer fact of disagree-
ment, the exercise of practical reason generates no apparent tendency to
convergence on a philosophy of life. Furthermore, third, no compelling
theory of the operations of practical reason predicts convergence of com-
prehensive moralities after sufficient evidence or reflection or argument
is brought to bear. For moral deliberation, we have nothing comparable
to the economists’ proof of the existence of a general equilibrium of a
competitive market economy with certain minimal optimality proper-
ties—much less an argument for convergence on such an equilibrium.
Nor, fourth, are there any marginally attractive social or political mecha-
nisms that might generate comprehensive agreement. Finally, we can
identify natural explanations for the persistence of disagreement that do
not require accusations of “unreason”: exercise of practical reason of-
ten proceeds within distinct traditions of thought with complex internal
structures; personal circumstance and sensibility predispose people to dif-
ferent traditions; and empirical constraints (including the constraints of
considered evaluative judgments) are typically too weak to swamp such
differences, in part because evaluative concepts themselves are imprecise
and their explications are contested.8
These considerations may appear to prove too much: to create troubles
for the deliberative view, too. For the deliberative view, it will emerge, re-
quires some agreement on political ideas. Why, then, don’t the observa-
tions that support the fact of reasonable pluralism exclude that agree-
ment, too? By way of response, I need to say something more about the
fourth point: the absence of convergence-generating mechanisms in the
case of comprehensive philosophies of life. Consider the difference on
this point between a comprehensive moral consensus and a narrower po-
8. See Rawls’s discussion of the burdens of judgment in Political Liberalism, 54–58. Leif
Wenar argues that these burdens are not widely acknowledged, that, for example, “a religious
doctrine characteristically presents itself as universally accessible to clear minds and open
hearts,” and he cites the Vatican II statement on Divine Revelation in support. See Leif Wenar,
“Political Liberalism: An Internal Critique,” Ethics 106 (October 1995): 32–62. But to say that
certain religious truths require revelation is to acknowledge that they, unlike truths of natural
religion, are not simply available to clear minds and open hearts.
227
litical agreement.9 While it is implausible to expect agreement on politi-

cal values to result from a convergence of practical reasoning conducted
within different, independent moral traditions, it is not so implausible to
expect important elements of political consensus to emerge from the ac-
quisition of ideas and principles embodied in shared institutions. The
acquisition of political ideals and values proceeds in part through partici-
pation in common, public institutions of various kinds—families, associ-
ations, the state. And the formation of moral-political ideas and sensibili-
ties proceeds less by reasoning or explicit instruction (which may be
important in the case of comprehensive moral views) than by mastering
ideas and principles that are expressed in and serve to interpret these in-
stitutions. Thus, people living within institutions and a political culture
shaped by certain ideas and principles are likely to come to understand
those ideas and principles and to develop some attachment to them.
Take the idea of citizens as equal moral persons. This idea is, in several
ways, manifest in the norms and traditions of interpretation associated
with citizenship in a democracy (though practice often fails to conform
to those norms)—for example, equality before the law, and equal civil
and political rights. We can understand how citizens quite generally
might acquire an understanding of one another as moral equals by hold-
ing the position of citizen and living in a political culture in which ideas
of equality associated with that position play a central role in political dis-
course.10 Different comprehensive views that accept this political under-
standing of equality will have different ways of fitting it into their broader
conceptions. Some will accept political equality as following from a fun-
damental religious conviction about human equality or from a general
moral conception of human beings as equally intrinsically valuable; oth-
ers will accept political equality as an important, nonderivative value.11
But what makes agreement possible in this case is that citizens who grow
up within a reasonably stable democracy will find this (self-) conception
familiar and attractive: the political ideas “expressed” in common, public
9. The remarks that follow draw on my “A More Democratic Liberalism.”
10. Consider in this connection the virtually unanimous popular endorsement of political
equality and equality of opportunity indicated in Herbert McClosky and John Zaller, The
American Ethos: Public Attitudes Toward Capitalism and Democracy (Cambridge: Cambridge
University Press, 1985), 74, 83.
11. Thus, Dahl formulates the “principle of equal consideration of interests,” as well as the
“idea of intrinsic equality” on which it rests, to apply solely to processes for making binding
collective decisions, and indicates that both might be adopted by adherents to religious morali-
ties, as well as utilitarian and Kantian moral views. See Democracy and Its Critics, 85–87.
228
institutions and appealed to in the culture to justify those institutions will

shape the moral-political education of citizens.
Of course, the acquisition of moral ideas does not proceed exclusively
through institutions. So citizens will need to be able to accommodate the
political ideas and the self-conceptions they acquire through institutions
within their different philosophies of life: to find a way to combine, for
example, a conception of human beings as servants of God bound by
natural duties with a political conception of citizens as free, equal, and
self-governing. And many views—religious, moral, philosophical—have
sufficient internal flexibility, or openness, to make such accommodations
possible.12 But while this accommodation may take place when a com-
prehensive moral or religious view is elaborated in ways that make it
compatible with a political conception, we have no reason to expect it to
produce agreement extending beyond political values; for no institu-
tional mechanism in a democratic society imposes pressure to reach
agreement in ways that would erase fundamental differences between
moral, religious, and philosophical traditions. The pressure of the shared
institutions in explaining political agreement ends even as considerable
disagreement remains.
In short, political values are institutionalized in a democratic society in
ways that comprehensive moral (or religious, or philosophical) ideas are
not. More precisely, comprehensive ideas are sustained through more
particular social associations that are not shared: different churches, for
example, advance different comprehensive views. So citizens acquire
12. Consider, for example, the changes in the Catholic doctrine on toleration that emerge
in Vatican II. The idea of human dignity, always a central element in Catholic moral and so-
cial thought, is developed along new lines as the basis for an account of political legitimacy
with principled limits on the state’s authority in matters of religious faith and practice. See The
Revised Documents of Vatican II, “Declaration on Religious Freedom,” 1.2, ed. Austin Flannery
(Northport: Costello, 1976), 800–801. Dignity imposes an obligation to seek the truth and em-
brace it. But though the “one true religion subsists in the Catholic and apostolic Church,” the
pursuit and embrace of truth must comport with our nature as free beings “endowed with rea-
son” and the dignity owing to that nature. And this requires immunity from “external coercion”
as well as from “psychological freedom.” The introduction to the Declaration ties the force of
the sense of dignity and an understanding of its implications to modern experience (the “con-
sciousness of contemporary man”). In an interesting essay on modern Confucian humanism
and human rights, Tu Wei-ming suggests a way to reinterpret Confucian doctrine as incorpo-
rating a conception of dignity, tied to obligations in social relationships, that could serve in
turn as a basis for a conception of human rights. Here, too, the conception of dignity is tied to
central Confucian notions, but its formulation is prompted by modern political sensibilities, as
articulated through international institutions. See Tu Wei-ming, “A Confucian Perspective on
Human Rights,” unpublished, 1995.
229
conflicting comprehensive views through those associations. An account

of how consensus might emerge on political values among citizens living
in a political society, then, has resources unavailable to an account of a
more comprehensive moral consensus.
Despite these considerations, one might still hold out hope for com-
prehensive moral agreement and for a political community based on
such agreement. In response to the observation that there is no tendency
to reach agreement, we might, for example, take certain points of moral
convergence—on the injustice of slavery or on the value of religious tol-
eration—as at least suggesting a broader tendency toward moral conver-
gence.
Moreover, a political society with comprehensive moral agreement has
at least four attractive qualities that might make it a society to be hoped
for. In such a society, for example, members respect one another for their
determinate, concrete commitments—for the values that animate their
lives—and not simply for their abstract though perhaps unrealized hu-
man capacities, not merely for their inner, but perhaps unexpressed, dig-
nity. Furthermore, all may believe the truth, whereas moral pluralism
implies that some members are bound to endorse false moral ideas; even
if we accept that false views have public rights,13 still, endorsing the truth
is an important human good; and endorsing it in common, deepens mu-
tual respect. In addition, agreement gives the members of a society a con-
fidence in the bases of their conduct that is plausibly lacking under con-
ditions of disagreement, thus mitigating pressures to reject the objectivity
of ethical thought and embrace skepticism, nihilism, or simple alien-
ation. Finally, as a practical matter, the absence of conflict on compre-
hensive views may ease communication and coordination.
In response to the proposed extrapolation from cases of convergence—
slavery and religious toleration—it must be said that these are not cases
in which people agree on comprehensive moral outlooks. Instead, peo-
ple who belong to different moral and religious traditions come to agree
on the injustice of certain especially injurious practices, despite their per-
sistent disagreement on other matters. Thus, the condemnation of slavery
is common ground among Catholics, Muslims, Jews, and Protestants;
and among Kantians and utilitarians. So the examples are not strong evi-
13. For critical discussion of the idea of the “exclusive rights of truth,” see John Courtney
Murray, “The Problem of Religious Freedom,” in J. Leon Hooper, S.J., ed., Religious Liberty:
Catholic Struggles with Pluralism (Louisville: Westminster John Knox Press, 1993), ch. 2.
230
dence for the thesis that practical reason generates a more general moral
convergence among people working within different moral and religious
traditions; rather, they are important cases of agreement on political val-
ues among people who have fundamental moral disagreements.
Moreover, let us suppose that comprehensive agreement has impor-
tant virtues. Still, if the exercise of practical reason does not generate
convergence among people who begin with very different outlooks, asso-
ciated with different traditions of moral, religious, and philosophical
thought and practice, then how is comprehensive moral agreement to be
achieved? Perhaps through some form of common moral education in a
comprehensive view. But how is such education—as distinct from educa-
tion about requirements of civic responsibility and decency—to proceed
in the face of different and competing forms of moral thought, each with
its own associational life? Perhaps through the state’s coercive means. But
it is not so clear that the state can produce genuine moral agreement, as
opposed to public spectacles of conformity; and if it could, the price
seems unacceptable, despite the values associated with agreement.
I assume the fact of reasonable pluralism, then. And this fact gives
shape to the conception of citizens as free and equal that constitutes part
of the deliberative conception of democracy I want to explore here. To
say that citizens are free is to say, inter alia, that no comprehensive moral
or religious view provides a defining condition of membership or the
foundation of the authorization to exercise political power. Not that reli-
gious or moral views are, religiously, morally, or metaphysically speaking,
matters of choice. To someone who has a religious view, for example, be-
lieving the view is a matter of believing what is true, and acting on it, a
matter of fulfilling obligations that are not self-legislated and are perhaps
more fundamental than political obligations.14 But politically speaking,
citizens are free in that it is open to them to accept or reject such views
without loss of status. To say citizens are equal is to say that each is recog-
nized as having the capacities required for participating in discussion
aimed at authorizing the exercise of power.
What, more particularly then, can we say about a conception of de-
14. Michael McConnell says, “It would come as some surprise to a devout Jew to find that
he has ‘selected the day of the week in which to refrain from labor’ since the Jewish people
have been under the impression from some 3,000 years that this choice was made by God.”
“Religious Freedom at a Crossroads,” University of Chicago Law Review 59 (1992): 115. The
source of the quotation to which McConnell is responding is Estate of Thornton v. Caldor,
Inc., 472 U.S. 703, 711 (1985) (O’Connor, J., concurring).
231
mocracy suited to conditions of reasonable pluralism? By excluding a

comprehensive consensus on values, the fact of reasonable pluralism
may suggest that a procedural conception of democracy, limited to such
values as openness and impartiality associated with fair process, is the
only remaining option. After all, that fact deprives us of a background of
shared moral or religious premises—shared reasons—that would give
more determinate content to the idea of popular authorization. Without
that background, we are left, it may seem, with no basis for agreement on
anything more than fair procedures—perhaps not even with that. Faced
with disagreement on comprehensive views, what legitimate complaint
can a person raise about a framework of collective decision, beyond the
complaint that the framework fails to take her interests into account?
I think this conclusion is incorrect and will sketch a view that com-
bines an assumption of reasonable pluralism with a more substantive
conception of democracy. I will suggest as well that this combination is a
natural result of adopting a deliberative understanding of the collective
decisions that constitute democratic governance.
3. Public Reasoning
A deliberative conception of democracy puts public reasoning at the cen-
ter of political justification. I say public reasoning rather than public dis-
cussion because a deliberative view cannot be distinguished simply by its
emphasis on discussion rather than bargaining or voting as methods of
collective decision making. On any view of democracy—indeed any
view of intelligent political decision making—discussion is important, if
only because of its essential role in pooling private information, against a
background of asymmetries in its distribution.
According to the deliberative interpretation of democracy, then, de-
mocracy is a system of social and political arrangements that institution-
ally ties the exercise of power to free reasoning among equals. This con-
ception of justification through public reasoning can be represented in
an idealized procedure of political deliberation, constructed to capture
the notions of free, equal, and reason that figure in the deliberative ideal.
The point of the idealized procedure is to provide a model characteriza-
tion of free reasoning among equals, which can, in turn, serve as a model
for arrangements of collective decision making—arrangements that es-
tablish a framework of free reasoning among equals. Using the model, we
can work out the content of the deliberative democratic ideal and its con-
232
ception of public reasoning by considering features of such reasoning in

the idealized case and then aiming to build those features into institu-
tions.
Thus, in an ideal deliberative procedure, participants are, and regard
one another, as free: recognizing the fact of reasonable pluralism, they ac-
knowledge, as I noted earlier, that no comprehensive moral or religious
view provides a defining condition of participation or a test of the accept-
ability of arguments in support of the exercise of political power. More-
over, participants regard one another as formally and substantively equal.
They are formally equal in that the rules regulating the ideal procedure
do not single out individuals for special advantage or disadvantage. In-
stead, everyone with the deliberative capacities—which is to say, more or
less all human beings—has, and is recognized as having, equal standing
at each stage of the deliberative process. Each, that is, can propose issues
for the agenda, propose solutions to the issues on the agenda, offer rea-
sons in support of or in criticism of proposed solutions. And each has an
equal voice in the decision. The participants are substantively equal in
that the existing distribution of power and resources does not shape their
chances to contribute to deliberation, nor does that distribution play an
authoritative role in their deliberation. In saying that it does not play an
authoritative role in their deliberation, I mean that the participants in the
deliberative procedure do not regard themselves as collectively morally
bound by the existing system of rights, except insofar as that system estab-
lishes the framework of free deliberation among equals. Instead, they re-
gard that system as a potential object of their deliberative judgment.
In addition, they are reasonable in that they aim to defend and criticize
institutions and programs in terms of considerations that others, as free
and equal, have reason to accept, given the fact of reasonable pluralism
and on the assumption that those others are themselves concerned to
provide suitable justifications.
Which considerations count as reasons? Generically speaking, a rea-
son is a consideration that counts in favor of something: in particular, a
belief or an action. Not an illuminating analysis: I doubt that illuminat-
ing analysis is available or that it would be helpful in answering our ques-
tion.15 What is needed is an account not of what a reason is, but of which
considerations count as reasons. And the answer to this question depends
15. Here I follow discussion in T. M. Scanlon’s What We Owe to Each Other (Cambridge,
MA: Harvard University Press, 1997), ch. 1.
233
on context. Whether considerations count in favor in the relevant way

depends on the setting in which they are advanced. Applying this point to
the issue at hand: a suitable account of which considerations count as
reasons for the purposes of an account of democratic deliberation will
take the form not of a generic account of what a reason is but of a state-
ment of which considerations count in favor of proposals within a de-
liberative setting suited to the case of free association among equals, un-
derstood to include an acknowledgment of reasonable pluralism. This
background is reflected in the kinds of reasons that will be acceptable—
meaning, as always, acceptable to individuals as free and equal citizens.
I have already specified the relevant deliberative setting as one in
which people are understood as free, equal, and reasonable and as having
conflicting philosophies of life. Within the idealized deliberative setting
that captures these conditions, it will not do simply to advance consider-
ations that one takes to be true or compelling. Such considerations may
well be rejected by others who are themselves reasonable—prepared to
live with others on terms acceptable to those others, given their different
comprehensive views—and endorse conflicting philosophies of life. One
needs instead to find reasons that are compelling to others, where those
others are regarded as (and regard themselves as) equals and have diverse
reasonable commitments. How wide a range of commitments? Because
we are addressing the institutional framework for making collective deci-
sions and assume the participants to be free—not bound to their de facto
commitments—the range of commitments is similarly wide: not ex-
hausted by de facto commitments. Considerations that do not meet these
tests will be rejected in the idealized setting and so do not count as ac-
ceptable political reasons. Let us say, then, that a consideration is an ac-
ceptable political reason just in case it has the support of the different
comprehensive views that might be endorsed by reasonable citizens.
To illustrate these points about the role of the background—the con-
ception of citizens as free, equal, and reasonable—in constraining the set
of reasons, let us consider three implications; the first two will be particu-
larly important in my discussion of the liberties.
First, people hold religious commitments on faith, and those commit-
ments impose what they take to be overriding obligations. Such commit-
ments are not, as such, unreasonable: though faith transcends reason,
even as “reason” is understood within the tradition to which the commit-
ments belong, citizens are not unreasonable for holding beliefs on faith.
234
But beliefs held on faith—perhaps beliefs in what are understood to be

revealed truths—can reasonably be rejected by others who rely only on
the darkness of an unconverted heart and so cannot serve to justify legis-
lation.
Thus, it matters to our response to the case against abortion stated in
the encyclical, Evangelium Vitae, that Pope John Paul II claims that the
gospel of life “can . . . be known in its essential traits by human reason”
and that the “Law of God” that condemns abortion is “written in every
human heart [and] knowable by reason itself”—in short, that the argu-
ment is presented as independent of a particular faith position.16 Though
I see no reason to agree with this claim about what lies within the com-
pass of reason, our response to the argument must be different from what
that response would be if the argument appealed openly to revealed
truths or beliefs held on faith. We must show that the conception of rea-
son it appeals to is itself sectarian and that the argument fails on a con-
ception of reason that is not.
Second, the adequacy of a consideration as a political reason—its
weight in political justification—will depend on the nature of the regu-
lated conduct, in particular on the strength of the reasons that support
that conduct. Thus, considerations of public order provide acceptable
reasons for regulating conduct. Different views have different ways of ex-
plaining that value: utilitarians will found it on considerations of aggre-
gate happiness, Kantians on the preconditions of autonomous conduct,
others on the intrinsic value of human life and human sociability. More-
over, people are bound to disagree about the requirements of public
order: that disagreement may extend to whether a state is necessary to
secure the conditions of order. But it will not be acceptable to suppose
that, as a general matter, the value of public order transcends all other
political values. Except perhaps in the most extreme circumstances, for
example, a state may not impose a blanket prohibition on alcohol con-
sumption—including consumption in religious services—in the name of
public order. The reasons that support such consumption include con-
siderations of religious duty—more generally, considerations of funda-
mental duty, which are normally overriding. And those considerations
16. Pope John Paul II, Evangelium Vitae (New York: Times Books, 1995), §§29, 62. For criti-
cal discussion of these claims and the argument based on them, see Judith Jarvis Thomson,
“Abortion,” Boston Review 20, 3 (Summer 1995): 11–15.
235
will provide a suitable basis for rejecting a justification cast in terms of

the value of public order, except in the most extreme conditions.
These first two points about reasons both generate pressure for liberty.
The first point underscores that reasonable pluralism will lead to the re-
jection of some bases for restricting liberty as politically weightless; the
second indicates that other bases of restriction will not be weightless, but
they will be insufficient to outweigh reasons that can be acknowledged,
consistent with reasonable pluralism, as commending or commanding a
course of conduct.
Finally, third, the fact that a policy is most beneficial to me arguably
provides me with a reason to support that policy.17 But this reason carries
no weight in public deliberation of the relevant kind, because others,
concerned with their advantage and willing to find mutually acceptable
reasons, will not accept it as a reason; moreover, it is reasonable for them
not to accept it, in part because they can dismiss it while at the same time
treating me as an equal and giving my good the same weight in their de-
liberations that they insist I give to theirs. Thus, I may prefer the arrange-
ment that gives me the greatest advantage, and so have a personal reason
for promoting it. But in the context of ideal deliberation, I must find con-
siderations in favor of the arrangement that do not neglect the good of
others. Similarly, I may wish to reject an arrangement that leaves me less
well-off than some others. But I cannot offer as a reason against it that it
leaves me less well-off, because every arrangement will leave some peo-
ple less well-off than some others. So if I need to find reasons acceptable
to others, I cannot reject a proposal simply because it does to me what
each arrangement must do to someone—and, again, every arrangement
leaves some less well-off than others. I can, however, reject it on grounds
that an arrangement leaves me less well-off than anyone needs to be.
In presenting the deliberative view in terms of an ideal deliberative
procedure in which parties are required to find reasons acceptable to oth-
ers, I may appear to be tying the deliberative conception to an implausi-
ble requirement of political consensus—to the view that “deliberation
leads to convergence.”18
I make no such assumption. Instead, I assume that different views will
17. I say “arguably” because it might be said that reasons are essentially public and capable
of being shared. If that is right, then the fact stated in the text would not constitute a reason.
18. See Adam Przeworski, Democracy and the Market (Cambridge: Cambridge University
Press, 1991), 17.
236
have different interpretations of the acceptable reasons and of how differ-

ent reasons are to be weighted—for example, reasons of equality and of
aggregate well-being. As a result, even an ideal deliberative procedure
will not, in general, issue in consensus. But even if there is such disagree-
ment, and a need to submit the decision to majority rule, still, partici-
pants in the ideal case will need to appeal to considerations that are quite
generally recognized as having considerable weight and having a suitable
basis for collective choice, even among people who disagree about the
right result: agreement on political values is not agreement on the proper
combination of them. But when people do appeal to considerations that
are quite generally recognized as having considerable weight, then the
fact that a proposal has majority support will itself commonly count as a
reason for endorsing it. Even people who disagree may, then, accept the
results of a deliberative procedure as legitimate.
4. Discussion, Deliberation, Motivation

I said earlier that a deliberative conception of democracy cannot be char-
acterized by its emphasis on discussion, that any view of democracy will
have an important place for discussion because of its essential role in
pooling dispersed, private information. Of course, discussion is not al-
ways so helpful. As Przeworski puts it, “If people behave strategically in
pursuit of their interests, they also emit messages in this way”:19 behavior
does not lose its strategic character simply because it involves the use of
language. And if people “emit messages”—that is, communicate—strate-
gically, they may well have incentives to misrepresent private informa-
tion, in which case discussion may play an essential role in creating de-
ception and spreading misinformation. The mere fact that conduct is
linguistically mediated does not, of course, imply that agents are pre-
pared to constrain their conduct by referring to norms of honesty, sincer-
ity, and full disclosure rather than simply taking the most effective means
to their ends. The use of language may, as a matter of conversational
implicature, commit the speaker to such norms—to endorsing them as
proper standards of guidance and criticism. But I don’t propose to rest
anything on this hypothesis.
Though the strategic use of language to advance one’s aims always car-
ries the potential of deviating from norms of honesty, the strength of in-
19. Ibid.
237
centives to engage in misrepresentation depends, inter alia, on the un-

derlying diversity of citizen preferences. Intuitively, the more diverse the
preferences of individuals—the more they disagree about the best out-
come—the greater are the gains from strategically lying, misinforming,
or distorting; the greater the gains, the greater are the temptations to
undertake such manipulation.20 The point is familiar in the setting of
legislative decision making in legislatures with committee structures.
The more extreme the preferences of committee members relative to
median legislative preferences (the more they are outliers), the less in-
formative the committees will be (more noise, less signal). For this
reason, majoritarian legislatures will not—unless they have a collec-
tive preference for self-deception—want to leave important decisions
in the hands of a committee composed of preference outliers: mem-
bers will not expect the committee to provide truthful and complete in-
formation.
But—and this is the crucial point—the extent of preference diversity
is not fixed, not given in advance of political deliberation. Not that the
aim of such deliberation is to change citizen preferences by reducing
their diversity: the aim is to make collective decisions. Still, one thought
behind a deliberative conception is that public reasoning itself can help
to reduce the diversity of politically relevant preferences, because such
preferences are shaped and even formed in the process of public reason-
ing itself. And if it does help to reduce that diversity, then it mitigates ten-
dencies toward distortion even in strategic communication.
Two points are essential here, one concerning reasons, the other con-
cerning connections between reasons and motives. First, the reasoning
that figures in collective decisions need not be exclusively instrumen-
tal—only a matter of determining the most effective means for achieving
settled aims, given perhaps by desires. Indeed, practical reasoning—un-
derstood as reflection on and discussion about what reasons for action
agents have—may proceed along deliberative paths with only the most
20. See Keith Krehbiel, Information and Legislative Organization (Ann Arbor: University of
Michigan Press, 1991), 81–84, 95–96. Apart from this dependence on preference diversity, the
effectiveness of speech depends on the ease of verifying information and on whether discus-
sion proceeds sequentially and in public. See David Austen-Smith, “Strategic Models of Talk
in Political Decision-Making,” International Political Science Review 13, 1 (1992): 45–58; on
economic institutions that promote verification and sequential, public conversation, see
Charles Sabel, “Learning by Monitoring: The Institutions of Economic Development,” in
Neil Smelser and Richard Swedberg, eds., The Handbook of Economic Sociology (Princeton:
Princeton University Press, 1995), 137–165.
238
attenuated connections to the agent’s current aims.21 Citizens are capa-

ble of recognizing as reasons considerations that conflict with their ante-
cedent preferences and interests, ranking alternatives in accordance with
such considerations, and acting on those rankings. I might now recog-
nize that I have good reason to refrain from harming others but not think
that I have reason ever to help them, nor have any desire at all to help.
Suppose, however, that reflection on why I have reason not to harm leads
me to see that the explanation for that reason also implies that I have rea-
son to help. Though I have no desire to help, a new reason emerges from
a search for reflective equilibrium that, as in this example, proceeds by
considering the justification for settled reasons and the implications of
the justification for other reasons that agents have. To the extent that I
also have preferences that conflict with these reasons, I will continue to
have incentives—perhaps strong ones—to strategically misrepresent in-
formation. But, coming now to the second point, seeing that certain of
my antecedent preferences and interests cannot be expressed in the form
of acceptable reasons may help to limit the force of such preferences as
political motives.22
To illustrate: assume a commitment to deliberative justification—as-
sume, that is, the shared belief that political justification requires finding
reasons acceptable to others, understood as free and equal, who endorse
that commitment. And assume, alongside that, a desire that others serve
my aims, regardless of their own.23 Though this desire may prompt me to
advance a proposal, it does not count as a reason in public argument. To
defend the proposal consistent with my commitment to finding justifica-
tions of the appropriate kind, then, I need to advance reasons indepen-
dent of the desire—which I will be prepared to do only if I believe that
there are acceptable reasons. And presenting such reasons may lead to
the formation of a new desire, say a desire to coordinate with others on
mutually beneficial terms. Merely believing that I have such a reason
21. For discussion, see Christine Korsgaard, “Skepticism about Practical Reason,” Journal of
Philosophy 83, 1 (January 1986): 5–25; Scanlon, What We Owe to Each Other, ch. 1. For illumi-
nating criticism of instrumental rationality, see Robert Nozick, The Nature of Rationality
(Princeton: Princeton University Press, 1993), ch. 5. Particularly important for purposes here
are the remarks on the symbolic utility of acting on principles.
22. Using Nozick’s terminology (see note 21), when I see that a proposal cannot be defended
with acceptable reasons, its symbolic utility declines. Assuming that symbolic utility is
motivationally important, the motivation for advancing the proposal declines.
23. I am not worrying here about distinctions between “desire” and “prefer,” in particular
that the latter is a polyadic relation.
239
may suffice to refashion preferences, but the motivational force of that

recognition is likely to be greater if I must state the reasons, thereby lend-
ing greater salience to them. That desire, unlike the desire that others
serve my aims regardless of their own, is naturally expressed in a reason
that is acceptable to others. Moreover, if I develop the desire to cooperate
on mutually beneficial terms, my incentive to strategically misrepresent
private information will decline. And that means that even if my commu-
nications are in part strategically motivated, I will be more likely to pro-
vide information that is commonly beneficial.
It should be clear, but it is nevertheless worth emphasizing, that the
preference changes with beneficial effects on strategic communication
are not simply changes of induced preference that result from the acquisi-
tion of new information through discussion.24 Of course, new informa-
tion may well induce new preferences: I now prefer eating bread to eat-
ing cheese because I believe that bread is more nutritious and prefer
more nutritious to less nutritious food; if I learn that cheese is more nutri-
tious, and I am rational, I will prefer cheese to bread. And sometimes dis-
agreement among preferences is generated by simple differences of fac-
tual belief. Reducing differences owing to lack of common information
will often be a good thing. But the kinds of preference changes I am con-
templating reflect a sensitivity of motivations—understood as behavioral
dispositions—to reasons, understood as standards of criticism and guid-
ance, and not simply a sensitivity of some preferences to information
about how most effectively to satisfy other preferences.
Though these are not cases of induced preference change, they are
also not cases (like hypnotism or suggestion) of preferences changing
without rational explanation. In the kinds of cases I am considering, pref-
erences change because a person comes to understand—through practi-
cal reasoning—that his current preferences lack appropriate justification:
not because new empirical knowledge is acquired that bears on the
achievement of an aim (as in the bread to cheese example), but because
the preference cannot be supported by reasons of a suitable kind, the
agent recognizes that it cannot, and that recognition has sufficient sa-
lience to shape motivations. In the background is the view that the no-
24. On the distinction between primitive and induced preferences, and a case for the view
that deliberation-induced preference change is a matter of changes in induced preferences as a
result of new information, see John Ferejohn, “Must Preferences Be Respected in a Democ-
racy?,” in Copp, et al., eds., The Idea of Democracy, 236–237.
240
tion of a reason is essentially normative—a term of justification and of

criticism—and that a reason is not a kind of motivation. Practical reason-
ing, then, is a matter of reflecting on what one is to do, not what one is
motivated to do, though the results of such reasoning can motivate.25
5. Religious Liberty
I have focused thus far on the structure of the deliberative view. I turn
now to its substance, in particular to the thesis that democracy—on the
deliberative interpretation of collective choice—must ensure religious,
expressive, and moral liberties. This proposal departs from conventional
understandings of the relationship between democracy and these liber-
ties. To illustrate that understanding, I want to present a familiar di-
lemma associated with the idea of democratic legitimacy.
On the one hand, the value of democracy seems too procedural to pro-
vide a basis for an account of legitimacy; some democratic collective
choices are too repulsive to be legitimate, however attractive the proce-
dures that generate them. On the other hand, the idea of democracy ap-
pears to be the authoritative, sovereign requirement of collective deci-
sions. That is because democracy appears to be the form of collective
choice mandated by the fundamental political idea that citizens are to be
treated as equals. Because the ideal of treating people as equals is so fun-
damental, and so intimately linked to democratic procedures of binding
collective decision making, democracy is naturally identified not simply
as one political value to be combined with others but also as the way we
must settle the ordering of other political values—the way to ensure
equal standing in settling the common environment. To put issues off
the democratic agenda appears, by contrast, to establish objectionable
spheres of privilege. Thus, Robert Dahl says:
It seems to me highly reasonable to argue that no interests
should be inviolable beyond those integral or essential to the
democratic process. . . . [O]utside this broad domain [which in-
cludes rights of political expression, participation, and associa-
tion] a democratic people could freely choose the policies its
members feel best; they could decide how best to balance free-
25. For suggestive discussion of the role nonstrategic reasons can play in constraining discus-
sion and in improving its effectiveness, see James Johnson, “Is Talk Really Cheap? Prompting
Conversation Between Critical Theory and Rational Choice,” American Political Science Re-
view 87, 1 (1993): 74–86.
241
dom and control, how best to settle conflicts between the inter-
ests of some and the interests of others, how best to organize
and control their economy and so on. In short, outside the invi-
olable interests of democratic people in the preservation of the
democratic process [inviolable because of the roots of that pro-
cess in an ideal of equal intrinsic worth] would lie the proper
sphere for political decisions.26
Dahl immediately indicates some qualms about this view and explores
ways to ensure that conventional democratic process might better protect
fundamental interests that are not integral or essential to it.27 But he has
identified a genuine problem, whose most familiar expression arises in
connection with what Benjamin Constant called the “liberties of the
moderns”—religious liberty, liberty of conscience more generally, liberty
of thought and expression, rights of person and personal property. These
liberties lack any evident connection to conditions of democratic proce-
dure: to borrow Dahl’s words, they are neither integral nor essential to it.
So their protection is commonly understood as constraining democratic
process—limiting its appropriate scope. In that respect they differ from
political liberties, including rights of association, of speech on political
questions, and of participation. If a constitution disables a majority from
restricting political participation or regulating the content of political
speech, that constitution can be interpreted as safeguarding the essen-
tials of democratic process. Assurances of such political liberties help
to ensure a connection between popular authorization and political
outcome—to preserve the continuing authority of the people, and not
simply the majority of them.28 Those liberties—the liberties of the an-
cients—are constitutive elements of democratic process.
The liberties of the moderns appear, then, to be based on independent
26. Dahl, Democracy and Its Critics, 182.
27. Dahl has long been skeptical about the role of courts with powers of judicial review in
providing such protection. See his remarkable essay, “Decision-Making in a Democracy: The
Supreme Court as a National Policy-Maker,” Journal of Public Law 6 (Fall 1957): 279–295; and
Democracy and Its Critics, chs. 12, 13.
28. See John Hart Ely, Democracy and Distrust (Cambridge, MA: Harvard University Press,
1980); Dahl, Democracy and Its Critics; and, more generally, on constitutional requirements
as enabling democracy, Stephen Holmes, “Precommitment and the Paradox of Democracy,”
in Jon Elster and Rune Slagstad, eds., Constitutionalism and Democracy (Cambridge: Cam-
bridge University Press, 1988), esp. 195–240; and Samuel Freeman, “Original Meaning, Demo-
cratic Interpretation, and the Constitution,” Philosophy and Public Affairs 21 (Winter 1992):
3–42.
242
values, separate from the ideal of treating people as equals in arenas of

collective choice that underlies the appeal of democracy. And that may
suggest that, from the perspective of democratic thought, these liber-
ties have roots no deeper than contingent popular consensus. Though
abridgments of nonpolitical liberties that emerge from a fair democratic
process may be unjust, they face no problems of democratic legitimacy.29
On the deliberative conception of democracy, this conclusion is
wrong: a deliberative view provides a basis for wider guarantees of basic
liberties. The explanation of this feature is that the deliberative con-
ception requires more than that the interests of all be given equal con-
sideration in binding collective decisions; it requires, too, that we find
politically acceptable reasons—reasons acceptable to others, given a
background of reasonable differences of conscientious conviction. I call
this requirement the principle of deliberative inclusion.30
To illustrate the roots and implications of this principle, I want to start
with the case of religious liberty, one of the principal liberties of the mod-
erns and the one that most sharply illustrates the analytical structure.
As I mentioned earlier, religious views set demands of an especially
high order—perhaps transcendent obligations—on their adherents.
Moreover, if we see these requirements from the believer’s standpoint,
we cannot see them as self-imposed—chosen by the agent. To put the
point without benefit of ocular metaphor: if we believe about these re-
quirements (say, as to day and manner of worship) what the adherent be-
lieves about them, then we do not believe that the adherent chooses to
place herself under these demands. The content and stringency of the
demands are fixed instead by the content of the convictions, which the
adherent believes true, not by the adherent’s endorsement of those con-
victions. To be sure, if a believer did not endorse the convictions, then
she would not believe herself to be bound by them: but given that she
does endorse them, she thinks that she would then hold false beliefs and
would be more likely to do what is wrong.
29. This is, I believe, Dahl’s view. Critics of Roberto Unger’s conception of empowered de-
mocracy have (mistakenly, I believe) assumed that he endorses it. See his False Necessity
(Cambridge: Cambridge University Press, 1987), 508–539. And it bears a strong family resem-
blance to the democracy-based interpretations of the U.S. Constitution advanced by Ely in De-
mocracy and Distrust; and Bruce Ackerman in We, the People (Cambridge, MA: Harvard Uni-
versity Press, 1991), esp. ch. 1.
30. Of course, not all differences of conviction are reasonable. One implication is that the
problem of toleration for the intolerant is a separate issue in an account of religious liberty. Re-
ligious liberty generally ought not to be treated as a response to a problem of unreasonableness.
243
Liberal political conceptions are sometimes said to endorse, if only by

way of implicit commitment, a conception of human beings as “bound
only by ends and roles we choose” and correspondingly to deny “that we
can ever be claimed by ends we have not chosen—ends given by God,
for example, or by our identities as members of families, peoples, cul-
tures, or traditions.”31 Liberalism, on this view, rests on moral volunta-
rism. And such voluntarism implies that religious moralities are false. It
is difficult, I think, to find political conceptions that embrace uncompro-
mising moral voluntarism—that conceive of human beings as, in Mi-
chael Sandel’s compelling phrase, “unencumbered selves.” In any case,
that philosophy of life cannot possibly serve as common political ground.
And once it is rejected as such, we see that reasonable adherents cannot
accept, as sufficient reasons in support of a law or a system of policy, con-
siderations that would preclude their compliance with fundamental reli-
gious demands or require that they treat those demands as matters of
choice.
What, then, of citizens who do not share those views, who reject them
as false—or, perhaps, as meaningless? (I will describe the issue from the
point of view of a citizen who has fundamental moral convictions but not
religious convictions. Broadly parallel remarks could be made from the
standpoint of different religious convictions.) They might respond in one
of three ways.
First, they might regard all religious views that impose such stringent
demands, whatever their content and foundation, as unreasonable. This
response might issue from the conviction that all religious views are intol-
erant, and for that reason politically unreasonable, or that religious con-
victions cannot withstand reflective scrutiny. But neither of these views is
acceptable. The first is simply false. Nothing in religious conviction it-
self—any more or less than in secular moral conviction—requires en-
dorsement of the view that “error has no rights” or that truth suffices for
justification. As to the second, there may be conceptions of “reflective
scrutiny” on which religious views cannot withstand such scrutiny. But
those conceptions themselves are almost certain to belong to compre-
hensive views (say, empiricist philosophies) that cannot be permitted to
set the bounds on public reasoning, any more than can natural theology’s
conception of natural reason. Moreover, any account of reflective scru-
31. Michael Sandel, Democracy’s Discontent (Cambridge, MA: Harvard University Press,
1996), 322.
244
tiny that condemns religious conviction will almost certainly condemn

secular moral ideas as equally unreasonable.
A second possibility is to treat concerns to fulfill religious obligations as
intense preferences, to be given equal consideration along with other
preferences of equal intensity. This response urges us to put aside the
content of the convictions and their special role—as first principles of
practical justification—in practical reasoning. The roots of this response
lie, I believe, in a misinterpretation of the value of neutrality. Neutrality
requires that political justification in a democracy not depend on any
particular reasonable view. But it does not require that we neglect the
content of views, treat all as matters of mere preference, and let the
strength of claims be fixed by the intensity of those preferences.32 Doing
so indicates a failure to take into account the special weight of reli-
gious or fundamental moral convictions to the adherent, particularly the
weight of requirements that the religious or the moral outlook itself desig-
nates as fundamental demands: an unwillingness to see how the adher-
ent’s convictions, in virtue of their content, state or imply that the re-
quirements provide especially compelling reasons, and not simply strong
preferences.
But if we are not prepared to treat convictions as (without qualifica-
tion) self-imposed, nor to accept them as true, nor to dismiss them as
false, or—putting their truth or falsity to the side—nor to let their weight
be fixed by their intensity as preferences, what is left? The alternative is to
take seriously that the demands impose what the adherent reasonably re-
gards as fundamental obligations (paradigmatically compelling practical
reasons), to accept the requirement—associated with the deliberative
view—of finding reasons that might override these obligations, and to
acknowledge that such reasons cannot normally be found.33 The result is
religious liberty, understood to include freedom of conscience, which
condemns disabilities imposed on grounds of religious belief, and free
exercise of religion, which condemns, in particular, limits on public
worship.34 It emerges as the product of three elements. The first is the
32. See Scanlon’s suggestion that proponents of subjective criteria for interpersonal compar-
isons might defend those criteria by arguing that they “would be agreed on by people to the ex-
tent that they seek a principle recognizing them as equal, independent agents whose judgment
must be accorded equal weight.” T. M. Scanlon, “Preference and Urgency,” Journal of Philoso-
phy 72 (1975): 655–669.
33. On the encumbered self, see Sandel, Democracy’s Discontent, 14.
34. On the distinction between these two aspects of religious liberty and on the connections
between the arguments for them, see Murray, “The Problem of Religious Freedom,” 141–144,
148–151.
245
demanding character of religious requirements, which from the point of

view of those who are subject to them, are matters of fundamental obliga-
tion. It accepts the idea that free citizens—who accept that no com-
prehensive moral or religious view provides a defining condition of par-
ticipation or a test of the acceptability of arguments in support of the
exercise of political power—are, in a way, “encumbered”: it proposes a
rendering of the idea that such citizens have obligations and commit-
ments that are not properly understood, for purposes of political argu-
ment, as matters of choice. Second, it draws on the shared concern—fun-
damental to the deliberative conception—to find reasons that citizens
who are subject to what they regard as basic obligations can reasonably
be expected to acknowledge. And third, it draws on the fact that citizens
who are not religious have fundamental convictions that they take to im-
pose especially compelling obligations.
The first two points are by themselves sufficient, but the third under-
scores the unreasonableness of failing to acknowledge religious liberty;
for those who might be prepared to deny freedom of conscience and lib-
erty of worship to others will typically want to claim freedom of con-
science for themselves. And if they are unable to defend that freedom by
appealing to the truth of their views, then they will need to defend it by
reference to the stringency of the demands imposed by their fundamen-
tal and not unreasonable convictions. And then treating others as equals
will require that they give similar weight to other demands belonging to
that general category.
Suppose, then, that we prevent a person from fulfilling such demands
for reasons the person is compelled to regard as insufficient: “com-
pelled,” because denying the sufficiency of these reasons follows from a
religious or a moral philosophy that not unreasonably commands the
person’s conviction. This is to deny the person standing as an equal citi-
zen—to deny full and equal membership in the people whose collective
actions authorize the exercise of power. And that, according to the delib-
erative conception, is a failure of democracy. We have denied full mem-
bership by failing to provide a justification for the exercise of collective
power by reference to considerations that all who are members of the sov-
ereign body that authorizes the exercise of power and who are subject to
that power, and are prepared to cooperate on reasonable terms, can ac-
cept. There are many ways to exclude individuals and groups from The
People, and this surely is one.
To conclude, I want to make two observations about this account of re-
246
ligious liberty. First, my remarks are limited. I have not said anything di-
rectly about how to handle claims for religious exemption from general
obligations with a strong secular justification (including obligations to
educate children); or whether special provision is to be made for spe-
cifically religious convictions, as distinct from conscientious ethical con-
victions with no religious roots;35 or about tolerating the intolerant. My
aim here is not to resolve, or even to address, these issues: any view recog-
nizing rights of free exercise will need to face those hard questions. I am
interested only in making the more restricted point that a deliberative
conception of democracy is not barred—by its emphasis on an ideal of
democracy—from acknowledging a fundamental role for rights of reli-
gious liberty: indeed, that it must provide a place for such rights. The ba-
sis of such rights, on the deliberative view, lies deeper than contingent
popular consensus. Like rights of political expression, they are founded
on the idea of democracy itself.
Second, I emphasize that the rationale for the guarantees of religious
liberty that fall under the requirement of deliberative inclusion is neither
narrowly political nor antipolitical. It is not narrowly political, because
those guarantees are not simply about enabling people to participate in
normal politics (or to participate without fear), nor simply about im-
proving public discussion by adding more diverse voices to it. It is not
antipolitical, because they are not simply about ensuring the strength of
organized associations (churches among them) that help to protect indi-
viduals from the state’s power.36 The argument does not deny the links
between religious liberty and associational liberty. The idea instead is
that abridgments of such liberties would constitute denials to citizens of
standing as equal members of the sovereign people, by imposing in ways
that deny the force of reasons that are, by the lights of their own views,
compelling. The reasons for abridgment are unacceptably exclusionary,
because they are unsuited to the ideal of guiding the exercise of power by
a process of reason-giving suited to a system of free and equal citizens.
The view I am presenting might, then, be contrasted with an approach
35. On this last point: the key to the case for religious liberty is that the content of a view as-
signs stringent obligations to a person who holds it. But specifically religious content is not es-
sential.
36. See the discussion of this rationale in Stephen Carter, The Culture of Disbelief (New
York: Basic Books, 1993), 17–18, 35–39. As a general matter, Carter’s defense of religious liberty
seems too exclusively focused on the parallels between religious and associational liberty and,
correspondingly, too dismissive of the continuities between freedom of conscience and free-
dom of public worship. On those continuities, see Murray, “The Problem of Religious Free-
dom,” 148–151.
247
suggested by Roberto Unger’s conception of empowered democracy, as

well as with the approach I sketched in an earlier essay, “Deliberation
and Democratic Legitimacy.”37 According to Unger, a system of immuni-
ties—negative liberties—is one component of a democratic order be-
cause “[f]reedom as participation presupposes freedom as immunity.” The
mistake of “critics of traditional democratic theory” is to believe that “par-
ticipatory opportunities [are] a more than satisfactory substitute for im-
munity guarantees.” According to Unger, participation is no substitute;
instead, immunity rights are necessary if a citizen is to have the “safety that
encourages him to participate actively and independently in collective
decision-making.” I do not disagree with the claim that immunity rights are
necessary, nor with the criticism of other views. But I now think that
the deliberative conception of democracy provides the basis for a less in-
strumental, less strategic rationale for certain liberties, even when those
liberties are not needed to ensure appropriate inputs to democratic
procedure.
6. Expressive Liberty
The principle of deliberative inclusion extends naturally from religious
liberty to a wide guarantee of expressive liberty. By a “wide guarantee,” I
mean a guarantee not confined to political speech, even on very capa-
cious understandings of political speech. Cass Sunstein, for example, de-
fines political speech as speech that is “intended and received as a con-
tribution to public deliberation about some issue.”38 I believe that a
deliberative view supports stringent protections of expressive liberty, even
when the expression falls outside the political, thus understood.
The deliberative view thus extends a more familiar democracy-based
strand of free-speech theory, which defends stringent protections of spe-
cifically political speech as one prerequisite for a democratic framework
of collective choice.39 Alexander Meiklejohn’s version of this theory lo-
cates the roots of a strong free-speech guarantee in the U.S. constitu-
tional design of popular self-government. Because popular sovereignty re-
37. Unger, False Necessity, 525, emphases added; Joshua Cohen, “Deliberation and Demo-
cratic Legitimacy,” in Alan Hamlin and Phillip Petit, eds., The Good Polity (Oxford: Blackwell,
1989), 17–34 [reprinted as essay 1 in this collection].
38. See Cass Sunstein, Democracy and the Problem of Free Speech (New York: Free Press,
1993), 130.
39. See Alexander Meiklejohn, Political Freedom (New York: Harper & Brothers, 1960);
Sunstein, Democracy and the Problem of Free Speech; Robert Bork, “Neutral Principles and
Some First Amendment Problems,” Indiana Law Journal 47, 1 (Fall 1971): 1–35; Ely, Democ-
racy and Distrust; Owen Fiss, Liberalism Divided (Boulder, CO: Westview Press, 1996).
248
quires free and open discussion among citizens, the government

undercuts the Constitution’s defining principle—treats citizens as sub-
jects of government rather than its sovereign masters—when it interferes
with such discussion. Others who favor the democracy defense add three
considerations that supplement Meiklejohn’s constitutional argument:
(a) because citizens have diverse views, regulation of speech owing to its
content establishes a regime of political inequality by silencing certain
views or topics that may be important to some citizens; (b) content regu-
lation effectively restricts the flow of information, perhaps reducing the
quality of democratic discussion and decision; (c) content restrictions
might limit the range of views in the discussion, and limits on range
might confine the capacity of discussion itself to challenge received views
and preferences by presenting unconventional outlooks.
Each of these considerations—fairness, quality, and reflectiveness—
plays an important role in a full treatment of free expression as essential
to deliberative democracy. Here, however, my aim is to indicate how
the deliberative view supplements these considerations and thus extends
stringent protections beyond political speech—and thus also forestalls
the need to stretch the category of “political speech” to cover, for exam-
ple, Bleak House, Ulysses, and Mapplethorpe’s photographs (Sunstein’s
examples) so that they are stringently protected. Restriction to political
speech may seem natural, once one has decided to base rights of expres-
sion on potential contribution as input to a discussion about the proper
use of public power. But a deliberative conception must be cautious
about accepting such a limit. Though the idea of reasonable discussion
aimed at agreement is fundamental to the deliberative view, it does not
follow that the protection of expression is to be confined to speech that
contributes to such discussion. It may also need to extend to speech that
cannot permissibly be regulated as an outcome of such discussion.
Consider, then, expression that is not part of any process of political
discussion—not intended nor received as a contribution to public delib-
eration about some issue. But assume, too, that it reflects what a citizen
reasonably takes to be compelling considerations in support of expres-
sion. Such expression advances what I will call an “expressive interest”: a
direct interest in articulating thoughts and feelings on matters of personal
or broader human concern, whether that articulation influences the
thought and the conduct of others.40
40. For discussion, see Joshua Cohen, “Freedom of Expression,” Philosophy and Public Af-
fairs 22, 3 (Summer 1993): 207–263 [reprinted as essay 4 in this collection].
249
As examples, consider artistic expression driven by a concern to create

something of beauty; or bearing religious witness with no intention to
persuade others; or giving professional advice out of a sense of profes-
sional obligation, with no intention to shape broader processes of collec-
tive decision making. In the case of bearing witness, an agent endorses a
view that places him under an obligation to articulate that view and per-
haps to urge on others a different course of thought, feeling, or conduct.
Restricting expression would prevent the agent from fulfilling what she
takes to be an obligation, thus imposing a burden that the agent reason-
ably takes to be unacceptable. To acknowledge the weight of those rea-
sons, the deliberative view extends stringent protection to such expres-
sion. Given the background of reasonable pluralism, the failure to extend
such protection represents a failure to give due weight to the reasons that
support forms of expression that are not inputs to public discussion. As
such, it constitutes a denial of equal standing, and decisions to deny pro-
tection are not suitably collective.
Or take expression on matters of political justice. Here, the impor-
tance of the issue—indicated by its being a matter of justice—provides
a substantial reason for addressing it, regardless of how the message is
received. The precise content and weight of the reason are matters of
controversy. Aristotelian views identify public engagement as the highest
human good; and Brandeis urged that “public discussion is a politi-
cal duty.”41 But even if political expression is neither the highest good
nor a matter of duty, still, it is a requisite for being a good citizen, some-
times a matter of sheer decency. Typically, then, such expression has sup-
port from substantial reasons within different moral-political concep-
tions.
Bearing witness; speaking out on matters of justice; creating things of
beauty; giving professional advice: such cases suffice to underscore the
importance of the expressive interest. They work outward from fully con-
scientious expression—the paradigm of expression supported by sub-
stantial reasons from the agent’s point of view, and therefore expression
whose protection is supported by the principle of deliberative inclusion.
To be sure, different evaluative conceptions have different implications
for what is reasonable to say and to do. But all conceptions assign to those
who hold them substantial reasons for expression, quite apart from the
41. Whitney v. California, 274 U.S. 357, 375 (1927) (Brandeis, J., concurring).
250
value of the expression to the audience, even if there is no audience at

all. For this reason, the deliberative view endorses a strong presumption
against content regulation, but it does not confine that presumption to
political speech.
Other reasons may also support that presumption, understood as part
of a wide guarantee of expressive liberty: for example, considerations of
reflectiveness (discussed earlier), which suggest that all manner of speech
helps to form values and beliefs that also figure in public deliberation.
But we need not confine ourselves to considerations of this kind. Con-
tent regulation is to be rejected because of the reasons for speech that are
captured in the expressive interest, and not simply because such regula-
tions prematurely foreclose public discussion.
To illustrate the point about the deliberative framework, the expres-
sive interest, and a wide guarantee of expressive liberty, let us consider
the case of regulations of sexual expression: in particular, regulations of
pornography. Part of the trouble with such regulations—for example,
the pornography regulations urged by Catharine MacKinnon—lies in
this area.42 An example of such regulation is an Indianapolis ordinance,
adopted in 1986, which defines pornography as:
the graphic sexually explicit subordination of women, whether
in pictures or in words, that also includes one or more of the
following:
• Women are presented as sexual objects who enjoy pain or hu-
miliation;
• Women are presented as sexual objects who experience sexual
pleasure in being raped;
• Women are presented as sexual objects tied up or cut up or mu-
tilated or bruised or physically hurt;
• Women are presented being penetrated by objects or animals;
• Women are presented in scenarios of degradation, injury, tor-
ture, shown as filthy or inferior, bleeding, bruised or hurt in a
context that makes these conditions sexual;
• Women are presented as sexual objects for domination, con-
42. The discussion that follows is taken from Joshua Cohen, “Freedom, Equality, Pornogra-
phy,” in Austin Sarat and Thomas Kearns, eds., Justice and Injustice in Law and Legal Theory
(Ann Arbor: University of Michigan Press, 1996), 99–139.
251
quest, violation, exploitation, possession, or use, or through pos-

tures or positions of servility or submission or display.43
In a nutshell, the regulation targets the graphic fusion of sexuality and

subordination.
Turning, then, to the connections of sexual expression and expressive
interest, suppose that concerns about human welfare and the quality of
human life prompt expression; the evident importance of those concerns
provides substantial reasons for the expression. A paradigm is expression
about sex and sexuality—say, artistic expression that displays an antipathy
to existing sexual conventions, the limited sensibilities revealed in those
conventions, and the harm they impose. In a culture that is, as Kathy
Acker says, “horrendously moralistic,” it is understandable that such writ-
ers as Acker challenge understandings of sexuality “under the aegis of art,
[where] you’re allowed to actually deal with matters of sexuality.”44 Again,
in an interview, Kathy Acker says:
I think you’d agree there are various things in us—not all of

which are kind, gentle, and tender—readers of de Sade and
Genet would probably agree on this point! But I think you can
explore these things without becoming a mass murderer . . .
without causing real damage, without turning to real crime.
One way of exploring these things is through art; there are vari-
ous ways of doing this. We have . . . to find out what it is to be
human—and yet not wreak total havoc on the society.45
The human significance of sexuality lends special urgency to the ex-

plorations Acker describes. Moreover, and here I join the issue about
pornography and the expressive interest, that urgency does not decline
when, as in the case of pornography, sexuality mixes with power and sub-
ordination—when, as in materials covered by proposed regulations, it is
not “kind, gentle, and tender.” On the contrary, a writer may reasonably
think—as Acker apparently does—that coming to terms with such mix-
43. Indianapolis, Ind., City-Council General Ordinance No. 35 (June 11, 1984). The full text
is cited in Catharine MacKinnon, Feminism Unmodified (Cambridge, MA: Harvard University
Press, 1987), 274 note 1. The regulation was overturned in American Booksellers Ass’n. v.
Hudnut, 771 F.2d 323 (7th Cir. 1985), affirmed without opinion, 475 U.S. 1001 (1986).
44. See Kathy Acker, “Devoured by Myths: An Interview with Sylvere Lotringer,” in
Hannibal Lecter, My Father (New York: Semiotext(e), 1991).
45. Interview with Andrea Juno, in Andrea Juno and V. Vale, eds., Angry Women (San Fran-
cisco: Re/Search Publications, 1991), 184–185.
252
ing is especially important, precisely because, in the world as it is, power

is so deeply implicated in sexual identity and desire. To stay away from
the erotization of dominance and submission—as pornography regula-
tions require—is to avoid sexuality as it, to some indeterminate degree, is.
The connections between pornography (materials covered by the regu-
lation) and the expressive interest may actually be strengthened because,
in a world of unequal power, it engages our sexual desires, categories,
identities, and fantasies as they are—even if our aim is to transform them.
Regulations targeted particularly on the fusion of sexuality and subordi-
nation—on the apparent extremes of heterosexual and phallic conven-
tions—will cover too much. For it may be in part by working with that fu-
sion and acknowledging its force, rather than by simply depicting a world
of erotic possibilities beyond power, that we establish the basis for trans-
forming existing forms of sexuality.46
One difficulty with the regulations, then, is that they make no provi-
sion for the importance of the expressive interest—for the weight of the
reasons that move at least some people to produce sexually explicit mate-
rials that conflict with the regulations. Underscoring that lack of provi-
sion, MacKinnon criticizes the exception in current obscenity law for
materials with “literary, political, artistic, or scientific value”: “The inef-
fectualness of obscenity law is due in some part to exempting materials of
literary, political, artistic, or scientific value. Value can be found in any-
thing, depending, I have come to think, not only on one’s adherence to
postmodernism, but on how much one is being paid. And never underes-
timate the power of an erection, these days termed ‘entertainment,’ to
give a thing value.” Of course, the expressive interest may be overridden,
but the conventional rationales for regulation fail to acknowledge it, and
thus fall afoul of the requirement of deliberative inclusion. More particu-
larly, though the connections with the expressive interest do not settle the
issue, they do help to increase the burden of argument that must be car-
ried in justifying such regulations: those connections mean that defend-
ers of regulations must make a more compelling showing of the harms of
46. See Judith Butler, Gender Trouble: Feminism and the Subversion of Identity (New York:
Routledge, 1990); Susan Keller, “Viewing and Doing: Complicating Pornography’s Meaning,”
Georgetown Law Review 81 (1993): 2195–2228; Duncan Kennedy, Sexy Dressing: Essays on the
Power and Politics of Cultural Identity (Cambridge, MA: Harvard University Press, 1993), 126–
213.
253
pornography and not simply advance the speculative arguments that are
commonly proposed.
7. Moral Liberty
I want now to discuss the implications of the principle of deliberative in-
clusion in the area of moral liberty, what John Stuart Mill called the “lib-
erty of tastes and pursuits.”47 I propose to concentrate, in particular, on
the enforcement of morality.
My principal focus here will be on the permissibility of imposing crim-
inal punishment on citizens for violating the ethical code shared by the
majority in a society, even when that conduct is neither injurious nor of-
fensive to others: Is it permissible for a political society to use its criminal
law to force members to lead lives that are not, by the lights of the major-
ity, immoral or perverse?48 Criminalization is, of course, a special case,
and problems of moral liberty extend well beyond it. Mill’s defense of
moral liberty was as much a criticism of intrusive collective opinion as of
state regulation. And, confining attention to state action, we need to ad-
dress the codification of morality through regulations that are not backed
by criminal sanction: for example, bans on same-sex marriage. But be-
cause the issues have important affinities, and can be seen more sharply
in the area of criminalization, I will focus on it here.
In two important cases in the 1980s and 1990s—Bowers v. Hardwick
and Barnes v. Glen Theatre—the U.S. Supreme Court affirmed the con-
stitutional permissibility of criminalizing immoral conduct. In the first,
“the presumed belief of a majority of the electorate in Georgia that ho-
mosexual sodomy is immoral and unacceptable” was offered as a suitable
rationale for a law imposing criminal sanctions on consensual homosex-
ual sodomy.49 To be more precise, the Georgia law itself imposed crimi-
nal sanctions on consensual sodomy quite generally, but it was upheld by
the Court only as applied to homosexual sodomy. In the second, the
Court upheld an Indiana public indecency law requiring dancers to
wear pasties and G-strings. As rationale for the law, Justice Rehnquist of-
fers the public’s “moral disapproval of people appearing in the nude
among strangers in public places.” Moreover, he notes traditional com-
mon law restrictions on public nudity and the view underlying those re-
47. See J. S. Mill, On Liberty (Indianapolis: Hackett, 1978), ch. 3.

48. Ronald Dworkin, “Liberal Community,” California Law Review 77, 3 (1989): 479–504.
49. 476 U.S. 186, 196 (1986).
254
strictions: that public nudity was an act malum in se. Writing in concur-
rence, Justice Scalia emphasizes that the conduct is prohibited because it
is “immoral”—in conflict with “traditional moral belief”—not because it
is offensive: “The purpose of Indiana’s nudity law would be violated, I
think, if 60,000 fully consenting adults crowded into the Hoosier Dome
to display their genitals to one another, even if there was not an offended
innocent in the crowd.” Rejecting Mill’s harm principle—which requires
a showing of harm to others as a necessary condition for criminal sanc-
tion—Scalia says, “Our society prohibits, and all human societies have
prohibited, certain activities not because they harm others, but because
they are considered, in the traditional phrase, contra bonos mores, i.e.,
immoral”—and he goes on to mention “sadomasochism, cockfighting,
bestiality, suicide, drug use, prostitution, and sodomy” as areas in which
legal regulation of conduct is constitutionally permitted, though the ra-
tionale for the regulations is rooted in “traditional moral belief.”
As Scalia’s examples indicate, the debate about the enforcement of
morals is not confined to issues of sexual morality. Still, that has been a
central historical focus—as in Bowers and Barnes, as well as in disagree-
ments about generic anti-sodomy laws and in disagreements about regu-
lations of prostitution and pornography. I will maintain that focus here.
The debate about enforcing morality implicates in especially profound
ways the value of democracy. Indeed, the debate about the permissibility
of enforcing conventional ethics is commonly presented as a conflict be-
tween democracy, which is said to support the enforcement of morals,
and some other value—say, personal liberty or autonomy—which is seen
to be compromised by and to condemn that enforcement. Ronald
Dworkin, for example, says that the argument from democracy is “politi-
cally the most powerful argument against liberal tolerance.”50 And in his
classical critique of Lord Devlin’s defense of the enforcement of morals,
H. L. A. Hart says, “It seems fatally easy to believe that loyalty to demo-
cratic principles entails acceptance of what may be termed moral popu-
lism: the view that the majority have a moral right to dictate how all
should live.”51
To state the views of the moral populist in more generous terms: on
moral issues, the values of the majority ought to be decisive because no
50. Dworkin, “Liberal Community,” 483.

51. H. L. A. Hart, Law, Liberty, and Morality (Stanford, CA: Stanford University Press,
1963), 79.
255
other basis for determining our shared “moral environment” is compati-

ble with the equality of citizens—no other basis is fair to citizens as
equals. Moral disagreements in the community ought, then, to be re-
solved by a procedure of decision making that treats citizens as equals—
say, majority rule. Lord Devlin himself suggests this case for enforcement
in an essay titled “Democracy and Morality”: “Those who do not believe
in God must ask themselves what they mean when they say that they be-
lieve in democracy. Not that all men are born with equal brains—we
cannot believe that; but that they have at their command—and that in
this they are all born in the same degree—the faculty of telling right from
wrong. This is the whole meaning of democracy, for if in this endowment
men were not equal, it would be pernicious that in the government of
any society they should have equal rights.”52
A first response to the moral populist argues that majority rule on
moral questions—like majority rule decisions to establish racially segre-
gated schools—does not treat citizens as equals, because it permits peo-
ple to act on their view that some citizens are worth less than others. But
we ought to resist acceding to this temptation too quickly. It is at least
possible to condemn a way of life and the conduct it comprises without
condemning those who lead it as being worth less than others: that is, for
example, the point of current Catholic doctrine on homosexuality. We
may wish to argue, in the end, that this view is hostility barely con-
cealed—that when it comes to matters sexual, moral traditionalism is ho-
mophobia and misogyny carried out by other means. Still, the means
are different: a complex structure of religious and moral argument. We
ought, then, to be cautious about simply identifying a willingness to pun-
ish conduct judged immoral with either racism or sexism—which are, on
their face and without reinterpretation, about people, not ways of life.
“Attempting to preclude the entire population from acting in ways that
are perceived as immoral is not assimilable to comparatively disadvantag-
ing a given group out of simple hostility to its members.”53 To be sure,
regulating ways of life as immoral may in the end be as objectionable as
condemning people as being of lesser worth, but we need to distinguish
the troubles.
A second response to the argument from democracy is to defend a
52. Lord Devlin, “Democracy and Morality,” in The Enforcement of Morals (Oxford: Oxford
University Press, 1965).
53. See Ely, Democracy and Distrust, 256. See also John Hart Ely, “Professor Dworkin’s Ex-
ternal/Personal Preference Distinction,” Duke Law Review (1983): 985.
256
Millian harm principle on the basis of the value of autonomy or utilitar-

ian principles. But this strategy may unintentionally lend added force to
Devlin-style arguments. Let me explain, concentrating on appeals to the
value of autonomy.
The argument from autonomy takes two principal forms. According to
the first, the autonomy protected by restrictions on the enforcement of
morality is principally the autonomy of the agent whose conduct would
otherwise be regulated. According to the second, moral toleration bene-
fits others, whose autonomy is enhanced by a greater range of alterna-
tives.54 The first version is more familiar, and I will sketch a version of it
presented by Ronald Dworkin, who offers it as a reply to one kind of
communitarian argument for the enforcement of an ethical code.55 What
matters here are not the details but the central thesis: that we cannot
make people’s lives go better by requiring them to conform, through
threats of criminal punishment, to an ethical code that they reject and
would otherwise violate. Dworkin rests this thesis on a theory of the best
human life in which the value of autonomy or self-government plays a
central role. According to this theory, a person’s life is good only if the
person chooses the values that guide it—or, if choice conveys the wrong
picture, reflectively endorses those values.56
Dworkin endorses an especially strong form of the requirement of re-
flective endorsement, though he needs just such a strong form to defeat
the argument for enforcement.57 Reflective endorsement, he says, has
constitutive, not merely additive value.58 In particular, no part of a life
54. For an illustration of this strategy of argument, see Joseph Raz, The Morality of Freedom
(Oxford: Oxford University Press, 1986), ch. 15. According to Raz, the duty of toleration is “an
aspect of the duty of respect for autonomy.” Autonomy requires a range of choices among dif-
ferent ways of life that include “distinct and incompatible moral virtues.” People who endorse
those ways of life tend, however, to be intolerant of one another. So ensuring the structure of
alternatives required to foster autonomy requires guarantees against intolerance. But as this
sketch indicates, the rationale for the guarantees lies not in the benefits conferred on individuals
who receive protection but in the preservation of a system of alternative possibilities for others.
55. Dworkin, “Liberal Community.”
56. I take the term “reflective endorsement” from ibid., 485–486.
57. The argument closely resembles the “maximalist” strategies for defending freedom of ex-
pression that I discuss in my “Freedom of Expression.” The difficulties are correspondingly par-
allel.
58. The roots of this claim about the constitutive value of reflective endorsement lie in what
Dworkin calls the “challenge” model of value: the view that a good human life is a life that re-
sponds suitably to life’s challenges. Dworkin appears to think that an agent’s conduct counts as
meeting a challenge if and only if the agent engaging in the conduct reflectively endorses it.
See his “Foundations of Liberal Equality,” in The Tanner Lectures on Human Values 1990, vol.
11 (Salt Lake City: University of Utah Press, 1991). I don’t find Dworkin’s case very plausible but
will not pursue the reasons here. I am indebted to John Tully for discussion of Dworkin’s view.
257
contributes to the value of that life unless it is reflectively endorsed. So,

for example, the development of intellectual powers makes a life better
only if that development takes the form of self-development—develop-
ment guided by the reflectively accepted values of the agent. Altruistic
behavior—saving the lives of others, say—makes the saver’s life better
only if she endorses the value of altruistic behavior.
Dworkin’s conclusions about the importance of reflective endorse-
ment closely parallel Locke’s defense of religious toleration; indeed,
Dworkin’s defense of moral toleration might be seen as generalizing
Locke’s argument. According to Locke, “Although the magistrate’s opin-
ion in religion be sound, and the way that he appoints be truly evangeli-
cal, yet, if I be not thoroughly persuaded thereof in my own mind, there
will be no safety for me in following it. No way whatsoever that I shall
walk in against the dictates of my own conscience will ever bring me to
the mansions of the blessed. . . . Faith only and inward sincerity are the
things that procure acceptance with God.”59 Just as worship is of no
value unless accompanied by inner conviction, by faith freely embraced,
so, too, decent or socially beneficial conduct—a good impact on the
world—adds nothing to the value of a life unless the behavior is accom-
panied by a freely embraced, inner conviction about the value of the
conduct.
Reasoning from this premise about the dependence of salvation on in-
ner faith, Locke argued that the state could not save souls through en-
forced religious rituals and should not try. Similarly, premising the con-
stitutive value of autonomy, Dworkin concludes that the state cannot
make a person’s life better by forcing that person to live according to an
ethical code he rejects. The problem with moral paternalism, then, is not
that it fails to be suitably skeptical about the good, or that it provokes civil
conflict, but that it rests on an incorrect theory of the good life; it fails to
appreciate the constitutive role of reflective endorsement in the value of
a life. If this is right, then we cannot defend the enforcement of morals
by claiming to show equal concern for the true good of all citizens. In-
deed, that argument is, in Dworkin’s words, “self-defeating” because, ac-
cording to the autonomy theory, a life of enforced moral conformity is
not better than a life of chosen immorality.60 On the contrary, the re-
quirement of reflective endorsement supports freedom of personal
59. John Locke, Letter Concerning Toleration (Indianapolis: Bobbs-Merrill, 1955), 34.
60. See Dworkin, “Liberal Community,” 487.
258
choice under favorable conditions for considering how best to live (e.g.,
conditions in which people pursue different “experiments in living,” to
borrow Mill’s phrase).
This argument has considerable force. The premise about autonomy
serves as an axiom in a family of reasonable comprehensive moralities;
and for the sake of discussion, I assume that the conclusions are well sup-
ported by the premises. But as political argument, it has an important
shortcoming: it depends on a comprehensive philosophy of life, and the
deliberative view requires that, under conditions of reasonable pluralism,
we free political argument from such dependence, particularly when
such argument bears on the fundamentals of conduct.
Thus, the autonomy theory is a form of comprehensive moral liberal-
ism, rejected by citizens who think, not unreasonably, that human lives
are made good at least in part by their compliance with divine law, or
their conformity to the order of the universe, or the quality of their im-
pact on the world, or the extent to which they realize human powers. To
be sure, citizens who endorse such views may themselves reject the en-
forcement of morality because they judge it worse or inappropriate to
force lives to be as good as they can be. But just as the belief in religious
toleration is not and should not be presented as contingent on a religious
view about the sufficiency of inner faith to salvation, acceptance of moral
toleration is not and should not be presented as contingent on the view
that personal autonomy is the supreme moral value and the comprehen-
sive guide to conduct. I mentioned earlier that the autonomy argument
for moral toleration generalizes a Lockean argument for religious tolera-
tion; whereas the latter rejects enforcement of a religious code because
inner faith is required for salvation, the former, more abstractly, treats re-
flective endorsement as constitutive of the goodness of a life for the per-
son who lives it. I think there is something right in this use of religious
toleration as a model, but as I explain in some detail below, the delibera-
tive view presents the parallel in a different way: it emphasizes, in partic-
ular, the weight of the reasons that lie behind the regulated conduct and
the unacceptability for the purposes of political argument of the consid-
erations that would justify the regulations.
The failure of this argument might suggest that majority rule ought, af-
ter all, to extend to moral issues. Precisely by underscoring moral dis-
agreement—indeed, reasonable moral disagreement—the limits of the
autonomy argument might suggest that treating citizens as equals re-
259
quires that the majority be left free to fix the shared moral landscape.
Lord Devlin makes just this point. Agreeing that moral differences often
cannot be resolved through a good faith exercise of practical reason—
that “after centuries of debate, men of undoubted reasoning power and
honesty of purpose have shown themselves unable to agree on what the
moral law should be”61—he concludes that the arbiter of social morality
cannot be reason but must instead be the people—“the ordinary man,
the man in the jury box, who might also be called the reasonable man or
the right-minded man.”62
But the deliberative account of democracy rejects that conclusion. To
see why, notice first that controversies about the enforcement of moral-
ity characteristically track moral controversies: when enforcement is con-
troversial, so, too, is the moral question itself. The issue, then, is not
whether conduct, assumed not to be injurious to others, can permissibly
be regulated for moral reasons on which there is general agreement. In-
stead, the issue is whether conduct can permissibly be regulated for
moral reasons, despite deep and apparently unresolvable disagreement
about the morality of the conduct and the grounds for regulating it.
This point seems easy to miss. In his criticisms of Devlin, for example,
Hart asks “why we should not summon all the resources of our reason,
sympathetic understanding, as well as critical intelligence, and insist that
before general moral feeling is turned into criminal law it is submitted to
scrutiny of a different kind from Sir Patrick’s. Surely the legislator should
ask whether the general morality is based on ignorance, superstition, or
misunderstanding. . . .”63 Such questions are certainly in order. But in
controversial cases of enforcement, the assumption that regulations en-
force “general morality” or “general moral feeling” is typically not, unless
by “general” we mean “majority,” in which case a very large step needs to
be filled in—the step that authorizes the majority to speak in the name of
the community.
Take the case of homosexuality. It would be preposterous to say that
the moral consensus of the community condemns homosexuality, but
some people want nevertheless to practice it despite its accepted immo-
rality. Instead, some citizens, for religious or perhaps other reasons, con-
demn it and perhaps wish to regulate it (though not all who condemn
61. Devlin, “Democracy and Morality,” 93.

62. Ibid., 90.
63. H. L. A. Hart, “Immorality and Treason,” The Listener (30 July 1959): 3, my emphases.
260
also wish to regulate). Putting aside concededly insufficient appeals to

Scripture, two principal arguments have been advanced for condemna-
tion and regulation.64 First, procreation is the natural end of sexuality
(the natural end being the end for which God established sexual desire),
natural ends ought to be practically authoritative, and conduct that by
its nature is disconnected from or contravenes those ends is base, per-
verse, and worthless.65 Second, homosexuality violates a principle of
“complementarity,” warring against the differentiation—including sexual
differentiation—that is essential to God’s ordering of the universe: “Hu-
man beings are . . . nothing less than the work of God himself; and in the
complementarity of the sexes they are called to reflect the inner unity of
the Creator.”66
These lines of argument are parts of philosophies of life that others rea-
sonably reject (indeed, they are very much contested within the tradi-
tions of thought to which they belong).67 Others believe that there is
nothing sinful, immoral, or in any other way objectionable about being
gay or lesbian. Some reject the view that procreation is the natural end of
sexuality or, more generally, that human conduct has natural ends or that
those ends, such as they are, ought to be authoritative in settling the best
way to live. And some reject the particular metaphysics of creation that
founds the principle of complementarity.
Parallel points can be made for the other disputes: about sodomy, por-
nography, or nude dancing. In each case, we ought not to ask whether
the state may enforce the general morality of the community: on these is-
sues, though the majority may share a morality, the community does
64. A third argument, advanced by Roger Scruton, claims that homosexuality exhibits an
objectionable form of narcissism and obscene perception, that instead of “mov[ing] out from
my body towards the other, whose flesh is unknown to me . . . I remain locked within my body,
narcissistically contemplating in the other an excitement that is the mirror of my own.” Sexual
Desire (New York: Free Press, 1986), 310. Though Scruton presents the antinarcissism argu-
ment as an alternative to arguments from the natural end of sexuality and complementarity, it
is not clear that it stands independently of the latter.
65. For a crisp statement, see C. H. Peschke, Christian Ethics, vol. 2 (London: C. Goodliffe
Neale, 1978), 379.
66. For discussion of the idea of complementarity, see Scruton, Sexual Desire, 309; Andrew
Sullivan, Virtually Normal: An Argument About Homosexuality (New York: Knopf, 1995); and
especially Gareth Moore, The Body in Context: Sex and Catholicism (London: SCM Press,
1992), ch. 7. Moore rightly points out that, on natural interpretations, the idea of comple-
mentarity depends on the procreation-as-natural-end doctrine. The passage on com-
plementarity comes from the Letter on the Pastoral Care of Homosexual Persons of the Catho-
lic Church’s Congregation for the Doctrine of the Faith.
67. See, e.g., Sullivan, Virtually Normal; and Moore, The Body in Context.
261
not—the democratic community, constituted by free and equal citizens.

Disagreements are fundamental and deeply rooted in reasonable differ-
ences of outlook, associated with, among other things, different views
about our bodies, about the role of our embodiment and the pleasures as-
sociated with it in the conduct of our lives, about how to respond to the
independence from rational control characteristic of sexual pleasure.
Some citizens find the law of sin in our members: they see in the body an
obstacle to our highest purposes, or at least the source of temptation to
do wrong. Others think, not unreasonably, that embodiment is essential
to our nature, that bodily pleasures provide ways to break free of conven-
tional constraint, and that our capacity to transcend such constraints is
fundamental to our nature as free agents. These are matters on which—I
say this as platitude—people disagree, often profoundly, in thought, sen-
sibility, and conduct: a public basis for justification is absent. The con-
straints of shared evidence and conceptual precision required for agree-
ment are simply not in view. Law has no place here, not in a democracy
committed to treating its members as equals. As Justice Blackmun said in
dissent in Bowers: “That certain, but by no means all, religious groups
condemn the behavior at issue gives the state no license to impose their
judgments on the entire citizenry.”68
Fundamental, reasonable disagreement, then, puts strong pressure
against the enforcement of conventional ethics: the principles in the
name of which such enforcement might be justified cannot be accepted
by all who are subject to them. And that pressure is particularly strong
when, as in the case of sexual conduct, the regulations impinge deeply
on the lives of those who are regulated. Given the underlying moral divi-
sion, some members of the community suppose that the regulated con-
duct is not only permissible but also essential: that sexual intimacy is a
fundamental human good and that—particularly in view of the facts of
human diversity—its value is contingent in part on its being guided by
the judgments, feelings, and sensibilities of the parties to it.69 Regulations
impinge deeply, then, because the reasons that support such intimacy are
substantial and can be acknowledged as such by people who reject those
68. Bowers v. Hardwick, 476 U.S. 186 (1986).

69. For suggestive remarks about the importance of sexual intimacy—connecting its value
to the importance of individuality and imagination, while separating that value from concerns
about procreation, see Stuart Hampshire, Innocence and Experience (Cambridge, MA: Har-
vard University Press, 1989), 124–131.
262
reasons. And because the reasons are so substantial, there is a correspond-

ingly strong case against regulating them in the name of considerations
drawn from philosophies of life that some citizens reasonably reject.
At the same time, should regulations not impinge very deeply—should
the reasons supporting the conduct be less substantial—then the case
against enforcement is less compelling, even if the reasons come from
conventional ethics. Thus, regulating boxing because it is immoral—or
betting because it is—may be less deeply objectionable than regulating
sexuality. Though some may reject the reasons used to support the regu-
lations, they may nevertheless accept majority support as itself sufficient
reason. More generally, I think it is very difficult to argue for a blanket
condemnation of the enforcement of community morality—in particu-
lar, a principled condemnation that does not depend on the complexities
of legal categorization and associated slippery slopes—unless we premise
the autonomy version of moral liberalism. In any case, no such blanket
condemnation is intended here.
I said earlier that the proponents of enforcement often appeal to the
value of democracy, urging that the equality of citizens requires that ma-
jority values fix the moral environment. This point has considerable
force when a collective choice is necessary, as, for example, in the area of
security policy: there, we need to arrive at a common decision, so the ma-
jority may speak in the name of the community—the majority principle
itself may be a matter of general agreement among people who disagree
about the right policy.70 But where regulation is unnecessary, as in the
area of sexual morality, this rationale is unavailable.
The case for the majority as tribune of the community is not confined
to issues on which collective choice is mandatory. It also has consider-
able force when regulations do not impinge deeply: when they do not
cover conduct rooted in fundamental obligations or supported in other
ways by substantial reasons. In such cases, even if a regulation is not re-
quired, it is permissible to adopt one with majority support. For example,
taxing citizens to support research and development may be unnecessary,
but it is unobjectionable: given differences of judgment and interest, it is
70. I borrow the example of security policy from the discussion of toleration in Thomas
Nagel, Equality and Partiality (Oxford: Oxford University Press, 1991), 164–165. The general
point—that the case for majority rule weakens when a collective policy is not necessary and
that such policy is not required when it comes to moral issues—can be found in both Nagel
and Joel Feinberg, Harmless Wrongdoing (Oxford: Oxford University Press, 1990), 51.
263
to be settled by a procedure that treats people as equals. But fundamental

interests and substantial reasons are at stake in the area of sexual morality
(once more, the key issue of contest about the enforcement of morality).
Given the reasonable rejection of the moralities in the name of which
the regulations are imposed, the presence of such fundamental interests
condemns the regulations.
In short, then, moral liberty—like religious and expressive liberty—is
an ingredient in the democratic idea of collective choice by free and
equal citizens. The decision to regulate cannot be collective, in a suit-
able sense: it cannot arise from free reasoning among equals. It is for that
reason undemocratic.
8. Community, Legitimacy, Democracy

To conclude, I will sketch some remarks on a fundamental question that
I have postponed for the end: What makes the deliberative conception of
democracy compelling as an interpretation of the fundamental demo-
cratic idea—that the authorization to exercise state power is to arise from
the collective decisions of those whose decisions are to be governed by
that power?
The principal virtues of the deliberative conception are allied closely
with the way that the conception understands binding collective choice.
By emphasizing the importance of articulating shared reasons, the delib-
erative view expresses an especially compelling picture of the possible re-
lations among people within a democratic order; moreover, it states a
forceful ideal of political legitimacy for a democracy. I take up these two
points in turn.
First, the deliberative conception offers a more forceful rendering than
aggregative conceptions do of the fundamental democratic idea—the
idea that decisions about the exercise of state power are collective. It re-
quires that we offer considerations acceptable to others, understood to be
free, equal, and reasonable, and whose conduct will be governed by the
decisions. It requires more than that we count their interests, while keep-
ing our fingers crossed that those interests are outweighed.
This point about the attractions of the deliberative interpretation of
collective decisions can be stated in terms of ideas of political autonomy
and political community. If a political community is a group of people
sharing a comprehensive moral or religious view, then reasonable plural-
ism ruins political community. But on an alternative conception of politi-
264
cal community, deliberative democracy is a form of political community.

To see how, notice first that by requiring justification on terms accept-
able to others, deliberative democracy provides for political autonomy.
Without denying the coercive aspects of common political life, it re-
quires that all who are governed by collective decisions, who are ex-
pected to govern their own conduct by those decisions, must find the
bases of those decisions—the political values that support them—accept-
able, even when they disagree with the details of the decision.
Through this assurance of political autonomy, deliberative democracy
achieves one important element of the ideal of community—not because
collective decisions crystallize a shared ethical outlook that informs all
social life generally, nor because the collective good takes precedence
over liberties of members. Rather, deliberative democracy is connected
to political community because the requirement of shared reasons for the
exercise of political power—a requirement absent from the aggregative
view—itself expresses the full and equal membership of all in the sover-
eign body responsible for authorizing the exercise of that power and es-
tablishes the common reason and will of that body.
When I say that it expresses “full membership,” not simply equal mem-
bership, I mean membership in the collective sovereign that authorizes
the exercise of power, and not simply membership as a subject of that
power. To be sure, an alternative conception of full membership is avail-
able: persons might be said to be full members of a political society just
in case the values their philosophy of life comprises coincide with the
values that guide the exercise of political power. Under conditions of rea-
sonable pluralism, the deliberative view rejects such full coincidence,
even as an ideal of practical reason. It acknowledges a separation be-
tween, as Michael Sandel puts it, “our identity as citizens [and] our iden-
tity as persons more broadly conceived.”71
But why this separation? Why, Sandel asks, “should political delibera-
tion not reflect our best understanding of the highest human ends?” After
all, when it does so reflect, we have an experience of political community
unavailable from more truncated political argument, confined to com-
mon ground that can be occupied by alternative reasonable views.
The answer is contained in the idea of reasonable pluralism, and I will
not repeat the details here. Suffice to say that if we take the fact of reason-
71. Sandel, Democracy’s Discontent, 322.
265
able pluralism seriously, then we need to watch our third-person plurals

as we move from “our identity as citizens” to “our identity as persons”:
our identity as citizens is shared, our identity as persons is not. Acknowl-
edging this, what are we to make of “our best understanding of the high-
est human ends?” As citizens, we do not have—nor can we expect to se-
cure—a common view about the highest human ends. So the request
that we make “our best understanding” the basis of political deliberation
is empty—or, in practice, gets content from the conception endorsed by
a particular group of citizens. In contrast, as persons, we each may have
such understandings, but they are plural—they are “our best understand-
ings.” Because those views are incompatible, we cannot fully incorporate
them all into political justification, and to use any one in particular is un-
acceptable.
This suggestion about deliberative democracy and the value of com-
munity may seem strained in light of the role of religious, expressive, and
moral liberties in the deliberative view. For such liberties are commonly
represented as—for better or for worse—the solvent of community. And
that is especially true when we reject the enforcement of community mo-
rality.
But the deliberative view offers a reason for skepticism about that
claim. Under conditions of reasonable pluralism, the protection of the
liberties of the moderns is not a solvent of community. Reasonable plu-
ralism itself may be such a solvent, at least if we define community in
terms of a shared philosophy of life. But once we assume reasonable plu-
ralism, the protection of the liberties of the moderns turns out to be a
necessary, though insufficient, condition for the only plausible form of
political community. For those liberties fall under the “principle of inclu-
sion.” As that term indicates, they are conditions required to ensure the
equal standing of citizens as members of the collective body whose au-
thorization is required for the legitimate exercise of public power.
Finally, the deliberative conception of democracy presents an account
of when decisions made in a democracy are politically legitimate and
how to shape institutions and forms of argument so as to make legitimate
decisions.
Generally speaking, we have a strong case for political legitimacy
when the exercise of political power has sufficient justification. But, as a
conceptual matter, a person can believe that the exercise of power is well
justified—therefore legitimate—while also acknowledging that others
266
over whom it is exercised reject the justification. As a conceptual matter,

legitimacy does not require that the relevant justification be acknowl-
edged as such by those who are subject to the legitimate power: there
need be no justification to them. But the background of democracy—the
idea of citizens as free and equal—and the fact of reasonable pluralism
are important in characterizing a more limited conception of justifica-
tion: because of these conditions, the relevant justification must be ad-
dressed to citizens, by which I mean that its terms must be acknowledged
as suitable by those subject to political power. Given that citizens have
equal standing and are understood as free, and given the fact of reason-
able pluralism, we have an especially strong showing of legitimacy when
the exercise of state power is supported by considerations acknowledged
as reasons by the different views endorsed by reasonable citizens, who are
understood as equals. No other account of reasons is suited for this case.
267
8
M O N E Y, P O L I T I C S , P O L I T I C A L E Q U A L I T Y
1. Introduction
In this paper, I discuss and criticize the current system of electoral
finance in the United States and the constraints on the reform of that sys-
tem imposed by the Supreme Court.
1. I begin by stating and discussing a three-part principle of political
equality (section 2), which I present as a partial statement of a normative
ideal of democracy.
2. I argue that the current system of campaign finance conflicts with
the principle of political equality, in particular its requirement of equal
opportunity for political influence. Current arrangements establish, in
effect, a framework of inequalities of opportunity (sections 3, 4).
3. I discuss the constitutional limits on reform initially set down by the
Supreme Court in the 1976 case of Buckley v. Valeo and reinforced in a
number of cases decided since then. These decisions substantially limit
the role that the value of political equality can play in shaping our system
of campaign finance. A regulatory scheme that gave weight to that value
by aiming to equalize opportunities for political influence would (barring
This essay originally appeared in a volume dedicated to Judith Jarvis Thomson. In it, as in
everything I write, I aspire to meet her high standards of clarity. I am sure that I have not suc-
ceeded but am deeply indebted to Judy for demonstrating in all her work that it is possible to
say important things without sacrificing clarity. The essay started as a talk to a meeting of the
Northeast Citizen Action Resource Center. I have presented subsequent and expanded ver-
sions to the MIT Club in Washington, D.C., the Tufts University philosophy colloquium,
McGill University departments of philosophy and political science, and a Brown University
Conference on political equality. I also presented a draft at a meeting of the September Group
and earlier versions of the main ideas to political philosophy seminars (in fall 1995 and spring
1998). I am grateful for the comments I received and wish particularly to thank Philippe van
Parijs, Erik Olin Wright, and David Estlund for suggestions. I am indebted to Stephen
Ansolabehere and James Snyder for discussions of the current system of election finance and
to Leonardo Avritzer for discussions of the persisting importance of the Schumpeterian view of
democracy in contemporary democratic thought. As always, my debt to John Rawls runs
throughout.
268
special assumptions about diminishing returns to political investment) re-

duce the overall quantity of electoral speech. But the Court has held that
fundamental constitutional principles preclude any restrictions on the
amount of speech in the name of equalizing opportunities for political
influence (sections 5, 6).
4. I argue that the limits imposed by the Court reflect an unduly
narrow conception of democracy and the role of citizens in it, a con-
ception that—like the elite theories of democracy that trace to Joseph
Schumpeter—casts citizens exclusively in the role of audience for the
messages of elite competitors rather than in the role of political actor, as
listeners rather than, so to speak, content-providers.1 And I suggest that al-
ternatives to the current system, founded on a less narrow conception of
democracy and the role of citizens in it—a conception that does not
treat the elite-mass distinction as the central fact of political sociology—
might achieve a better reconciliation of expressive liberty and political
equality (section 7). In short, my central point is that the current system is
deeply troubling, not simply because it subordinates democracy to some-
thing else—to property, or to an abstract and absolutist view of free-
dom of expression—but because it can be seen as founded on and as
constitutionalizing a narrow conception of democracy and citizenship
and thus as precluding experimentation aimed at more fully realizing
democratic values.
I will not defend a particular proposal for reforming the system of cam-
paign finance, though for purposes of illustration, I will, from time to
time, refer to the voluntary public financing scheme adopted by Maine
voters in 1996, a variant of which was adopted in Massachusetts in 1998.
In essence, that scheme—whose constitutional standing remains uncer-
tain (though it has been upheld thus far)2—finances candidate cam-
paigns through a public fund, on condition that those candidates do not
raise or spend private money. So it combines public subsidy with volun-
1. See Joseph Schumpeter, Capitalism, Socialism, and Democracy (New York: Harper and
Row, 1942), chaps. 21, 22. Not that Schumpeter himself was especially concerned about ensur-
ing more informed electoral judgments. What comes from him is the thesis that we should
think of democracy as a particular way of organizing competition for political leadership—that
instead of using “birth, lot, wealth, violence, co-optation, learning, appointment, or examina-
tion” to resolve the contest for political power, democracies resolve it through voting in regular
elections—and think of the role of citizens as analogous to that of consumers in the product
market.
2. See Daggett v. Webster, U.S. District Court, District of Maine, slip op. (November 5,
1999 and January 7, 2000).
269
tary spending limits, incentives to accept the limits as condition for re-
ceiving the subsidy (therefore incentives to reduce the overall quantity of
speech), some constraints (in the form of reasonably low contribution
limits) on private money nonparticipants, and additional support for pub-
lic money candidates who face large private spending by opponents or
large independent expenditures.3
One final prefatory note: as I was adding final touches to this paper (in
January 2000), the Supreme Court announced its decision in Nixon v.
Shrink, in which it upheld Missouri’s statutory limits on campaign con-
tributions. Were I rewriting the paper now, I would make two changes in
light of the opinions in Nixon. First, I would change the discussion in
section 6 to take notice of the fact that Nixon v. Shrink rejects the idea
that restrictions on the size of campaign contributions should be sub-
jected to the most demanding level of scrutiny. Second, and more impor-
tant, I would underscore that the vast majority of the Court now seems
willing to uphold campaign finance regulations enacted to ensure a
more fair democratic process. This willingness is explicitly stated in Jus-
tice Breyer’s concurrence and is suggested as well by Justice Souter (writ-
ing for the Court), who indicates that corruption of democratic process is
not confined to the financial quid pro quo. The implications of this shift
remain to be seen. But the apparent departure in constitutional philoso-
phy is cause for hope.
2. A Principle of Political Equality

In a democratic society, the members are conceived of as free and equal
persons. A principle of political equality for a democracy presents norms
that are suited to persons thus conceived; it articulates values that apply
to democratic arrangements for making binding—authoritative and en-
forceable—collective decisions; and it aims to provide guidance about
the appropriate design of such arrangements. In particular, the norms
are to guide judgments about the right to vote, the rules of election orga-
nization and vote aggregation (ballot access, systems of representation,
electoral finance), and the organization of legislative and executive deci-
sion making.4 Thus, a principle of political equality applies to the frame-
3. More precisely, the system allows no private funds beyond the initial seed money required
to qualify for public funds. For discussion, see Ellen Miller, David Donnelly, and Janice Fine,
“Going Public,” in Joshua Cohen and Joel Rogers, eds., Money and Politics: Financing Elec-
tions Democratically, foreword by Gore Vidal (Boston: Beacon Press, 1999).
4. For illuminating discussion of the terrain, see Charles Beitz, Political Equality (Prince-
ton: Princeton University Press, 1989).
270
work for making authoritative and enforceable collective decisions and

specifies, inter alia, the system of rights and opportunities for free and
equal members to exercise political influence over decisions that they are
expected to comply with and that are made in their name. It does not ap-
ply to the dispersed networks of political-cultural discussion, founded on
the associational life of civil society—what Habermas calls the “informal
public sphere” and Rawls calls the “background culture.”5 It presents,
and is framed for the purpose of presenting, an account of, inter alia, de-
mands that free and equal members can legitimately make on the high-
est-level systems of authoritative collective decision making.
The principle of political equality I rely on here has three components.
It states that arrangements for making binding collective decisions are to
accommodate the following three norms:
1. Equal rights of participation, including rights of voting, associa-

tion, and office-holding, as well as rights of political expression,
with a strong presumption against restrictions on the content or
the viewpoint of expression, and against restrictions that are un-
duly burdensome to some individuals or groups;
2. A strong presumption in favor of equally weighted votes; and
3. Equal opportunities for effective political influence. This last re-
quirement, what Rawls has called “the fair value of political lib-
erty,” condemns inequalities in the opportunities for holding office
and influencing political decisions (by influencing the outcomes
of elections, the positions of candidates, and the conduct of inter-
election legislative and administrative decision making).6
To be sure, a principle of political equality is not the only requirement on

the authoritative system of collective decision making. Decisions should
5. I return to this limitation and note some reasons for it later, at 277–279. On the informal
public sphere, see Jürgen Habermas, Between Facts and Norms, trans. William Rehg (Cam-
bridge, MA: MIT Press, 1996); on the background culture, see John Rawls, Political Liberalism
(New York: Columbia University Press, 1996 [1993]), 14, 382n. 13.
6. See Political Liberalism, 327–330. The general idea is familiar. For example, in the 1986
case of Davis v. Bandemer, which concerned political gerrymandering, the Supreme Court in-
dicates that equal protection problems emerge when an “electoral system is arranged in a man-
ner that will consistently degrade influence on the political process as a whole,” Davis v.
Bandemer, 478 U.S. 109, 132 (1986). Notice the importance attached to “influence on the po-
litical process as a whole,” and not simply electoral influence. Lani Guinier refers to the norm
that “each voter should enjoy the same opportunity to influence political outcomes,” in The
Tyranny of the Majority: Fundamental Fairness in Representative Democracy (New York: Free
Press, 1994), 152. She emphasizes “the importance of an equal opportunity to influence public
policy, and not just to cast a ballot” (134).
271
also be substantively just, according to some reasonable conception of

justice, and effective at advancing the general welfare. But a principle of
political equality states norms that will normally override other consider-
ations, apart from the most fundamental requirements of justice. To be
sure, conflicts may emerge between and among the norms comprised by
the principle. So the force of saying that arrangements for making bind-
ing collective decisions are to accommodate all three components is that,
when conflicts emerge, we can’t say a priori which value is to give way. In
particular, if we accept this three-part principle, then we allow that we
may need to regulate speech to avoid certain kinds of inequalities in op-
portunities for political influence.7
This third requirement is modeled on the familiar norm of equality of
opportunity. Stated intuitively and abstractly, that norm says that one per-
son ought not to have greater chances than another to attain a desirable
position because of some quality that is irrelevant to performance in the
position. Using some familiar jargon, I will say that this expresses the
concept of equal opportunity, and that different conceptions of equal op-
portunity are distinguished by the interpretations they give to “irrele-
vant to performance.” For the sake of discussion here, I rely on Rawls’s
conception of equal opportunity, which specifies “irrelevant to perfor-
mance in the position” as follows: that people who are equally motivated
and equally able ought to have equal chances to attain the position.8
When this conception of equal opportunity is applied to the political
system, the relevant position is active citizen in the formal arrangements
of binding collective decision making. The requirement, then, is that
people who are equally motivated and equally able to play this role, by
influencing binding collective decisions, ought to have equal chances to
exercise such influence.9 The Constitution and the surrounding rules
governing elections as well as legislative, executive, and administrative
decision making establish this position. When suffrage was restricted to
property owners, economic position was a formal qualification for hold-
7. To use the standard constitutional jargon, equality of opportunity would provide a com-
pelling interest.
8. John Rawls, A Theory of Justice, rev. ed. (Cambridge, MA: Harvard University Press,
1999), 63. Similar requirements of equal political opportunity are found in a variety of accounts
of democracy. See, for example, Robert Dahl who attaches considerable importance to equal
opportunities to express preferences and citizen control over the political agenda. Robert Dahl,
Democracy and Its Critics (New Haven: Yale University Press, 1989). I am indebted to
Chappell Lawson for underscoring the consistency with Dahl’s view.
9. Theory of Justice, 197.
272
ing that position. We now agree that economic position is not a relevant
formal qualification. But if economic position is not a relevant formal
qualification for voting and other forms of political influence, how could
it be acceptable to organize the highest-level system for exercising politi-
cal influence in a way that makes the opportunity for such influence de-
pendent on economic position? How could it be acceptable to organize
the framework so that greater opportunity comes with greater resources?
I do not propose here to defend this principle (though see 292–302 for
some relevant considerations) but will confine myself to four comments
on the idea of equal opportunity for political influence. Three of the com-
ments bear on the content of the principle—why equal opportunity for
political influence, why equal opportunity for political influence, and
why equal opportunity for political influence—and one bears on its sta-
tus as “autonomous” or independent.
First, then, political equality demands equal opportunity for effective
political influence rather than equality of effective influence itself. In-
equalities of effective influence are sometimes acceptable, on any rea-
sonable view of political equality. Some citizens may be more influential
because, for example, they care more about politics. Differences of influ-
ence that trace to such differences in values and choices seem unobjec-
tionable. Similarly, if a person is more influential because her views are
widely shared, or her judgment widely trusted, and others are therefore
likely to be swayed by her position on the issue at hand: the differences of
influence trace to the distribution of political values and commitments in
the population, not to the organization of the structure of collective
choice. The requirement of equal opportunity for effective influence
condemns certain kinds of effective exclusion or dilution, but it does not
support charges of objectionable exclusion or dilution merely because I
am unwilling to make reasonable efforts to persuade others, or because
others regard my views as ridiculous, or because they lack confidence in
my judgment.
What about inequalities due to differences in persuasiveness, or in
physical attractiveness? In neither case are the greater opportunities for
influence due to aspects of the design of arrangements for making collec-
tive decisions that we can permissibly control. To be sure, we could make
collective efforts to reduce the importance of differential persuasiveness,
for example, by investing more in civic education. But the legitimacy
and importance of making such investment do not imply that it would
273
ever be permissible to regulate the activities of the persuasive in order to

achieve greater equality of opportunity. To regulate those activities would
go to the core of the free speech guarantee, by establishing regulations
that control viewpoint and are unduly burdensome. Moreover, it would
defeat the point of political discussion. After all, differences in persuasive-
ness are not irrelevant to performance in the position. Similarly, we
could try to control the power that flows from being attractive (such as it
is) but only by taking measures that would keep people from appearing
before one another (only radio spots, no TV). And such regulations
would, on their face, be damaging to political judgment.
Underlying this focus on opportunity is the idea that it is unreasonable
to demand influence irrespective of one’s own actions or of the consid-
ered convictions of other citizens. That demand is unreasonable, be-
cause a compelling interpretation of the idea of political equality must
ensure a place for individual responsibility. Members of a democratic so-
ciety are represented as free and equal. As free, they are to be treated as
responsible for their political judgments and conduct. So if I demand in-
fluence irrespective of the judgments of other citizens, then I deny the
importance of such responsibility. Once we accept it, then we accept,
too, that a regime with equal opportunity for effective influence is almost
certain to be associated with inequalities of actual influence.
Second, the norm of equal opportunity for political influence assigns
autonomous importance to political equality rather than merely depen-
dent or derivative importance.10 Thus, suppose we eliminate all inequali-
ties of political influence due to causes that we can identify as unjust
apart from their effects on political influence. Assume, in particular, that
the distribution of economic resources is fair and that effective participa-
tion is not impeded by stereotype or group hostility. Still, unequal oppor-
tunities for effective influence might result from inequalities—assumed
by stipulation to be fair—in the distribution of resources. In condemning
these unequal opportunities, the principle assigns autonomous impor-
tance to political equality. It does not require political equality simply as
a way to discourage independently cognizable forms of injustice.
Third, the principle requires that we ensure equal opportunity for po-
10. Here, I disagree with Dworkin’s account of political equality in “What Is Equality? Part
4: Political Equality,” University of San Francisco Law Review (Fall 1987). Dworkin there re-
jects the idea that political equality has autonomous importance.
274
litical influence, not simply that we ensure a certain threshold level of

opportunity—a principle of sufficiency or adequacy of opportunity11—or
a maximin level of opportunity.12 Thus, consider a public auction, the
winners of which get free television time to present their political views,
in particular their electoral views. But the proceeds of the auction go to a
fund that subsidizes political activity by low-income citizens: perhaps it
subsidizes media access, or internet access with assistance for content
provision. So holding the auction expands general opportunities for polit-
ical influence, but the opportunities are unequal in that greater opportu-
nities for influence are available only to those who have the resources to
win the auction.
One response is to deny the premise that underlies the distinction be-
tween equalization and the alternatives: that it is possible to improve ev-
eryone’s opportunity for influence. But that seems mistaken. If we estab-
lish a lottery, the winners of which get free television time for presenting
their political views, then everyone has greater opportunities for influ-
ence (anyone can win the lottery).
Putting this zero-sum response to the side, then, I note two points
about the merits of a limit on inequality of opportunity that is more mod-
est than the limit I endorse here. First, from the point of view of the issue
that motivates this essay, the distinction between equal opportunity and,
say, maximin opportunity is idle. I am concerned here with the issues
about liberty, equality, and democracy raised by the (in)famous sentence
in Buckley: “The concept that government may restrict the speech of
some elements of our society in order to enhance the relative voice of
others is wholly foreign to the first amendment.” Whether the restrictions
would serve to equalize opportunity, “adequatize” it, or maximin it, the
same issues emerge—unless it could be shown that the need for restric-
tions emerges only when we are concerned to achieve equal opportunity
for political influence.
Second, part of the reason for thinking that a maximin (or sufficiency)
view of opportunity is reasonable is that such a view makes sense when it
11. In discussions of equal opportunity in the context of education, the focus is often on ade-
quacy, in part because a number of state constitutions in the United States guarantee an ade-
quate level of education.
12. David Estlund explores these concerns in his excellent paper, “Political Quality,” Social
Philosophy and Policy 17, 1 (Winter 2000): 127–160.
275
comes to the distribution of economic resources. Although the equality

of citizens as moral persons imposes some pressure to reduce socioeco-
nomic inequalities—as a way to express the respect owed to equals—that
pressure is limited by the mutual benefits that can flow from inequalities.
A parallel case can be constructed, so it may seem, for inequalities of po-
litical opportunity. Although the equality of citizens as moral persons im-
poses some pressure to reduce inequalities of political opportunity—as a
way to express the respect owed to equals—that pressure is limited, it
might be argued, by the mutual benefits that can flow from inequalities
of opportunity. But this parallel is in part illusory. One reason that re-
source inequalities are not troubling in a world of moral equals is pre-
cisely because their equality is already expressed through the equal stand-
ing of individuals as citizens in the system of authoritative collective
decision making: “The basis for self-respect in a just society is not . . .
one’s income share but the publicly affirmed distribution of fundamental
rights and liberties. And this distribution being equal, everyone has a sim-
ilar and secure status when they meet to conduct the common affairs of
the wider society. No one is inclined to look beyond the constitutional af-
firmation of equality for further political ways of securing his status.”13
But if inequalities of opportunity extend to the political system itself, as
the authoritative system for making collective decisions, then the public
basis of mutual respect is less secure. To be sure, an explanation might be
given for the inequalities that does not depend on the idea that citizens
are unequal (namely, that the inequalities of political opportunity benefit
all). But when citizens lack assured equal standing, that explanation may
itself provoke suspicion.
Coming back to the principle of political equality, then, my final ob-
servation is that the principle requires equal opportunity for political in-
fluence. To clarify the force of this condition, I distinguish three inter-
pretations of the idea of political equality, each of which supplements
the requirement that votes not be diluted. Thus, equal opportunity for
electoral influence condemns inequalities in chances to hold office or
to influence the outcome of elections, but it is confined to the electoral
13. Rawls, Theory of Justice, 477. I trace this idea to Rousseau, in Joshua Cohen, “The Natu-
ral Goodness of Humanity,” in Christine Korsgaard, Barbara Herman, and Andrews Reath,
eds., Learning from the History of Ethics (Cambridge: Cambridge University Press, 1996), 102–
139.
276
setting.14 Equal opportunity for political influence (the requirement I en-

dorse here) extends beyond equal opportunity for electoral influence by
condemning inequalities in chances to influence decisions made by for-
mal political institutions. Thus, it condemns conditions in which citizens
have equal chances to influence the outcome of elections but unequal
chances to form or join groups that influence the outcome of legislative
decisions. Equal opportunity for public influence requires equal opportu-
nities to influence the formation of opinion in the wider, informal public
sphere as well as the decisions taken by formal political institutions.
The principle of political equality requires equal opportunity for politi-
cal influence and is thus more stringent than the norm of equal opportu-
nity for electoral influence. But it does not go as far as the requirement of
equal opportunity for public influence. Why not endorse this wider re-
quirement?15 After all, just as political influence is more important than
electoral influence (because of the nonelectoral ways to influence legis-
lative or executive decisions), public influence is arguably more impor-
tant than political influence. Thus, make the assumption that public
opinion is translated into legitimate law. Surely, then, it seems especially
desirable to have opportunities for shaping public opinion. So we might
14. Bruce Cain and Kathleen Sullivan both accept equal voting influence but reject equal
opportunity for electoral influence. See Cain, “Moralism and Realism in Campaign Finance
Reform,” University of Chicago Legal Forum (1995). At p. 136, Cain indicates that equalizing
electoral influence through restrictions on political expenditures threatens excessive respon-
siveness to “ill-formed majoritarian preferences.” The basis for that presumption is unclear but
appears to derive from the idea that spending limits restrict the flow of information, and thus
give too much sway to uniformed preferences. Sullivan’s case is far more plausible. See
Kathleen M. Sullivan, “Political Money and Freedom of Speech,” University of California Da-
vis Law Review 30, 31 (Spring 1997): esp. 674–675. She points out, rightly, that the equalizing
opportunities for electoral influence will require some regulations of election-related expres-
sion, but that no such regulations are required by equally weighted votes, however broadly we
interpret the range of unacceptable gerrymanders. So we will need to draw some lines between
electoral and political speech that occurs in informal political discussion. The result may be ei-
ther unacceptable restrictions of political speech in the informal public sphere, if the bound-
aries around electoral speech are loosely drawn, or only minimal corrections for unequal
chances for influence, if those boundaries are drawn more crisply. For if we know one thing
from our experience with regulation in this field, it is that every regulation represents an invita-
tion to invest in political strategies that are equally effective but that circumvent the regulation.
One might have thought that these “practical difficulties,” as Sullivan calls them, would
prompt efforts at legal invention. Sullivan puzzlingly treats them as insuperable hurdles.
15. To be sure, the boundaries are vague, as is amply demonstrated by the problem of regu-
lating issue advertising.
277
suppose that a case for ensuring equal chances for political influence
would support equal chances for public influence as well.
I think, however, that we should resist this conclusion and reject the
wide interpretation of the principle of political equality.
First, the content of the requirement of equal opportunity for public
influence is obscure. The informal process of opinion-formation is not at
all well defined, bounded, or understood: it extends throughout life and
spreads through all its spheres. So it is not clear what the requirement de-
mands: not clear, that is, when opportunities for influence are suitably
equal—when individuals who are equally motivated and equally able
have equal chances to influence the formation of public opinion. To
be sure, effective chances to persuade others and to discuss cultural
and political issues are important, but those chances are ensured both by
the protections of expressive and associative liberties that fall under the
first part of the principle of political equality and by a fair distribution
of resources. Here, I am asking whether there is a further, independent
requirement of equal opportunities for public influence. To be more
precise, I am asking whether such a requirement ought to be included
in a principle of political equality, as a fundamental political value to
be accommodated along with equal rights of participation and equally
weighted votes. If it is not included, it may still be legitimate to reduce
inequalities of opportunity for public influence by subsidizing opportuni-
ties for people with limited resources: through ensuring more traditional
public fora and expanding access to the new fora by, for example, sub-
sidizing internet access (addressing the so-called digital divide) and op-
portunities for content provision. But the obscurity of the norm speaks
against including it in a first principle of political equality, for such inclu-
sion might lead to excessive restrictions on expression.
Second, part of the reason for requiring equal chances for political in-
fluence is that the state speaks in the name of citizens, claiming authori-
zation for its binding collective decisions from its equal members; more-
over, its decisions are enforceable. So we want to be sure that that claim
is founded on arrangements that manifestly treat citizens as equals. But
in the wider public sphere, we have no such authoritative statement of
results. Although citizens have fundamental interests in chances for pub-
lic influence, the equality requirement is less compelling.
Third, part of the reason for ensuring equal opportunities for political
influence is to establish, in a visible, public way, the respect for citizens
278
as equal members of the collective body that authorizes the exercise of

political power. Given the uncertain content of the wider principle of
equal opportunity for public influence, it is perhaps unnecessary for en-
suring such mutual respect.
3. Facts and Trends

I want now to shift attention in two ways: from political equality in gen-
eral to the particulars of campaign finance, and from political norms to
facts and trends about current campaign finance in the United States.
The current system of financing, then, has four fundamental features:
Increasing costs: In the 1996 election cycle, $2.4 billion was raised, and
$2.2 billion was spent on candidate campaigns.16 In addition, another
$175 million was spent on independent expenditures and issue advocacy.
Independent expenditures are funds—roughly $25 million for the 1996
elections—that are used expressly to advocate the election of one candi-
date or the defeat of another but that are not spent in explicit coordina-
tion with a candidate’s campaign. In issue advocacy, money—roughly
$150 million for the 1996 elections—is spent supporting or opposing the
stand of an elected official or a challenger on some issue, but without ex-
pressly advocating the election or defeat of the candidate. These aggre-
gates nearly doubled the previous record.
Who gives? Though spending is growing steadily, the number of contrib-

utors remains small. In 1996, for example, just 0.1 percent of the popula-
tion gave more than $1000 to candidates and parties. Altogether, the
$1000+ contributors accounted for $638 million for the 1996 elections:
$477 million to candidates and parties, and another $161 million in soft
money and PAC contributions.
Moreover, business spending continues to dominate the scene. “In
1996 . . . the biggest source of campaign money—by far—was the busi-
ness community. Overall . . . business outspent labor by a factor of 11:1
and ideological groups by 19:1. Looking strictly at contributions to candi-
16. I use the 1996 numbers because the essay was originally written in 1999. The $2.4 billion
comprises public money for the presidential campaign ($211m), small donors contributing less
than $200 ($734m), larger donors contributing more than $200 ($597m), PACs ($234m), “soft
money,” which is contributed to the parties but not to be spent in connection with federal elec-
tions ($262m), and candidates themselves ($262m, led by Steve Forbes’s $37m).
279
dates, business gave nine times as much money as organized labor, and
fifteen times as much as ideological donors.”17
In Voice and Equality, Verba, Schlozman, and Brady provide two find-
ings that bear on our understanding of this relatively small pool of citi-
zens who participate in American politics by making financial contribu-
tions and who are responsible for a large share of contributions and
spending. First, willingness to contribute money is largely explained by
income—by the capacity to contribute—and not by political interest.
Whereas every other political-participatory act—voting, talking, giving
time to a campaign—is substantially explained by the participant’s gen-
eral interest in politics, contributing is explained very little by general po-
litical interest and very strongly by income.18 Second, the pool of contrib-
utors is unrepresentative of the citizenry: for example, contributors tend
to be more conservative on economic issues.19
Unregulated flows. The current system of finance is complex, and contri-

butions to candidate election campaigns are regulated. But here I want to
emphasize that certain areas of growing importance are entirely unregu-
lated:
• Soft money: Soft money given to political parties for activities alleg-
edly unrelated to federal elections—for example, get-out-the-vote
campaigns by a state Democratic or Republican Party—is entirely
unrestricted by federal law.20 Such soft money contributions grew by
17. From an online publication by Center for Responsive Politics, http://www.opensecrets
.org/pubs/bigpicture/overview/bpoverview.htm.
18. Level of political interest is measured by responses to survey questions that ask about the
respondent’s interest in local and national affairs. See Sidney Verba, Kay Lehman Schlozman,
and Henry E. Brady, Voice and Equality (Cambridge, MA: Harvard University Press, 1995),
553. The finding is striking, but not surprising. Someone with little political interest, thus mea-
sured, might be highly motivated to give to a candidate because of a concern about some par-
ticular issue, and assuming a declining marginal utility of money, the cost to the contributor is
very small. Moreover, people with high capacity but low interest are more likely to give than
people with low capacity and comparably low interest, because the former are more likely to
be asked for money. My guess is that the finding that financial contributions (unlike other
forms of activity) are largely explained by capacity rather than by interest is probably true for a
wide range of activities and is almost certainly true of any activity in which professional fund-
raisers are involved, because they target capacity, not motivation. Perhaps contributions to reli-
gious organizations are an exception.
19. Voice and Equality, 303, 358, 361–364, 477, 512, 516.
20. Federal candidates are, however, permitted to solicit soft money. For discussion of the
complexities of soft money, see Note, “Soft Money: The Current Rules and the Case for Re-
form,” Harvard Law Review 111 (1998).
280
206 percent between 1992 and 1996, to $262 million. Whereas cor-
porations and unions are prohibited from contributing money from
their treasury to a candidate, they can contribute soft money, with
no restrictions on amounts.
• Issue ads: Spending on issue ads is also unregulated by federal
law, because such advocacy is not explicit in its endorsement of
candidates. So corporations and unions can spend as they wish
on issue advocacy, with no disclosure requirements. Absent such
requirements, the estimate of $150 million in 1996 is inevitably
speculative, but everyone agrees that issue ads are growing in im-
portance.
• Candidate spending: Out-of-pocket spending by candidates is unreg-
ulated: the cases of Ross Perot, Steve Forbes, and Michael
Huffington are the most famously large doses of such spending.
• Independent expenditures: Whereas contributions to organizations
that engage in independent expenditures are regulated, the extent of
such expenditures cannot be regulated. The importance of such
spending has grown—fourfold between 1994 and 1996, to $22 mil-
lion—because the Supreme Court decided that spending by parties
on candidate elections cannot be regulated unless that spending is
expressly coordinated with the candidate. But contribution limits for
donations to parties are much higher than limits on giving to indi-
vidual campaigns: individuals can give $40,000 to a party in an elec-
tion cycle (half during the primary season, and half for the general
election), but only $2000 to a candidate. So I can give $20,000 to
the Democratic Party to spend on vote-for-Kennedy ads promoting
Ted Kennedy over other Democratic hopefuls for senator and then
another $20,000 to support Kennedy in the general election. So
long as the Democrats don’t ask Kennedy how to spend the money,
there is no problem.
Money matters. In 1996, the candidate who outspent his or her oppo-
nent won 92 percent of the House races and 88 percent of the Senate
races. These high correlations of spending and winning are typical. But
they leave open questions of fact and interpretation about the political
difference that money makes, even in the relatively well-defined arena of
candidate elections, much less in the wider arena of political influence.
For three things are true:
281
i. The bigger spender tends to win.

ii. Incumbents tend to win.
iii. Incumbents tend to be better fund-raisers.
The trick is to provide a consistent and empirically tenable interpretation

of these facts. For example, the correlation between spending and elec-
toral advantage may be spurious, as incumbency may directly confer
both. Or perhaps, instead, incumbency confers some fund-raising advan-
tage, and the money in turn directly confers electoral advantage—apart
from any direct, nonpecuniary incumbency advantage. The truth ap-
pears to be the latter: whereas incumbency makes it easier to raise money
and independently easier to win elections, the money itself confers elec-
toral benefit, as we see in open-seat races. Moreover, challengers who
spend more than incumbents do have considerably greater chances of
winning than challengers who spend less.21
Second, if incumbents are good at raising money (which confers elec-
toral benefit), that might be because incumbents are a survivor popu-
lation of especially talented candidates, and talent attracts money. Or
it might be that the powers of officeholding confer an advantage in fund-
raising, because contributors (individuals and particularly organized
groups) want to curry favor with officeholders as a result of the powers
associated with offices, and/or because reelection-seeking officeholders
need to please potential contributors and have a capacity to please ac-
cording to the powers of their office. On this issue, the answer seems not
to be that officeholders are a survivor population of high-quality candi-
dates but that officeholding itself creates an advantage in fund-raising.
Contributors care about the capacity to deliver results; they therefore pay
attention to the offices held by elected officials, and invest in those who,
by virtue of their official positions, have that capacity.22
21. The literature is vast. See Gary Jacobson, Money in Congressional Elections (New Ha-
ven: Yale University Press, 1980); Jonathan Krasno and Donald P. Green, “Preempting Quality
Challengers in House Elections,” Journal of Politics 50 (1988): 920–936; Stephen Ansolabehere
and James Snyder, “Money, Elections, and Candidate Quality” (unpublished paper, on file
with author).
22. If the capacity to raise money (especially from organized groups) reflects the powers of
office, we should not conclude that power is therefore a source of money rather than money is
a source of power, as Ansolabehere and Snyder suggest in “Money and Institutional Power.” Af-
ter all, it is not implausible that greater decision-making capacity (due to greater powers associ-
ated with office) is associated with greater fund-raising capacity because funders are interested
in influencing the exercise of official powers, and they target their investments accordingly. So
powers of office beget money, because money is a source of influence (over the exercise of
282
Putting the complexities to the side, what seems undeniable is that the
success of candidates depends on their fund-raising success, that the ca-
pacity to raise money depends on their performance, that the ability to at-
tract support from the groups that give depends on their conduct; contrib-
utors, by providing such support, gain some measure of influence over
electoral outcomes.
To summarize these four observations, then: formal politics is get-
ting more expensive, just as the flow of money unregulated by sum or
by source is increasing. Because of these increasing costs, and because
money is important to electoral success, candidates must be especially—
arguably increasingly—attentive to the interests and the concerns of the
relatively small and unrepresentative group of citizens who spend money
on politics and thus provide essential resources for running a modern
campaign.
4. Getting the Problem Right

Contemporary discussion of reform tends to focus on one of three issues:
that too much money is being spent in the aggregate; that candidates are
spending too much time raising money and courting donors; and that do-
nors get political favors in return for their contributions or other forms of
spending. I don’t think that any of these three concerns gets to the heart
of the problem.
The first strikes me as weightless: if campaigns were well run, debated
real issues, genuinely reached most citizens, and provided them with es-
sential information, why would we think that $2 billion over a two-year
election cycle is too much to spend? Perhaps we are not spending
enough.
Are candidates spending too much time fund-raising? Perhaps. Dick
Morris reports that President Clinton complained “bitterly” about time
spent fund-raising: “I can’t think. I can’t act. I can’t do anything but go to
fund-raisers and shake hands.”23 And Vincent Blasi has made a forceful
case that time devoted to fund-raising injures the democratic process by
limiting the capacity of representatives to do their principal work-infor-
powers). Stephen Ansolabehere and James Snyder, “Money and Institutional Power,” Texas
Law Review, 77, 7 (June 1999): 1673–1704; Stephen Ansolabehere and James Snyder, “Money
and Office,” in David Brady, John Cogan, and Morris Fiorina, eds., Continuity and Change in
House Elections (Stanford: Stanford University Press, 2000).
23. Dick Morris, Behind the Oval Office, 2d ed. (Washington, D.C.: Renaissance Books,
1998), 150–151.
283
mation gathering, constituency service, deliberating, legislating.24 But

the case for reducing the sheer time spent raising funds is not so clear.
Suppose, once more, that we had a system of campaign finance in which
each citizen could spend up to $250 on a candidate election, and that
candidates were required to raise all their resources from such contribu-
tions. If they spent lots of time fund-raising, perhaps that would be a good
thing: they would be required to meet with large numbers of potential
contributors and might learn from those discussions, but without the cur-
rent bias in the pool.
Are contributors getting favors in return for their money? Perhaps; but
even if they are not, a large problem of political fairness remains.
The idea of political fairness is captured by the requirement in the
principle of political equality mandating that citizens have equal oppor-
tunities for political influence. The vote is one form of influence, and the
one-person/one-vote requirement is an important implication of the idea
of equalizing opportunities for effective political influence. But when
money is as important a political resource as it is in our current system,
control of it is an important source of political influence. It enables
people to run for office, to support electoral efforts financially, and to
join together with like-minded others with the aim of persuading fellow
citizens on some issue of public concern. A system that does not regulate
the flow of money—nor provide (as in a system of public finance) alter-
natives to relying on private money—provides unequal opportunities for
political influence. It provides to wealthier citizens channels of influence
that are effectively unavailable to others, who are equally motivated and
equally able but lack the resources required for using those channels.
Do these channels of influence overwhelm others? Do they establish de-
cisive forms of power? Clearly they are not always decisive. But it seems
clear, too, that we will never have conclusive answers to questions about
the relative importance of different avenues of influence. What we can
say is that the current legal structure establishes a channel of influence
that is effectively open to some but not to others. That is itself the prob-
lem, however precisely this opportunity translates into power over deci-
sions.
So the principle of political equality—in particular, the norm of equal
24. Vincent Blasi, “Free Speech and the Widening Gyre of Fund-Raising: Why Campaign
Spending Limits May Not Violate the First Amendment After All,” 94 Columbia Law Review
1281 (1994).
284
opportunity for political influence—raises serious troubles for the cur-

rent system of finance.
5. Constitutional Landscape
What might be done to remedy this situation? To answer this question, I
start with the constitutional landscape.
In the 1976 case of Buckley v. Valeo, the Supreme Court heard a chal-
lenge to the Federal Election Campaign Act (FECA) of 1971, as
amended in 1974.25 The Court’s assessment was mixed: some parts were
upheld, some not. But the details of the decision matter less than the
framework of analysis and the argument announced in it. That analytic
framework comprises two key elements.
First, the Buckley Court held that “money is speech”: meaning that
spending money on politics—both contributions to campaigns and ex-
penditures (by candidates or individual citizens or organizations)—has
First Amendment protection. Indeed, as political speech, it lies at the
core of the First Amendment. For the First Amendment is centrally
(though not exclusively) about protecting political speech from regula-
tion, as a necessary condition for assuring popular sovereignty—rather
than governmental sovereignty—that defines the American constitu-
tional system.
The argument that spending is, for constitutional purposes, protected
political speech proceeds as follows: “Contribution and spending limita-
tions impose direct quantity restrictions [emphasis added] on political
communication and association. . . . A restriction on the amount of
money a person or group can spend on political communication during
a campaign necessarily reduces the quantity of expression by restrict-
ing the number of issues discussed, the depth of their exploration, and
the size of the audience reached. This is because virtually every means
of communicating in today’s mass society requires the expenditure of
money. . . . The electorate’s increasing dependence on television, radio,
and other mass media for news and information has made these expen-
sive modes of communication indispensable instruments of effective po-
litical speech.”26
So sending messages requires money, and restrictions on money there-
fore restrict such sending: they limit the “quantity” of speech. The quan-
25. Buckley v. Valeo, 424 U.S. 1 (1976).
26. Ibid., 19.
285
tity of speech is an important constitutional value, not simply because

speakers have an interest in advancing their views but also because audi-
ences—citizens, as the ultimate political authority—have an interest in
the fullest airing of issues, without control by government over what is
said or how much is said. Citizens may, of course, tune the messages out,
but because of the audience/citizen interest, state restrictions on the
quantity of speech face a chilly reception.
More particularly, the Court held that contributions and expenditures
both have First Amendment protection, but that regulations of contribu-
tions are less offensive to the First Amendment than regulations of expen-
ditures. Contributions are lower in the constitutional scale in part be-
cause the principal value of a contribution lies in the fact that it is given,
quite apart from its size. Though contributing more reveals greater inten-
sity of support, it does not itself add to the content of the basic message,
which is “I support Jones.” This claim—here I plead against interest—
strikes me as preposterous. Giving lots of money might well express a dif-
ferent belief than giving a smaller amount: namely, the belief that the
candidate I contribute to is a much better candidate than the competitor,
and that it is very important that he or she be elected. Apart from this im-
plausible consideration about the independence of the content of the
message sent by a contribution from the magnitude of that contribution,
the Court also noted that if contributions are regulated, citizens still have
other ways to get their messages out—by spending in ways that are not co-
ordinated with a campaign.
Neither in Buckley nor elsewhere does the Court contemplate the pos-
sibility that electoral speech—though assuredly political—should be, as a
general matter, easier to regulate than political speech—more generally
easier, say, than nonelectoral, political speech in the public sphere. This
possibility might have been defended along the following lines:27 the
Court might have treated speech in the electoral setting generally along
the lines that it has treated speech in the setting of ballot access law. Thus
the Court has generally taken the view that restrictions on ballot access—
say, restrictions on write-in ballots that prevent voters from writing in
Daffy Duck or restrictions on fusion candidates that prevent third parties
from cross-nominating major party candidates—are permissible because
the point of ballots is to select officeholders, not to have open-ended de-
27. C. Edwin Baker, “Campaign Expenditures and Free Speech,” Harvard Civil Rights-
Civil Liberties Law Review 33, 1 (Winter 1998): 1–55.
286
bate of political ideas: “The purpose of casting, counting, and recording

votes is to elect public officials, not to serve as a general forum for politi-
cal expression.”28 Similarly, the Court might have said that the principal
forum for political expression is the informal public sphere, not the elec-
toral setting in particular. The latter is a specific institution, designed for
a particular purpose—the selection of officials—and can permissibly be
regulated in light of that purpose. So if the purpose of elections is to
translate public opinion into an authorization to exercise power—to pro-
vide an accurate register of the state of collective opinion rather than to
form public opinion itself—then regulations designed to ensure such
translation should be permissible, even if they have the effect of reducing
the quantity of speech, in just the way that it is permissible to restrict
write-in candidacies in light of the institutional purpose of ballots.
One reason for rejecting this approach is that—particularly in a world
of virtually permanent campaigning—it would require difficult line-
drawing exercises to distinguish electoral speech from other forms of po-
litical speech. Those distinctions are much crisper in the ballot setting,
where the issue is whether and how a particular person’s name will ap-
pear on a well-defined ballot. Moreover, some ways of drawing the line
and regulating electoral speech might end up providing excessive protec-
tion for incumbents. Still, I don’t think this criticism is compelling. After
all, line-drawing is already necessary, as, for example, in the area of issue
advocacy.
More fundamentally, I suspect that the Court would—and should—
reject the idea that electoral speech performs a mere “translation func-
tion” and the conception of democracy associated with that idea. Elec-
tions, they might say, are important not only to translating an anteced-
ently articulated collective opinion into political power but also to
crystallizing such opinion in ways that enable the exercise of power to
take guidance from it.29 So we ought not to treat electoral speech as nar-
rowly institutional speech, with a well-defined purpose, nor to make the
permissibility of regulation turn on such treatment.
28. Burdick v. Takushi, 504 U.S. 428, 445 (1992); see also Timmons v. Twin Cities Area New
Party, 117 S. Ct. 1364, 1377 (1997) (Stevens, J., joined by Ginsburg and Souter, JJ., dissenting),
where the dissent attributes this view to the majority.
29. Samuel L. Popkin, The Reasoning Voter: Communication and Persuasion in Presidential
Campaigns (Chicago: University of Chicago Press, 1991); Arthur Lupia and Matthew
McCubbins, The Democratic Dilemma: Can Citizens Learn What They Need to Know? (Cam-
bridge: Cambridge University Press, 1998).
287
Returning to Buckley: The second main idea is that the state has a
compelling interest in avoiding the appearance and the reality of quid
pro quo—dollars for votes—corruption. “Corruption,” the Court said in
1985, “is a subversion of the political process,” and the “hallmark of cor-
ruption is the financial quid pro quo: dollars for political favors.”30 The
essential point is that the corruption rationale is narrowly understood—in
effect, as a generalization of bribery law.
The Court allows that there may be other compelling rationales for
regulating spending but insists that none has yet been identified. In
particular, the state is said not to have a compelling interest in “leveling
the playing field”—ensuring equal opportunity for political influence.
FECA, the Court says, was “aimed in part at equalizing the relative abil-
ity of all voters to affect electoral outcomes by placing a ceiling on expen-
ditures for political expression by citizens and groups.” But the majority
opinion rejects this rationale: “The concept that government may restrict
the speech of some elements of our society in order to enhance the rela-
tive voice of others is wholly foreign to the First Amendment.”31 In this
important remark, the Court does not dispute that restricting the voice
of some may enhance the relative voice of others—indeed, that it might
be necessary to enhancing their voice. Nor does it deny that such en-
hancement would be a very good thing, a legitimate and perhaps sub-
stantial governmental objective. Instead, the majority asserts that the
First Amendment bars the door to achieving equalization through restric-
tion on First Amendment liberties.
With those two elements in place, the rest of the system follows pretty
straightforwardly. Because contributions merit lesser First Amendment
protection, and because restrictions on “large contributions” are well de-
signed to avoid the appearance and reality of political quid pro quo,
restrictions on such contributions are permissible, though only if they
are addressed to quid pro quo corruption, and that means only if the reg-
ulated contributions are sufficiently large to pose a genuine threat of
such corruption. Because expenditures merit especially stringent protec-
tion, and because restrictions on expenditures do not advance the one
concededly compelling interest in the arena of electoral finance—the in-
terest in avoiding the appearance or reality of quid pro quo corruption—
expenditure restrictions are impermissible, unless they are voluntary, as
30. FEC v. NCPAC, 470 U.S. 480, 497 (1985).
31. Buckley v. Valeo, 424 U.S. 1, 48–49 (1976).
288
under the public financing scheme for presidential elections that was
part of FECA.
6. Persisting Constraints
In the period since Buckley, the two fundamentals of this framework
have been restated and reinforced, but they have not been changed.
Thus, the Court continues to hold that the First Amendment protects
both contributions and expenditures and has continued to emphasize
the importance of spending in contributing to the quantity of speech,
and thus to the interest of the audience, even more than it has empha-
sized the importance of protecting the interests of speakers. Because of
this emphasis on quantity of speech, the Court has held that the iden-
tity of the speaker is not especially relevant to the permissibility of regula-
tion. Particularly important and revealing in this connection is the 1978
Bellotti decision, in which the Court held that states could not regulate
corporate spending on ballot initiatives. The fact that the speakers were
not individual citizens but corporations did not matter because the pro-
tected value was not the corporation’s interest in speaking but the audi-
ence’s interest in a full airing of views.32 This is “the type of speech indis-
pensable to decision making in a democracy,” and its value “in terms of
its capacity for informing the public does not depend upon the identity
of its source, whether corporation, association, union, or individual.”33
“The Constitution,” according to the Court majority, “often protects in-
terests broader than those of the party seeking their vindication. The First
Amendment, in particular, serves significant societal interests,” in partic-
ular the interest in the “free discussion of governmental affairs.” So the
Court vindicated the expressive liberty of the corporation (in this case,
the bank) not because of any special concern for the corporation’s inter-
ests, or because of a judgment that the regulation was especially burden-
some to those interests, but because of a concern for the wider public in-
terest in informed decision making. The essential idea is captured in a
paraphrase of Mill’s reason for thinking that it is as bad to silence one as
to silence all: “Were an opinion a personal possession of no value except
to the owner; if to be obstructed in the enjoyment of it were simply a pri-
vate injury, it would make some difference who the injury was inflicted
upon. But the peculiar evil of silencing the expression of an opinion is,
32. First National Bank v. Bellotti, 435 U.S. 765, 767 (1978).
33. Ibid., 777.
289
that it is robbing the human race”34—or if that seems excessively high-

minded, let’s just say “that it is robbing the voters of relevant informa-
tion.”
As to the second element, the Court shows virtually no disposition to
break from Buckley’s claim that there is no such thing as a process being
corrupt because it is unfair, because it provides citizens with fundamen-
tally unequal chances to influence the political process—more precisely,
that even if such inequality is a form of unfairness, it is not of the same
constitutional magnitude as quid pro quo corruption and therefore does
not justify restrictions on expenditures.35 Put otherwise, the Court contin-
ues to be very solicitous of the interests of citizens as spectators, informa-
tion gatherers, observers—as consumers of information and argument
who can decide for themselves which messages to listen to—but contin-
ues to show much less concern for the interests of citizens as activists and
participants, seeking fair chances to influence others in the political
arena.
In one post-Buckley case, the Court majority has acknowledged con-
cerns about fair access—about a corruption extending beyond quid pro
quo. In Austin v. Michigan Chamber of Commerce, the Court upheld a
Michigan law prohibiting corporations from using general treasury funds
for independent expenditures in connection with state candidate elec-
tions.36 They upheld it because of concerns about the “corrosive and dis-
torting effects of immense aggregations of wealth that are accumulated
with the help of the corporate form and that have little or no correlation
with the public’s support for the corporation’s political ideas.” This talk
about “corrosive and distorting effects” acknowledges a corruption of
democratic process that extends beyond quid pro quo. But the case,
which drew a strongly worded dissent from Justice Scalia, has been virtu-
ally without impact on subsequent decisions, largely because it has been
interpreted as arising specifically from traditional concerns about corpo-
rations and wealth accumulated with the help of the corporate form, and
not as standing for a more general proposition about the effects of “aggre-
gations of wealth that have little or no correlation with the public’s sup-
port for the political ideas of the holders of that wealth.”
With the two fundamentals of the Court’s analysis remaining essen-
34. Mill, On Liberty (Indianapolis: Hackett, 1978), chap. 2, paragraph 1.

35. As I noted earlier (270), Nixon v. Shrink, 528 U.S. 377 (2000), may signal a change of di-
rection on this essential point.
36. Austin v. Michigan Chamber of Commerce, 494 U.S. 652 (1990).
290
tially fixed, proposed regulations continue to face very stringent, in prac-

tice nearly insuperable, hurdles.
The situation with contribution regulations has not changed funda-
mentally, though it may be somewhat stricter than it was after Buckley,
because the Court, as mentioned, has focused principally on the impor-
tance of an anticorruption rationale and not on the lesser First Amend-
ment importance of contributions. Because of its focus on corruption,
the Court has said that states cannot limit contributions to groups run-
ning ballot initiatives: because there is no danger of quid pro quo with a
candidate, there is no problem.37 Similarly, lower courts have been over-
turning laws with “low limits” on contributions ($100 for state contests).
Contribution limits cannot, for example, be justified by “level playing
field” arguments, or by the importance of enabling most people to play,
or by bringing more citizens into the process. The limit must be set such
that there is a plausible concern about quid pro quo: because it seems
implausible that you can buy many favors from the mayor of St. Louis for
$100, a low limit of that kind provokes suspicion that the aim is to level
the playing field not to fight corruption. In addition, there may well be
an emerging Court majority for the view that party contributions to can-
didates are, as it were, born pure: because political parties are coalitions
of candidates, those parties cannot be corrupting candidates by directly
supporting their campaigns. According to this view, party contributions
are to be treated on a par with the candidates’ own expenditures, which
cannot be regulated because there is no threat of corruption. Justices
Rehnquist, Scalia, Thomas, and Kennedy have taken this position, and it
may eventually win support from Breyer, O’Connor, or Souter.38
The situation with expenditures is similarly crisp, largely stable, with a
few signs of increased hostility to anything that suggests limits. Apart from
the special case of independent corporate expenditures on candidate
campaigns (as in Austin), the Court has not upheld mandatory expendi-
ture restrictions, nor are there signs that they will. Thus, in 1985 the
Court held that “independent” expenditures cannot be regulated, even
if the candidate supported by those expenditures has accepted public
money with associated voluntary limits.39 Moreover, it has adopted a
pretty broad interpretation of “independent.” The key point is “uncoordi-
37. Citizens Against Rent Control v. City of Berkeley, 454 U.S. 290 (1981).
38. Colorado Republican Federal Campaign Committee v. FEC.
39. FEC v. NCPAC, 470 U.S. 480 (1985) (overturning limits on independent expenditures
on behalf of Presidential candidates who have accepted public funding).
291
nated”: if spending is not explicitly coordinated with a candidate, then

quid pro quo concerns are absent.
So in Colorado Republican Federal Campaign Committee v. FEC, the
Court opinion held that spending by the Republican Party to defeat Sen-
ator Tim Wirth was an independent expenditure because Wirth’s Repub-
lican opponent candidate had not yet been chosen and no exchange of
support for favors could have been in play. In short, party spending in
support of a candidate is not, as such, coordinated, and may therefore be
protected.
Finally, as the definition of “independent” is capacious, so, too, the so-
licitude for independent spenders, thus defined, is very great. In a 1994
Eighth Circuit decision, the circuit court rejected a provision of a Min-
nesota public financing law that would have provided increased support
for publicly financed candidates facing opposition from independent
spending by PACs.40 Efforts by the state to match that spending would
have amounted, in effect, to chilling the speech of those independent op-
ponents. The theme here is potentially very important: the trouble with
this regulation is that it puts the state in the position of trying to reduce
the quantity of speech, and that is objectionable.
One case that looks different is a decision in the Eighth Circuit up-
holding a provision of a Minnesota public financing law that removes ex-
penditure caps from candidates who have accepted such caps as a condi-
tion for receiving public money, but who face opponents who do not and
who spend more than a specified amount.41 The challengers said that the
state’s incentives were too good to be voluntary: that the state was in ef-
fect coercing people into the public system and trying to reduce the
quantity of speech—likely to be the chief objection to waivers on expen-
diture limits in public financing schemes. Similarly, the District Court
for the Maine District has upheld provisions of the Maine law that pro-
vide additional support for clean money candidates facing high-spending
challengers.42
7. Democracy and Campaign Finance

The current system of campaign finance appears to be at odds with the
principle of equal opportunity for political influence. In the name of a
40. Day v. Holahan, 34 F.3d 1356 (8th Cir. 1994), cert. denied, 513 U.S. 1127 (1995).
41. Rosensteil v. Rodriguez, 101 F.3d 1544 (8th Cir. 1996), cert. denied, 520 U.S. 1229 (1997).
Also, Gable v. Patton, 142 F.3d 940 (6th Cir. 1998).
42. Daggett v. Webster, 74 F. Supp. 2d 53 (D. Me. 1999).
292
constitutionally basic liberty of speech, however, the Court has resisted

reform efforts that appeal to that principle. It is essential to understand
exactly what is—and what is not—being said by the Court and the allied
critics of reform. To reiterate: the Court has not said that the current sys-
tem already ensures equal opportunity, or that equal opportunity for in-
fluence is a trivial or illegitimate political concern, or that all policies
aimed at promoting it are constitutionally infirm, or that proposed re-
forms would be ineffective at advancing that value. Thus it is not true, as
one recent discussion states, that “Buckley outright rejected the legiti-
macy of the asserted interest in equalizing the relative ability of individu-
als and groups to affect election outcomes.”43 Buckley speaks to the mag-
nitude of the asserted interest, not its legitimacy. It is hard to see what, in
Buckley, would stand in the way of a redistributive voucher scheme with
benefits targeted on low-income citizens, so long as the scheme was not
accompanied by expenditure restrictions.44 Instead, the Court has said
that neither governments nor citizens themselves acting directly through
initiative can legitimately seek to equalize opportunities for political in-
fluence by means of regulations that reduce the quantity of speech. Such
reduction conflicts with the First Amendment’s free speech guarantee.
In the name of equality, it puts illegitimate restrictions on freedom of
speech.
I want to focus on this claim about illegitimate restrictions. But before
getting there, I need to consider an argument to the effect that there is no
deep conflict between liberty and equality in this area, and that the
Buckley framework is not a hurdle to achieving fair equality. Thus it
might be said that an ideal scheme of financing would accommodate
both expressive liberty and political equality by providing subsidies to all
eligible candidates (or to political parties) while attaching no conditions
to the receipt of those subsidies—no restrictions on expenditures by can-
didates who accept them (the current system of financing of presiden-
tial elections does attach conditions to the acceptance of public money).
43. Deborah Goldberg, ed., Writing Reform: A Guide to Drafting State and Local Cam-
paign Finance Laws (New York: Brennan Center for Justice, 1998), 1–7. The only evidence
cited for the proposition quoted in the text is the infamous line in Buckley about the im-
permissibility of “restricting the speech of some elements of our society in order to enhance
the relative voice of others.” But the passage is specifically about restricting speech, not about
the legitimacy of the interest in ensuring equal opportunity for political influence.
44. Bruce Ackerman’s voucher scheme is so accompanied: it excludes real money and per-
mits only voucher-based expenditures. See Bruce Ackerman, “The Patriot Option,” in Joshua
Cohen and Joel Rogers, eds., Money and Politics: Financing Elections Democratically (Boston:
Beacon Press, 1999).
293
By establishing floors that enabled candidates to compete without hav-

ing to appease the interests of contributors, the scheme would go some
way to equalizing opportunity for influence. By excluding ceilings, it
would achieve that equalization without reducing the level of speech,
thus eliminating worries about conflict with the first part of the principle
of political equality. Worries about public subsidies, because they prompt
concerns about incumbency protection, or other forms of official manip-
ulation, could be addressed by using alternative strategies for providing
floors: for example, tax credits, deductions, or vouchers that enable indi-
vidual citizens to finance elections, while eliminating the cost to them of
contributing.45
Put aside questions about whether such an “all floors/no ceilings” ap-
proach, with its focus on candidates, fully addresses the concerns about
opportunities for citizen influence. Still, it faces an obvious objection.
Private contributions and expenditures may well swamp floors unaccom-
panied by restrictions, so that no real equalization of opportunities for in-
fluence results. In response, the floors-only proponent might say that the
benefits of spending more money decline as quantities of money in-
crease; the production function for votes has a negative second deriva-
tive. Although this response has some force, it hardly seems sufficient to
dismiss the liberty/equality issue. If we take equal opportunity for influ-
ence as a basic political value, then we cannot make its satisfaction con-
tingent on a speculative judgment of this kind about the responsiveness
of votes to spending.
So a scheme of public financing likely needs to be paired with some
limits and some incentives to accept the limits. Consider, for example, a
system of voluntary public financing in which public money goes only
to candidates who agree to forgo private money; in which nonpublic
candidates face reasonably low contribution limits (say, $250 for state-
wide offices); and in which additional subsidies go to public money
candidates who face independent expenditures or high-spending private
money challengers. The Maine system is of this kind, and critics com-
plain that it includes too wide a range of limits. The crux of the worry is
that the regulation has the state taking the position that less money, and
therefore less speech, is better. And surely they would object still more
strenuously to more straightforward limits—for example, a narrower con-
45. See Zach Polett, “Empower Citizens,” and Ackerman, “The Patriot Option,” in Cohen
and Rogers, eds., Money and Politics.
294
ception of issue advocacy that would result in a widening of regulable ex-

penditures, or expenditure ceilings, or a less capacious conception of an
independent expenditure. Though the floors-only idea seems very attrac-
tive, then, I don’t think we can so easily evade the issue.
Returning, then, to the issue of “illegitimate restrictions of speech,” I
note first that the phrase is not a pleonasm. We have bribery laws, child
pornography laws, and contribution limits; restrictions on the time, man-
ner, and place of speech are widely accepted, and sometimes—as with
restrictions on campaigning within 100 feet of polling places—those re-
strictions apply exclusively to political speech: in short, some restrictions
of speech are acceptable. Moreover, the kinds of restrictions of speech
that are most profoundly objectionable—that offend most directly against
the value of freedom of expression—are restrictions very different from
those contemplated by campaign finance regulations.46 First, they are
directed against speech with certain contents or viewpoints. Such regula-
tions threaten to freeze the existing state of opinion and, perhaps, to insu-
late the government from popular criticism. But campaign finance regu-
lations are neutral with respect to content and viewpoint.
Second, restrictions are objectionable when they are directed against
certain persons or groups. They say in effect that some person or group is
not worthy of being heard or have the objective effect of imposing an un-
due burden on the expression of some group. Again, the regulations un-
der contemplation appear not to be of this kind.
Suppose, then, that regulations are content- and viewpoint-neutral and
do not impose undue burdens on some citizens or groups. Why might
they still represent unacceptable burdens on freedom of speech? It might
be—third—that they restrict more speech than is necessary for achieving
their goal of ensuring equal opportunity for political influence: perhaps,
that is, we can find alternative regulations that are less restrictive but
more or less as effective. But absent optimistic and highly speculative as-
sumptions about declining marginal benefits of money, I see no reason to
suppose that the proposed regulations are, in this way, unreasonable.
Consider, then, a content- and viewpoint-neutral regulation that is not
unduly burdensome to any group and that is no more restrictive of
speech than is necessary—given available alternatives—for ensuring equal
opportunity for political influence. Why should the sheer fact that it re-
46. This paragraph and the next two draw on Rawls’s discussion of the three conditions that
an acceptable regulation must meet. See Political Liberalism, 357–358.
295
duces the quantity of speech make it so objectionable? Why does that

suffice to trump the importance of equal opportunity for influence?
Two answers come to mind. The first is instrumental and concerns
threats to the quality of decisions that might result from restrictions. Re-
call the Court’s statement in Buckley that “a restriction on the amount of
money a person or group can spend on political communication during a
campaign necessarily reduces the quantity of expression by restricting the
number of issues discussed, the depth of their exploration, and the size of
the audience reached.” Here, the restrictions on money, which lead to
limits on the quantity of speech, are tied to a threat of making worse col-
lective decisions, because the restrictions limit the flow of information
and prevent a sufficiently close examination of the issues. In short, the re-
strictions make the outcomes worse.
This first, instrumental argument against reducing the quantity of ex-
pression seems very weak. It is not true that restrictions on money “nec-
essarily” restrict issue range, depth of exploration, or audience size.
Though they do limit quantity, the effects of quantity limits—whether
they transform into limits on quality—are contingent on the extent and
character of the restrictions and on what the money would have been
used for: if the money goes to more attack ads, then quantity declines,
but not range, depth, or audience size. Indeed, if Steve Ansolabehere and
Shanto Iyengar are right, spending on negative ads turns voters off. So an
increase in expenditures may produce a decline in audience size.47
A second argument is intrinsic and plays a large role in hostility to reg-
ulation: it claims that restrictions on the quantity of speech are objec-
tionable, not because they worsen political outcomes, but because they
worsen the democratic process itself by distorting the proper role of citi-
zens within it. In short, such restrictions conflict with the ideal of democ-
racy itself. The intrinsic argument is founded on an idea about individual
responsibility and its role in democracy. It says that democratic process,
properly understood, assigns to individual citizens the right and the re-
sponsibility to decide how much information is sufficient and to distin-
guish between reliable and unreliable sources—just as democracy as-
signs to individual citizens the responsibility to decide how much they
wish to participate, as indicated by the embrace of an equality of opportu-
nity rather than an equality of influence principle. But this assignment
47. Stephen Ansolabehere and Shanto Iyengar, Going Negative (New York: Free Press,
1995).
296
of responsibility is undermined when collective judgments about ap-

propriate levels and kinds of information replace individual judgments,
whether those collective judgments come from legislatures or from citi-
zen majorities acting directly through referenda. It is incompatible with
this idea of democracy to seek to correct, through collective means, for
biases or imbalances in available information, except perhaps by increas-
ing the level of speech. We cannot restrict the quantity of speech on the
ground that citizens may be misled by what they hear or that they may be
put off because they hear too much or because what they hear is so re-
lentlessly negative. Thus the Court’s essential claim in Buckley: “The
First Amendment denies government the power to determine that spend-
ing is wasteful, or excessive, or unwise. In the free society ordained by our
constitution it is not the government, but the people—individually as cit-
izens and candidates and collectively as associations and political com-
mittees—who must retain control over the quantity and range of debate
on public issues in a political campaign.”48
“The people,” as the passage between the dashes underscores, must
here be understood distributively, as the set of individual citizens and as-
sociations of citizens, not as a single collective authority. The Court here
denies that collective responsibility extends to the issue of how much
should be said in an election, or to the range of issues that ought to be
covered. Though the intrinsic argument emphasizes the role of individ-
ual responsibility, it does not deny the importance of a division of labor
within democracy between collective and individual responsibility. In-
stead it holds that we discharge our collective responsibility to uphold de-
mocracy by ensuring an open process of communication—with no re-
strictions on the flow of information or the content of communications—
that enables citizens to act with political responsibility by making their
own judgments about political affairs, including judgments about what
to pay attention to.
This intrinsic argument has considerable force. It does not commit the
critic of regulation to saying, for example, that property rights or private
liberties take precedence over democracy. Kathleen Sullivan correctly
observes that “arguments for greater limits on political contributions and
expenditure typically suggest that any claims for individual liberty to
spend political money ought to yield to an overriding interest in a well-
48. Buckley v. Valeo, 424 U.S. 1, 57 (1976).
297
functioning democracy.”49 The critic I have described here turns that ar-
gument around. This critic accepts the overriding interest in a well-func-
tioning democracy but argues that a “well-functioning democracy,” prop-
erly conceived, does not permit regulation of speech in the name of
equal opportunity for political influence. The critic who endorses the in-
trinsic argument does not say: “Yes, the current regime of campaign
finance injures democracy, but this injury is justified by the need to en-
sure that citizens can freely use their private property” (though, of course,
some critics may say that). Instead, the argument is that the value of de-
mocracy itself condemns regulation, because of the conception of re-
sponsibility ingredient in the best conception of democracy. The dis-
sent in Nixon v. Shrink suggests just this point: “The right to free speech
is a right held by each American, not by Americans en masse. The Court
in Buckley provided no basis for suppressing the speech of an individ-
ual candidate simply because other candidates (or candidates in the ag-
gregate) may succeed in reaching the voting public. Any such reasoning
would fly in the face of the premise of our political system—liberty
vested in individual hands safeguards the functioning of our democ-
racy.”50
Observing the earlier discussion of equal opportunity for influence, the
critic argues that a plausible principle of political equality, suited to a po-
litical society of free and equal persons, needs to include some account of
individual political responsibility. So the argument might be put this
way: the principle of political equality includes a right of free political
speech and an associated idea of political responsibility, implicit in its
hostility to content and viewpoint regulation and the distinction between
equalizing opportunity for influence and equalizing influence itself.
That’s part of what is involved in treating democratic citizens as free. But
once we embrace this notion of political responsibility, we must accept,
too, that collective regulation of the quantity of speech is incompatible
with democracy.
Though forceful, this argument is doubly deficient. First, it miscon-
ceives the case for regulation by representing it as dependent on a judg-
ment about who is entitled to decide whether the quantity and the kind
of information are sufficient. The argument for regulation based on the
principle of equal opportunity for influence is not of this kind. Though it
49. Kathleen M. Sullivan, “Political Money and Freedom of Speech,” 671.
50. Slip op., 17.
298
leads to restrictions on the quantity of speech, those restrictions are the

by-product of a principle of political fairness not of the claim that the leg-
islature or the majority of citizens are better judges of the value of politi-
cal messages than are citizens and their associations acting separately.
The problem that the regulations are designed to address is not that citi-
zens may be misled or put off by what they hear, but that they have a
powerful objection to a process whose organization does not even make
an effort to ensure equality of opportunity for influence among citizens
who are said to be equal. No insult to the freedom of citizens, or to their
capacity for responsible judgment, is implied or suggested.
Second, though the intrinsic argument against restrictions stakes its
case on the value of democratic process, it neglects an essential point
about that process. The point might be put in terms of the different inter-
ests of citizens in a democracy or in terms of roles associated with those
interests.51 The Buckley framework—like much democratic theory in the
“elite” tradition associated with Schumpeter—casts citizens principally
in the role of audience. As participants in democratic process, they have
a fundamental interest in listening to debates, acquiring information
through both formal political communications and more informal pro-
cesses of discussion,52 arriving at judgments about policies and candi-
dates, and acting as political agents when they express those judgments at
the polls, making informed judgments among competing candidates. But
in a democracy, citizens are also agents, participants, speakers, who may
aim to reshape both the terms of political debate and its results, by run-
ning for office or by seeking to influence the views of candidates, the out-
comes of elections, and the interelection conduct of politics.53 A require-
ment of equal opportunity for political influence aims to ensure that they
are in a position to play that role, should they wish to take it on. Of
course, they may also wish to influence politics through conduct in the
informal public sphere. But, once again, the principle of political equal-
51. For a parallel discussion, see Ronald Dworkin, “The Curse of American Politics,” New
York Review of Books 17 (October 1996): 19–24. Dworkin emphasizes the dual role of citizens, as
judges of electoral contests and as participants in those contests.
52. See Samuel Popkin’s The Reasoning Voter, on the acquisition of information through in-
formal discussion.
53. See my discussion of the deliberative and expressive interests, in “Freedom of Expres-
sion,” Philosophy and Public Affairs 22, 3 (summer 1993), 224–229 [reprinted as essay 4 in this
collection].
299
ity is confined to the organization of the arrangements of authoritative

collective decision making.
The claim that “democracy” casts citizens in this role and respects
their expressive-participatory interests might appear to depend on some
special philosophical view, whether Aristotelian or Rousseauean, about
the value of political participation in a well-lived human life. But it need
not be presented as so dependent. The idea that citizens have a funda-
mental interest in bringing their conceptions of justice to bear on the
conduct of political life is common to a range of philosophies of life.54 A
characteristic feature of different philosophies—different comprehensive
doctrines, in Rawls’s phrase—is that they assign to us strong reasons for
exercising responsible judgment about the proper directions of collective
life and for aiming to correct those directions particularly when they are
unjust; and those reasons are all the more compelling when authoritative
collective decisions are made in the name of those over whom they are
enforced. Aristotelians found those reasons on the central role of civic en-
gagement in a flourishing human life; Rousseaueans on the fundamental
value of individual autonomy and the connection of such autonomy with
political participation in a democratic polity; and some religiously based
philosophies on the commanding personal obligation to ensure social
justice and to respect human dignity. These alternative philosophies of
life each acknowledges that citizens have substantial, sometimes compel-
ling, reasons for addressing political affairs and a correspondingly funda-
mental “expressive” interest in favorable conditions for forming judg-
ments about the proper directions of policy and for acting on those
judgments—by presenting them to others and by seeking to correct for
injustices by acting in the political arena. Failure to acknowledge both
the weight of those reasons for the agent and the claims to opportunities
for effective influence that emerge from them reflects a failure to respect
the democratic idea of citizens as equals.
The weight of these reasons is reflected in part by the first component
of the principle of political equality, which requires equal rights of politi-
cal speech, association, and participation. But these reasons do not sim-
ply support a right to participate. They also yield a right to opportunities
for effective influence on the political environment. Moreover, because
claims for effective influence reflect the standing of citizens as equals,
54. Ibid., 224–226.
300
those claims are for an equal chance to influence: a failure to provide

such is a failure to acknowledge that equal standing.
More particularly, the aim must be to mitigate the impact on effective-
ness in the role of citizen of irrelevant facts about economic position—
particularly when that impact is a result of the design of arrangements of
binding collective decision making. And that means a different under-
standing of the division of individual and collective labor. Individuals re-
main responsible for finding the signals in the political noise that sur-
rounds them and for judging how far they wish to go in taking on the
role of participant, agent, speaker. Thus we keep free political speech,
without content or viewpoint restrictions, and maintain the influence/
opportunity-to-influence distinction. When it comes to acquiring the in-
formation needed to play this role, collective responsibility is to en-
sure open communication and perhaps to encourage, in the familiar
Brandeisian phrase, “more speech.” But collective responsibility extends
to ensuring that when citizens do decide to operate as political agents,
they have a fair chance for influence. We cannot reasonably expect peo-
ple to respect the results of a political process whose basic organization
effectively assigns greater opportunities for political influence to those
who are economically advantaged.
What makes the current constitutional framework so disturbing is that
it says that the people cannot permissibly adopt this conception of de-
mocracy and citizenship nor experiment with ways to secure equal op-
portunities for political influence while also protecting political speech.
It says that the Constitution enacts Joseph Schumpeter’s Capitalism, So-
cialism, and Democracy. To underscore the point, I conclude by contrast-
ing the framework of constitutional reasoning described here with the
framework presented by the European Court of Human Rights in the
case of Bowman v. The United Kingdom (1998).55 The case involved a
challenge to a 1983 British law (the Representation of the People Act)
that prohibited individuals from spending more than five pounds either
favoring or opposing the election of a particular parliamentary candidate
in the period immediately preceding an election. The case was decided
under Article 10 of the European Convention on Human Rights, which
states that the exercise of freedom of expression “may be subject to such
formalities, conditions, restrictions or penalties as are prescribed by law
55. Bowman v. United Kingdom, European Court of Human Rights, 19 February 1998, slip
op.
301
and are necessary in a democratic society.” More particularly, the court

needed to decide whether the regulation was more stringent than neces-
sary to foster a democratic society, where such fostering was understood
to comprise three legitimate aims: establishing fair conditions for com-
peting candidates, ensuring the independence of candidates from inter-
est groups, and preventing political debate around election time from fo-
cusing on single issues rather than on matters of broad concern. The
court found the five-pound limit excessive. The crucial point here, how-
ever, is not the conclusion but the court’s recognition that the three
aforementioned values are aspects of democracy, and that promoting
them provides an entirely legitimate reason for restricting the quantity of
speech in the period just prior to an election. Whatever the wisdom of
the court’s judgment in the Bowman case, the framework—-with its rec-
ognition that political fairness and freedom of expression are both ingre-
dients of democracy—is more suited to a democracy than the Buckley
framework is.56
8. Conclusion
A fundamental proposition of democratic thought is that our collective
decisions should reflect our judgments (the judgments of individual citi-
zens), formed through open processes of communication, unconstrained
by collective judgments about what and how much we should hear. But
this important principle must not lead to the undemocratic proposition
that citizens are equals only when we sit in the audience, listening to
what others say, and unequals when we take to the political stage. The
principle of political equality requires that we accommodate the interests
of citizens as audience and actor. We need to preserve a system of open
political communication that enables citizens to exercise their delibera-
tive responsibilities by forming their views against a background of ade-
quate information and rich debate and also that ensures equal access to
the public arena: we should not organize the political arena as a system
of unequal opportunities. Designing a regulatory scheme that promises
both will be hard: we need some experimentation. But we do not solve
the conundrum by throwing out half the democratic ideal.
56. I should add that it is very much consonant with the view suggested in the concurrence
by Justice Breyer in Nixon v. Shrink, which states that contribution limits are based on the
“need for democratization” and not simply on the concerns about quid pro quo corruption.
Slip op., 10.
302
9
P R I VA C Y, P L U R A L I S M , A N D D E M O C R A C Y
So the spheres of the political and the public, of the nonpublic

and the private, fall out from the content and application of
the conception of justice and its principles. If the so-called pri-
vate sphere is alleged to be a space exempt from justice, then
there is no such thing.
Sometimes those who appear to reject the idea of public rea-

son actually mean to assert the need for full and open discus-
sion in the background culture. With this political liberalism
full agrees.
—john r awls 1
P rivacy is a controversial idea of law and political morality. The con-

troversy it provokes can be divided into two areas. The first is about
the right to privacy, which requires protections of certain kinds of in-
formation and conduct from public disclosure and regulation. Debate
about the privacy right considers how we should characterize its content;
whether that content (if there is such a right) transcends concrete rights
that protect reputation, property, and bodily integrity; what connection,
if any, there is between the protection of interests in personal indepen-
dence characteristic of constitutional privacy and the protections of inter-
I presented earlier versions of this essay at a meeting of the Eastern Division of the Ameri-
can Philosophical Association, at the Universidad Torcuato di Tella, at the University of Geor-
gia, at the Graduate Center, City University, at the Inland Northwest Philosophy Conference,
and as a Donald Kalish Memorial Lecture at UCLA. I am indebted to Martha Nussbaum, Ju-
dith Thomson, Frances Kamm, David Estlund, Eduardo Rivera-López, Andrew Sabl, Seana
Shiffrin, and Erik Olin Wright for their helpful comments on earlier versions. I would like to
think that my friend Don Kalish would have endorsed the combination of privacy rights and
cultural democracy that I endorse here.
1. John Rawls, “The Idea of Public Reason Revisited,” University of Chicago Law Review 64
(1997): 791, 769.
303
ests in nonintrusion and nondisclosure of personal information associ-

ated with tort privacy; and what special challenges to such protections
may arise from new communications technologies. Stylizing greatly, we
might distinguish two parties to the dispute: privacy skeptics think we
should reject, as uncontrollably capacious, a right to privacy that tran-
scends protections of reputation, property, and bodily integrity; anti-
skeptics think that an expansive privacy right protects a distinct set of
interests in personal independence and need not be objectionably open-
ended, even if the relevant set of interests is not crisply circumscribed.
A second concern, more cultural than constitutional, is about conven-
tions of privacy—about the informal social norms that distinguish public
topics, which are fit for discussion, revelation, or disclosure, from topics
that are unsuitable for public consideration. Once more, we have two
stylized positions. The skeptic worries that solicitude for privacy masks
protection for privilege. Fueled by reflection on the history of appeals to
the family and to the firm as spheres of private ordering, skeptics fear that
a crisp public-private distinction in cultural debate will put the roots of
too much injustice and unreason beyond the reach of public criticism.
The antiskeptic argues that inattention to the boundaries between private
and public damages both private and public life: it damages private life,
because it threatens to “normalize” private diversity and complexity by
subjecting them to public scrutiny; it damages public life by swamping
deliberation about common concerns with therapeutic self-disclosure
and the intrusive revelation of personal information. Thus Hannah Arendt
expressed concern about our “eagerness to see recorded, displayed and
discussed in public what were once strictly private affairs and nobody’s
business.”2 And Thomas Nagel has urged that “something has gone wrong,
in the United States, with the conventions of privacy”—that more or less
any topic is now seen as fair game for public discussion.3
In this essay, I explore both topics—privacy rights and privacy conven-
tions—within the framework of a deliberative conception of democracy.
I will proceed in three steps. First, I sketch the rudiments of the delibera-
tive conception and indicate how its three essential elements—reason-
able pluralism, reasoning, and equality—provide a basis for rights to per-
2. Cited in Seyla Benhabib, “The Personal is not the Political,” Boston Review 24 (1999): 45.
3. Thomas Nagel, “Concealment and Exposure,” in Concealment and Exposure and Other
Essays (Oxford: Oxford University Press, 2002), 3.
304
P R I V A C Y, P L U R A L I S M , A N D D E M O C R A C Y
sonal liberties, and thus assign an essential place in democracy to the

liberties of the moderns.
Second, I extend this framework to the specific case of privacy rights
and focus on the issues of life, death, and sex that have been the subject
of constitutional privacy debates in the United States since the mid-
1960s. Rejecting skepticism, I argue that privacy rights—which I under-
stand as protecting independence of judgment—have an essential place in
a democracy understood as a society of equals. Once we acknowledge the
pluralism of philosophies of life characteristic of such a society, I argue,
we should understand privacy rights as expressing democracy’s central
values not as constraining such expression.
Finally, third, I discuss conventions of privacy. Here I am more skepti-
cal. Drawing on the distinction between an informal sphere of public dis-
cussion and cultural argument and the formal political sphere of authori-
tative collective decision making, I argue that the case for strong norms
of reticence is more compelling in the latter than in the former. I accept
what Rawls has called the duty of civility in the political public sphere of
a democracy and agree that duty imposes restrictions on appropriate
forms of public justification. But I do not think there is a comparable
case for the application of that duty to the informal public sphere: to
what I will call the system of cultural democracy. The idea of democracy
as a society of equals has different implications in the political sphere
than in the informal public sphere. In particular, the fundamental plu-
ralism that makes privacy rights essential to a deliberative-democratic sys-
tem of authoritative collective decision making should achieve more
open expression in the informal, democratic public sphere.
Deliberative Democracy
1. The deliberative-democratic idea is that authoritative collective deci-
sions should issue from public reasoning among equals.4 Three ideas
play a central role in this conception:
1. The fact of reasonable pluralism, according to which citizens en-
dorse a plurality of politically reasonable philosophies of life;
2. The requirement of deliberation, which states that the justification
4. The discussion that follows draws on Joshua Cohen, “Democracy and Liberty,” in Jon
Elster, ed., Deliberative Democracy (Cambridge: Cambridge University Press, 1998) [reprinted
as essay 7 in this collection].
305
of authoritative collective decisions should be founded on public

reasoning;
3. A conception of the members of a democratic society as free and
equal.
These three conditions, I suggest, impose restrictions on the principles

and values that provide the proper terms of public-political justification
in a democracy: that is, they circumscribe the content of democracy’s
public reason. After sketching these three ideas, then, I will indicate how
they support a conception of the content of public reason that includes a
constitutive role of personal, nonpolitical liberties.
First, then, the fact of reasonable pluralism is the fact that there are dis-
tinct and incompatible philosophies of life to which people, who are rea-
sonable politically speaking, are drawn under favorable conditions for the
exercise of practical reason. By a philosophy of life—what Rawls calls a
“comprehensive doctrine”—I mean an all-embracing view, religious or
secular in foundation, liberal or traditionalist in substance, that includes
an account of all ethical values and, crucially, provides a general guide to
conduct, individual as well as collective. People are reasonable, politi-
cally speaking—in short, politically reasonable—only if they are con-
cerned to live with others on terms that those others, understood as free
and equal, can also reasonably be expected to accept.
I say “politically reasonable” because the relevant notion of reasonable-
ness is suited to the context of political questions. Generically speaking, a
reasonable person is someone who gives due attention to the consider-
ations that bear on an issue and whose judgments and conduct reflect
that attention. So a person is politically reasonable if he or she gives due
attention to the facts about the political relation of citizens in a democ-
racy—in particular, the fact that political power is the collective power of
equals: that it belongs to us all, and that government of the people is to
be by and for the people. The fact of reasonable pluralism, then, is the
fact that when politically reasonable people engage in conscientious,
good-faith efforts at the exercise of practical reason—aimed at deciding
how to live—those efforts do not converge on a particular philosophy of
life, but lead to different views, many of which are compatible with de-
mocracy itself.
The fact of reasonable pluralism does not express a skeptical or nihilist
or relativist view about conceptions of the right way to live and the place
306
of human beings in the world: it does not say that the truth in these mat-
ters is unknowable, or that there is no truth, or that the truth varies across
persons, or cultures, or places; nor does it deny any of these views. Cer-
tainly, people who hold different outlooks think of their views as true:
they believe them, and believing is believing true. The fact of reasonable
pluralism does not conflict with what they believe but says that the truth
in such matters, if there be such, transcends the exercise of practical rea-
son that we can appropriately expect of others, as equals in the political so-
ciety. So it affirms a kind of toleration of reasonable differences in ulti-
mate outlook. As the italicized formulation indicates, when we assert the
fact of reasonable pluralism, we are already operating within the domain
of political argument. We are not relying on a philosophical theory about
the scope and competence of reason that provides common ground for
different outlooks: the nature and competence of reason, both theoretical
and practical, is one matter on which such outlooks disagree. Instead, we
are making a point about what we can appropriately expect of others, as
free and equal.
A crucial point for the issue of privacy is that liberalism itself is both a
political outlook and a philosophy of life.5 As a philosophy of life, liberal-
ism emphasizes the importance of autonomous choice—of reflective
self-direction—as a guide to conduct. As a political outlook, liberalism is
committed, inter alia, to securing basic personal and political liberties
through a system of rights, whose precise content is a matter of contro-
versy. One such controversy is about the scope and content of privacy
rights, and about whether such rights depend on a liberal philosophy of
life. I believe that they do not, for reasons I will discuss later.
Second, the deliberative conception of democracy puts public reason-
ing at the center of political justification, which I understand as the jus-
tification of coercively enforced regulations. To clarify, let’s distinguish
two conceptions of democracy, aggregative and deliberative. The basic
distinction lies in their interpretations of the fundamental idea of a dem-
ocratic society as a society of equals, and thus of a collective decision
made or authorized by citizens as equals. Both views apply in the first in-
5. A central point of Rawls’s political liberalism is to show that liberalism as a political con-
ception does not depend on liberalism as a broader philosophy of life, with a distinctive moral,
metaphysical, and epistemological outlook. See John Rawls, Justice as Fairness: A Restatement,
ed. Erin Kelly (Cambridge, MA: Harvard University Press, 2001). For the alternative view, see
Roberto Unger, Knowledge and Politics (New York: Free Press, 1974); and Michael Sandel,
Liberalism and the Limits of Justice (Cambridge: Cambridge University Press, 1982).
307
stance to institutions of binding collective decision making, and each in-

terprets the basic democratic ideal that such institutions are to treat peo-
ple bound by collective decisions—subject to them and expected to
comply with them—as equals. According to aggregative conceptions of
democracy, decisions are collective just in case they arise from proce-
dures of binding collective choice that assign equal weight to—or, more
generically, are positively responsive to—the interests of each person
bound by the decisions.6 According to a deliberative conception, a deci-
sion is collective just in case it emerges from arrangements of binding
collective choice that foster free public reasoning among equals who are
governed by the decisions.
The two views—aggregative and deliberative—share a conception of
political power as the collective power of equals. And they share the idea
that that power is exercised over members who typically do not have ef-
fective exit options from the political society.7 The deliberative view then
adds that the exercise of this power is rendered suitably collective—ap-
propriately by and for the people—only if the considerations that figure
in public argument used to justify its exercise (at least on fundamentals)
belong to the common reason of members.
According to the deliberative interpretation, then, democracy is a sys-
tem of social and political arrangements that ties the exercise of collec-
tive power to free reasoning among equals. Which considerations count
as reasons? The answer will not take the form of a generic account of
what a reason is, but a statement of which considerations count in favor
of proposals within a deliberative setting suited to the case of free politi-
cal association among equals, understood to include an acknowledgment
of reasonable pluralism. This background is reflected in the kinds of rea-
6. See, for example, the discussion of intrinsic equality and equal consideration in Thomas
Christiano, The Rule of the Many (Boulder: Westview, 1996); and Robert Dahl, Democracy
and Its Critics (New Haven: Yale University Press, 1989), 85, 87. Also, conditions imposed on
social choice functions—in particular, anonymity and independence—are sometimes moti-
vated in ways that suggest an identification of the requirement that people be treated as equals
with the requirement that interests be given equal consideration. See William Riker, Liberal-
ism Against Populism: A Confrontation Between the Theory of Democracy and the Theory of So-
cial Choice (San Francisco: W.H. Freeman, 1982).
7. Rawls suggests that we model the political relation by thinking of it as a relation in a
“structure we enter only by birth and exit only by death.” See Rawls, “The Idea of Public Rea-
son Revisited,” 769. To be sure, this modeling assumption is not literally true, but it captures
the idea that political decisions are binding on citizens, and that they cannot be made compati-
ble with the freedom of citizens simply by establishing a right to emigrate.
308
sons that will be acceptable: meaning, as always, acceptable to individu-

als as free and equal.
Third, then, in democratic deliberation, participants are and regard
one another as free: recognizing the fact of reasonable pluralism, they
acknowledge that no comprehensive moral or religious view provides a
defining condition of participation or a test of the acceptability of argu-
ments in support of the exercise of political power. Moreover, a democ-
racy is a society of equals. Everyone with the deliberative capacities—
which is to say, more or less all human beings—has and is recognized as
having equal standing at each of the stages of the deliberative process.
Each, that is, can propose issues for the agenda, propose solutions to the
issues on the agenda, and offer reasons in support of or in criticism of
proposed solutions. And each has an equal voice in the decision.
2. I have sketched the deliberative conception of democracy, with its
three main ideas—reasonable pluralism, reasoning, and equality. Subject
to these assumptions, political justification cannot proceed simply by
advancing considerations one takes to be true or compelling. For those
considerations—whether about the importance of personal autonomy in
the conduct of life or the value or rightness of willing submission to di-
vine command—may well be rejected by politically reasonable equals,
with conflicting philosophies of life, founded on opposing religious and
moral outlooks. One needs instead to find considerations that can rea-
sonably be expected to be acceptable to others, as equals: with political
power as the power of all as equals—imposed on all and in the name of
all—political justification is to proceed on common ground. Interpreted
by reference to pluralism, equality, and deliberation, democracy emerges
as an arena not only of public rights but also of autonomous public argu-
ment—a space of public reasoning and justification open to all on equal
terms.
These considerations have implications for the content of democracy’s
public reason: they constrain suitable political reasons in a democracy.
What is of immediate interest is that they provide a way to show the es-
sential place of nonpolitical liberties in a democracy and underscore that
that place does not require an appeal to a moral-liberal philosophy of life.
In particular, pressure for liberty comes from at least two sources within
the political ideal of democracy.
First, because of the pluralism of philosophies of life among politically
reasonable citizens, some bases for regulating conduct are politically
309
weightless. To take the clearest case, people hold some commitments on

faith, and take those commitments to impose overriding obligations.
Such nonnegotiable commitments are not as such unreasonable, nor is
there anything unreasonable about embracing them as true. But because
they are expressly held as truths known through faith, they are matters on
which reasonable people disagree, and adherents cannot reasonably ex-
pect others to accept those considerations as having any weight and,
therefore, cannot use them in justifying regulations. And the fact that
they cannot, will impose pressure for personal liberties—say, religious
and moral liberty—which are often restricted for such unacceptable
reasons.
Second, acceptable considerations—which are not weightless—will
have different weights in political justification. Even if they are not weight-
less, they may be insufficient to override the reasons that can be acknowl-
edged, consistent with reasonable pluralism, as commending or com-
manding the conduct whose regulation is under contemplation. And the
weight will depend on the nature of the regulated conduct and, in partic-
ular, on the weight of the reasons that support the conduct.
Take, for example, the value of public order. It provides an acceptable
rationale for regulating conduct. Different outlooks have different ways
of explaining the value of public order—as a precondition of the full ex-
ercise of personal autonomy, as an expression of submission to God’s
law—and people are bound to disagree about what public order requires.
But it will not be acceptable to suppose that, as a general matter, the
value of public order transcends all other political values. Except perhaps
in the most extreme circumstances, for example, a state may not impose
a blanket prohibition on alcohol consumption—including consumption
in religious services—in the name of public order. After all, the consider-
ations that support such consumption include considerations of obliga-
tion, which will provide a suitable basis for rejecting a justification cast in
terms of the value of public order, except in the most extreme conditions.
To be sure, not all citizens acknowledge the obligations in question. But
even those who do not can see the weightiness of those reasons, within
the outlooks of other politically reasonable citizens.
I will come back later, in the discussion of privacy, to this idea about
the weightiness of reasons that support a course of conduct. Suffice to say
here that if we take these two considerations together, we have the basis
for a strong case for rights to religious, moral, and expressive liberties as
310
elements of democracy’s public reason. Thus, conduct in these areas is

supported by strong (perhaps compelling) reasons, as when religious
exercise is a matter of obligation according to a person’s reasonable reli-
gious outlook. At the same time, standard reasons for restriction—reli-
gious and sectarian moral reasons—will often be weightless. More gener-
ally, we can see, at least in general terms, how a case for personal liberties
emerges without resting on ideas—of personal autonomy, individuality,
or self-ownership—drawn from liberalism, understood as a general phi-
losophy of life. The ideas of pluralism and deliberative justification ad-
dressed to equals together establish a stricture against religious and mor-
alistic justification (along with the stricture against secular-liberal moral
justification); moreover, they require attention to the burdensomeness of
regulations, as defined by the nature of the reasons that support the regu-
lated conduct.
Given this rationale for personal liberties, we can see why they are ele-
ments of democracy. For imposing regulations in the name of reasons
that are either weightless or of insufficient force to override reasonable
demands violates the fundamental democratic idea: the idea that a de-
mocracy is a society of equals and that authorization to exercise state
power must arise from the collective decisions of the equal members of
a society who are governed by that power, that such authorization must
be supported by reasons that can be shared by the addressees of the regu-
lations.
Privacy Rights
1. Thus far I have sketched a general framework of ideas about democ-
racy and indicated how at least some personal liberties might be under-
stood to have a constitutive place in democracy’s public reason. Under
conditions of moral pluralism, the collective authorization that lies at the
heart of democracy has limited scope.
What, then, about the right of privacy—in particular, a right to per-
sonal independence in areas of life, death, and sex? Can it be defended
without appealing to a liberal philosophy of life? Skepticism in American
constitutional argument about a capacious right to privacy derives in part
from the concern that such a right does depend on a liberal philosophy of
life. In particular, constitutional theorists who interpret the constitution
as fundamentally a design of democracy, whether aggregative or deliber-
ative, have typically been skeptical about a constitutional privacy right, in
311
part for this reason.8 Because the constitution is a shared, fundamental,

public-political framework, they are troubled—as they should be—by the
constitutionalization of a liberal philosophy of life.
Although their concerns are reasonable, their conclusions are incor-
rect. A fundamental right of privacy does not depend on a liberal philoso-
phy of life. Or so I hope to show.
2. To focus the discussion, I will begin with a widely cited remark in
the 1977 Supreme Court case of Roe v. Whalen. The details of the case
do not matter here. Suffice to say that Justice Stevens there distinguishes
two kinds of interests that privacy rights are commonly thought to pro-
tect: “The individual interest in avoiding disclosure of personal matters
. . . and the interest in independence in making certain kinds of impor-
tant decisions.”9 The former is the constitutional analogue of one of the
interests protected by the privacy tort, and I will not discuss it here—apart
from noting that it overlaps the latter, inasmuch as disclosure may itself
undermine independence in decision making.10
So my focus will be on Stevens’s remark about the interest in “inde-
pendence in making certain kinds of important decisions”—what might
be described as independence of judgment. The phrase “certain kinds of
important decisions” invites the question: Which kinds of decisions? One
familiar kind of answer strikes me as not very helpful. In particular, I can-
not see how we could possibly identify the private arena with the family,
or with the economy, or with any arena of social life, identified—either
spatially or institutionally—prior to normative political argument. My
doubts are founded on a familiar line of thought. Once we acknowledge
that the organization of any sphere is in part a result of collective deci-
sions, and may have decisive bearing on the lives of equal citizens, then
we cannot adopt a hands-off policy to a sphere identified in advance. In
the United States, the New Deal taught this lesson about the economy;
and feminism, both as a movement and a theory, taught this lesson about
the family.11 Thus the thesis that a decision is private, and ought not
therefore to be regulated except for especially compelling reasons, is not
8. See John Ely, Democracy and Distrust (Cambridge, MA: Harvard University Press, 1980);
and Cass Sunstein, One Case at a Time (Cambridge, MA: Harvard University Press, 1999), esp.
250–252.
9. Roe v. Whalen, 429 U.S. 589, 599–600 (1977).
10. If having an abortion, for example, meant that your name appeared in the newspaper,
then the decision about whether to have one would to that extent be less independent.
11. See the quote from Rawls about the public-private distinction at the start of the essay.
312
best understood as a premise in political argument—because the decision

is private, we should not regulate it—but as a conclusion of such argu-
ment.
3. How, then, are we to characterize the kinds of decisions that ought
to be protected as private? I propose that we do so by reference to the
weight of the reasons that support them, in the judgments of those who
makes the decisions. I do not have a precise characterization of this
idea (I am not sure that a precise characterization is appropriate), and
I do not wish to suggest that there will be general agreement about the
relative weight of different sorts of reasons. Still, the idea strikes me as
essential.
Consider some examples that may help explain it. When I say that a
course of conduct is a matter of obligation, I claim that it has support
from an especially weighty reason. If, in contrast, I say that I prefer one
course of conduct to another, even that I prefer it intensely, I am down-
grading the kind of reason that supports the conduct. If I say that some-
thing is essential to a decent human life, then the reason has consider-
able weight. If I say that it is a good thing to do, then the weight is greater
than if it is a matter of preference (even intense preference) but is not
as substantial as when it is a matter of obligation or an essential of a de-
cent life.
Of course different philosophies of life will interpret these categories
differently and assign different content to them: such philosophies will
differ, for example, on the nature and content of our obligations to pre-
serve human life or on the scope and weight of duties of charity. The
essential point is that, once we accept the three essential elements of
the deliberative conception of democracy—reasonable pluralism, equal-
ity, and deliberative justification—then we need to let political ideas of
burdensomeness track the weight of reasons within the reasonable views
of those we are regulating. Thus it will not be acceptable to say that we
will make special efforts not to burden people by preventing them from
fulfilling their fundamental obligations, but then deny that their own
politically reasonable views about what those obligations are have any
weight—to say that we respect people’s obligation to keep the Sabbath,
but then to insist that, after all, the only Sabbath they have any obligation
to keep falls on Sunday.
4. Consider now how these observations apply to some of the issues
about life, death, and sex—issues that are arguably to be covered by a
313
constitutional privacy right and that have been at the center of the debate
about that right in the United States. I will be very brief on these extraor-
dinarily complex issues, because I mean to be illustrating the framework
rather than saying anything very distinctive about the issues themselves.
Thus, in the case of abortion, three considerations suggest the conclu-
sion that restrictive regulations are especially burdensome. First, there is
the burden on women’s equality if women are required to carry un-
wanted pregnancies to term. This point is acknowledged in the Supreme
Court’s 1992 Casey decision, with reference to the consequences of a
twenty-year adjustment to the Roe regime—“The ability of women to
participate equally in the economic and social life of the nation has been
facilitated by their ability to control their reproductive lives”12—but the
same point could have been made without the reference to issues of ad-
justment to Roe v. Wade.
Second, restrictive regulations burden women’s liberty, by restricting
choices of a deeply personal kind, thus impinging on a woman’s exercise
of personal responsibility about the course of her life. They represent a
substantial denial of judgment about the conduct of elementary aspects
of life. There is of course an equality aspect to the issue here, inasmuch
as no comparable denial is contemplated for men. But the relevant bur-
den would not be alleviated by generalizing it.
A third burden—the burden on judgment—results from the kinds of
reasons that commonly support decisions about continuing life and is es-
pecially pertinent here. In its Casey decision, the U.S. Supreme Court
says that “at the heart of liberty is the right to define one’s own concept of
existence, of meaning, of the universe, and of the mystery of human
life,”13 and it supposes that women take guidance from such judgments
12. Planned Parenthood of Southeastern PA v. Casey, 505 U.S. 833 (1992).

13. Ibid.; Justice Scalia disparages this remark as the “famed sweet-mystery-of-life passage” in
Lawrence et al. v. Texas, 539 U.S. 558 (2003). He finds two problems with it: laws, he says,
never restrict the right to define concepts (he has “never heard of a law” that restricts the “right
to define” concepts); and laws standardly restrict conduct based on a self-defined “concept of
existence.” Thus the famed passage is either vacuous or massively destructive (“the passage that
ate the rule of law”). But some laws do restrict beliefs—for example, they attach sanctions to
holding them—and laws restricting beliefs do not contain escape clauses that permit defen-
dants to plead that their belief is a matter of definition (“it is impermissible to deny the Trinity,
except when the essential unity of God is asserted to be a matter of definition”). As for “eating
the rule of law”: the famed passage asserts that conduct that is closely associated with certain
kinds of important beliefs (say, in the way that wine consumption is closely associated with be-
lief in the doctrine of transubstantiation) should only be regulated for especially compelling
reasons. To be sure, it is important to say which beliefs and in what way the conduct is closely
314
in making decisions about childbirth. And in Life’s Dominion, Ronald

Dworkin has characterized the relevant judgments as “spiritual” disagree-
ments about the relative importance of natural and human contributions
to the value of a human life.14
The burdens on equality and liberty are not especially controversial:
someone may think that they are overridden, but surely they are substan-
tial. Moreover, as to the burden on judgment, we can argue over how
precisely to characterize the relevant judgments, but however they are
characterized, the point is that they have considerable weight. Moreover,
we can debate how to explain the disagreements about these judgments
and offer views about why they have the weight that they do. But for the
purposes of political argument, this debate is unnecessary. Suffice to say
that the judgments are about when we have compelling reason to ensure
the continuation of life: that content indicates their weight and shows
why they are covered by a privacy right and can be overridden only by
reasons of considerable weight.
To be sure, that right could be overridden by an argument that abor-
tion is, as the papal encyclical Evangelium Vitae asserts, the taking of in-
nocent human life: such an argument would provide grounds for regula-
tion, despite the three burdens.15 But that argument cannot be made,
except by appealing to a particular outlook about the nature and value of
life that is rejected by many who are reasonable, politically speaking.
This criticism is driven not by the idea that judgments about the continu-
ation of life are in some way intrinsically private: instead, the claim that
they are private matters expresses the thesis that there is a strong case for
leaving women to their own judgment on these matters, in part because
of the weight of the reasons that support their conduct. Requirements of
reason giving under conditions of a pluralism of philosophies of life
drives the argument.
But there may appear to be a sleight of hand at work here. After all, the
pro-life and pro-choice positions seem completely symmetrical. That is,
either abortions are restrictively regulated or they are not. If they are re-
associated. But it is not clear why the effort to draw such distinctions is bound to undermine
the rule of law, except of course if the rule of law means a law of rigid rules. See Antonin
Scalia, “The Rule of Law as a Law of Rules,” University of Chicago Law Review 56 (1989): 1175–
1188.
14. Ronald Dworkin, Life’s Dominion (New York: Knopf, 1993), 91.
15. See Pope John Paul II, The Gospel of Life [Evangelium Vitae] (New York: Random
House, 1995).
315
strictively regulated, the side that favors restrictive regulation wins. If they
are not, then the side that has a liberal view wins. So yes, some people do
reject the justification for regulation: and let’s say that they are politically
reasonable. But why should their objections carry the day? After all, some
people reject the current settlement, which permits abortion. Why do
their objections to a permissive regulatory regime not carry as much
weight as the objections to a restrictive regime?
This alleged symmetry is illusory. The restrictive regime imposes un-
deniably substantial burdens on women’s liberty, equality, and indepen-
dence of judgment; in a society of equals, those burdens need to be justi-
fied, and the terms of that justification must carry some weight with those
whose liberty and equality are impaired. Otherwise, we fail to respect
them as equals. The restrictive regime cannot stand, then, because no
such acceptable justification is available.
The case for a right to die, understood as comprising a right to assisted
suicide, is importantly different. To be sure, there is a compelling case for
the conclusion that regulations of assisted suicide are burdensome: like
restrictive abortion regulations, they regulate conduct governed by judg-
ments about when we have compelling reason to ensure the continua-
tion of life. The considerations that support regulations in this area are,
however, of a different kind. In particular, the defense of such regulation
should appeal not to a conception of when life ends or what makes a
whole life (or parts of it) worth living but to concerns about when we
have a conclusive showing of a person’s willing decision to end her own
life, and worries about pressures to make that decision as a way to reduce
burdens on family and friends. A compelling case for the conclusion that
these decisions about death should be treated as private matters would
need to address these concerns. Absent such address, it is not unreason-
able for a democratic process to regulate, and thus to hold that the inter-
est in independent judgment, although present and substantial, is over-
ridden (though it would also be reasonable to have a more permissive
regime).
In the case of sexual intimacy, there is a strong case for concluding
that democracy’s public reason must treat it as a private matter—an arena
to be regulated by the independent and diverse judgments of individu-
als. On the one hand the reasons for regulation—a religiously founded
moral outlook—are arguably weightless because they are objectionably
316
sectarian.16 Moreover, the regulations are burdensome, for a reason that

Justice Harry Blackmun rightly identified in his powerful dissent in
Bowers v. Hardwick, and which the Supreme Court more recently em-
braced in Lawrence v. Texas: that intimate relations are an important set-
ting in which we do something of fundamental importance in a decently
lived life, namely, work out a sense of our identity. But this importance of
sexual intimacy—often one aspect of a personal relationship—is contin-
gent in part on its being guided by the judgments, feelings, and sensibili-
ties of the parties to it.17 Because of the importance of such intimacy, the
reasons that support it are substantial.18 So democracy’s public reason
provides a compelling case for treating this as a private matter. But, once
more, the claim that it is private does not serve as a premise in the case
against regulation. Instead it is a conclusion drawn from reflection on the
kinds of reasons that support the regulated conduct and the consider-
ations that might be used to defend a case for its regulation.19
Conventions
1. I turn now to the second privacy issue: the social conventions that mark
the boundaries between the public and the private. This is a sprawling
territory, and to narrow it I will focus here on a discussion of these issues
by Thomas Nagel. According to Nagel, we ought to embrace strong
16. The majority opinion in Lawrence v. Texas notes that the moral reasons that support reg-
ulation have, among other things, a religious background (they are “shaped by religious be-
liefs, conceptions of right and acceptable behavior, and respect for the traditional family”), and
asserts their weightlessness, though the weightlessness does not turn on their provenance in re-
ligious morality.
17. Bowers v. Hardwick, 478 U.S. 186 (1986).
18. The majority in Lawrence v. Texas cites Casey’s “sweet-mystery-of-life” passage in ex-
plaining the importance of the conduct regulated by restrictions on sexual intimacy. Auton-
omy in intimate conduct is important because such conduct is tied to independent self-defini-
tion.
19. The issue of gay and lesbian marriage raises questions that extend beyond my concerns
here, though the considerations I have presented bear on the issue in three ways, which I
will simply mention. First, the regulated conduct is supported by important reasons. Second,
some of the considerations appealed to in arguments against legal status are weightless: for
example, the argument in the Vatican’s “Considerations Regarding Proposals to Give Legal
Recognition to Unions Between Homosexual Persons” that “marriage is holy, while homo-
sexual acts go against the natural moral law.” Third, although arguments about the “best inter-
ests of the child” (also in the Vatican statement) are certainly not weightless, they need to be
made in a sufficiently compelling way, given the importance of the regulated conduct. The
statement is available at http://www.vatican.va/roman_curia/congregations/cfaith/documents/
rc_con_cfaith_doc_20030731_homosexual-unions_en.html.
317
norms of reticence to preserve “smooth functioning” in the public do-

main when we don’t need collective decisions and cannot expect agree-
ment. Nagel’s discussion of these issues was animated by what he per-
ceived as overreaching in the American public debate surrounding Bill
Clinton’s impeachment—overreaching by the public, the media, and the
political officials who presumed that Clinton’s sexual appetites and hab-
its were fit subjects of public discussion. Prompted by the Clinton contro-
versy, Nagel’s argument proceeds well beyond Clinton and sex to a de-
fense of “cultural liberalism.” Though fugitive in its precise content,
cultural liberalism is (roughly) the view that we ought to respond to fun-
damental disagreements in our informal public culture with a restraint—
a reticence about criticism—parallel to that which we arguably ought to
respect in our formal system of law and politics. Let me explain.
In the formal political system—what Habermas calls the “political
public sphere”—where the stakes are authoritative, coercively enforced
legal regulations, we should not enforce a particular moral or philosophi-
cal outlook; but we ought to justify laws and policies by reference to val-
ues that can reasonably be embraced by reasonable people who endorse
different outlooks: that is the idea of democracy’s public reason. Simi-
larly, in the public culture more broadly—what Habermas calls the “in-
formal public sphere,” and Rawls “the background culture”—where the
stakes include discursive presuppositions, as well as social conventions
and norms backed by decentralized sanctions, cultural liberalism re-
quires that, as a general matter, we steer clear of controversial topics
about which we cannot expect to reach agreement and that do not
demand a collective decision. In the particular case of sex, Nagel says:
“We should stop trying to achieve a common understanding in this area
and leave people to their mutual incomprehension, under the cover of
conventions of reticence.”20 Nagel extends these points to religion, and
concludes, quite generally, that it would be “a good idea to leave the pub-
lic space of a society comfortably habitable, without too much conflict,
by the main incompatible elements that are not about to disappear.”21 To
be sure, in a culturally liberal world, old-fashioned, “healthy mutual con-
tempt”22—for example between and among the communicants of differ-
20. Nagel, “Concealment and Exposure,” 23.

21. Ibid., 24.
22. Ibid.
318
ent religions—would flourish, but it would be obscured by tight-lipped,

ironic smiles.
So cultural liberalism means two things. First, it rejects communi-
tarian aspirations to control the public culture and to establish cultural
uniformity, and it requires instead what might be called a “cultural de-
mocracy,” in which all members are entitled to participate as equals in
shaping the public cultural environment. Second, cultural liberalism
takes a particular view about what such participation and cultural de-
mocracy involve: it commands reticence about the public expression of
fundamental disagreements. These two elements are independent; reti-
cence does not follow from opposition to communitarianism. To be sure,
the attractions of permanent Kulturkampf are themselves a matter of
persistent conflict. But those who are drawn to it—a group that argu-
ably extends from Machiavelli and Mill to the postmodernist friends of
an agonistic public realm—are all anticommunitarian. So we need addi-
tional argument in favor of cultural liberalism with its norms of reti-
cence as an alternative to cultural democracy with more open public
contestation yet with no expectation of agreement. And the case for it
does not seem very compelling. The connection observed earlier be-
tween democracy and liberalism (including rights of privacy) in the polit-
ical sphere does not extend to a comparable connection between democ-
racy and reticent liberalism in the cultural sphere.
To explain why not, I will make three points: about the costs of reti-
cence, about why the case for political reticence (a duty of civility) does
not extend to a more general case for public reticence, and about the
benefits of more open public argument even when no agreement is in
view.
2. As to the costs of reticence: let’s say that a strong convention of reti-
cence or restraint with respect to some area of disagreement requires that
we not get into that area at all (anyway, not outside the sphere of close
friends and intimates). Thus there might be a strong convention against
any open, public discussion of sex or religion or attitudes toward groups
different from one’s own. The case for such strong conventions of re-
straint is that they are mutually beneficial. Still—and as ever—legitimate
questions can be raised about the distribution of those benefits.
For example, we are all better off with a norm condemning all discus-
sion of who is sleeping with whom, than with a completely permissive
view of sexual gossip. But although such a blanket condemnation is fa-
319
cially neutral, men—especially in positions of power—may be especially

large beneficiaries of this particular convention: they get multiple sexual
partnerships with reduced danger of discovery (because of the conven-
tion). For this reason, strong reticence may be perceived—in Nagel’s
words—“as a form of male mutual self-protection.”23
To appreciate the force of this observation about the distribution of the
benefits, note that a range of different conventions might be adopted to-
ward regulating discussion in this area—all weaker in the restraints they
impose on discussion but better for all than an unregulated free-for-all.
Consider the norm: “Don’t talk about who is sleeping with whom, unless
the target is a powerful man who should know better.” That norm—what-
ever its ultimate merits—would be better for all than the regime of open
gossip, would be less likely to be rejected as a form of male mutual self-
protection, and would be arguably helpful in undermining cultural sensi-
bilities that support the abuse of power.
The general problem is familiar from discussion of problems of social
coordination with mixed motives: we have a plurality of conventions;
each such convention is mutually beneficial compared with the absence
of normative regulation; different conventions are associated with differ-
ent distributions of benefits; and there is nothing obviously right about
the status quo distribution. Under these circumstances, to opt for strong
norms of reticence—to insist on strong cultural norms of privacy that
would effectively put existing conventions beyond challenge—seems
simply to be a matter of endorsing the current distribution and seems to
deny the importance of cultural democracy. Thus, it is right to worry that
the “elevation of reticence” is “too protective of the status quo and that it
gives a kind of cultural veto to conservative forces who will resent any dis-
ruption,” at least if we interpret reticence as requiring a generalized re-
straint on any contentious, open public discussion about a topic.24
Nagel acknowledges this concern: “Those who favor confrontation
and invasion of privacy will think it necessary to overthrow pernicious
conventions like the double standard of sexual conduct and the un-
mentionability of homosexuality. To attack harmful prejudices, it is nec-
essary to give offense by overturning the conventions of reticence that
help to support them.”25 But he appears to believe that these costs must
23. Ibid., 23.

24. Ibid., 25.
25. Ibid.
320
be paid—perhaps worrying, with Lord Devlin, that a rejection of regnant

cultural norms may put us on the slippery slope that ends in the rejection
of civilization itself, on the road to hell in a handbasket. I conclude in-
stead that there is something to classic worries about conventions of pri-
vacy: that under the color of neutral norms of decorum, civility, decency,
and sheer discretion, they effectively mute criticism of social practices.
3. Strong conventions of reticence, then, impose costs. But perhaps we
ought to bear the costs to avoid even greater damage: after all, less reti-
cence may impose greater pressures to conformity not more disruptions
of a stifling status quo. This is a large issue, and I can only address a piece
of it here. In particular, I want to consider whether the case for restraint
in political argument, aimed at the collective exercise of power, carries
over to the informal politics of a democratic public culture—to public
conversation and interaction quite generally.
Let’s start with the idea—associated with democracy’s public reason—
that there is a duty of civility that requires that we cast our justifications
for fundamental laws and policies on the common ground of political
values.26 Although it is permissible to offer justifications that draw on our
own philosophy of life, arguments that lie on the common ground of de-
mocracy’s public reason must—here I follow Rawls—be available and
offered “in due course.”27 Thus moral liberals must, for example, present
a case for a right to die that does not rely on the thesis that autonomy is
the supreme value or that individuals own themselves. They can also
present arguments that do depend on such premises. But the duty of ci-
vility requires that they, roughly speaking, also present (“in due course”)
some argument that lies on common ground. And it requires, too, that
when they present an argument on a fundamental political question, and
do appeal to their own comprehensive outlook, they are not to expose
the underlying philosophical, religious, or moral outlooks of their oppo-
nents, and subject those outlooks to challenge: they are beyond reproach.
So the duty of civility requires a kind of restraint or reticence in the politi-
cal arena with respect to matters of fundamental disagreement.
I accept the duty of civility, understood as applying to public political
argument. But its plausibility, thus applied, reflects two special features
of the political domain. First, political debate issues in collective deci-
26. On the duty of civility, see John Rawls, Political Liberalism (New York: Columbia Uni-
versity Press, 1996), 216; and John Rawls, “The Idea of Public Reason Revisited,” 768–769.
27. Rawls, “The Idea of Public Reason Revisited,” 793.
321
sions that are made in the name of all: the majority is the tribune of the
people. But the fact that laws are made in the name of all, in the name of
citizens as equals, naturally suggests that they ought to be justifiable—
and be shown to be justifiable—by reference to values that all can rea-
sonably accept, and that they should provide for everyone a secure place
in the debate: this, as I indicated earlier, is the deliberative interpretation
of the idea that democratic decisions should be suitably collective. When
the supreme political authority speaks in a democracy, it ought to be
manifest that it speaks in the name of equal citizens. In the informal pub-
lic culture, we find no such corresponding authoritative statement of a
conclusion: “cultural democracy” names a social practice of argument
among equals, not a form of authority. With no agent authorized to speak
for all, there is no corresponding case for restraint in the terms of debate.
Second, collective decisions apply to all and are backed by the collec-
tive power of the state. If we wish to reconcile this coercive imposition of
regulation with the recognition of citizens as politically autonomous self-
legislators, then the best we can hope for is that the justification of basic
laws and policies rests on political values that can be shared. But, once
more, there is no comparable coercive imposition in the case of the pub-
lic culture, and so no comparable requirement of restraint.
I have said that cultural democracy is not a form of authority, and that
it does not involve a coercive imposition of constraints comparable to the
formal political system. I do not, however, wish to deny that cultural de-
mocracy is a form of power or to deny that cultural norms constrain con-
duct. To the contrary. Equal standing in the system of cultural democ-
racy is important in part because of the importance of the common
environment of cultural norms and standards of taste and value: we do
not want to leave these matters in the hands of others, even if we can
count on an exercise of collective political authority that is governed by
democracy’s public reason. For we may still find the public culture and
the constraints it fosters deeply uncongenial. Thus Scanlon says: “I have
no desire to dictate what others, individually, in couples or in groups, do
in their bedrooms, but I would prefer to live in a society in which sexual-
ity and sexual attractiveness, of whatever kind, was given less importance
than it is in our society today.” Or again, “What I fear is not merely the le-
gal enforcement of religion but its social predominance.”28 In both cases,
28. Thomas Scanlon, “The Difficulty of Tolerance” in David Heyd, ed., Toleration: An Elu-
sive Virtue (Princeton: Princeton University Press, 1996), 30.
322
it is difficult to see how we can challenge the regnant cultural norms,

while remaining within the bounds fixed by a duty of civility extended to
the informal public sphere. Nevertheless, the two differences—about au-
thority and constraints—remain, and because of these differences, the
case for a duty of civility—for reticence about fundamental criticism—
seems less forceful in the informal public sphere. A democratic society is
a society of equals; but the implications of this idea for the public politi-
cal sphere differ from its implications for the informal public sphere.29
4. So strong reticence carries costs, and there is no case for it compara-
ble to the case for the duty of civility in the political arena. Still, suppose
we accept the fact of reasonable pluralism and reject the communitarian
aspiration of achieving agreement on fundamentals through open-ended
public discussion. Once we accept that we cannot, in any case, expect
agreement on these matters, why bother getting into them? What is the
point of a more open, democratic public culture, freed from strong con-
ventions of reticence on matters on which agreement is not in view?
Why put up with the conflicts ingredient in the informal public sphere;
that’s a cost, too. Why not reticence?
Consider three kinds of public speech, each of which might break with
strong norms of reticence, none of which is presented with any expecta-
tion of achieving deeper agreement, but all of which can contribute
something important to informal public argument: bearing witness, ex-
pressive presentation of information, and fundamental criticism.
In bearing witness, I present my stand on an issue, act from a sense of
personal responsibility in presenting my view, and may feel it necessary
to condemn opponents’ views at their roots. I do not aim to change
minds or to present someone else with information that they lack but
only to state my convictions.
In the expressive presentation of information, people present their hos-
tility to some aspect of the public culture that they find stifling, oppres-
sive, humiliating, or demeaning, in view of who they are—in view of
some particular aspect of their identity. The form of this discourse is, “As
an X, I feel. . . .” I describe this as an expressive presentation of informa-
tion because I want to emphasize that others may learn something from
29. On this point, see my discussion of the distinction between equal opportunity for elec-
toral, political, and public influence in “Money, Politics, Political Equality,” in Fact and Value,
ed. Alex Byrne, Robert Stalnaker, and Ralph Wedgwood (Cambridge, MA: MIT Press, 2001),
153–157.
323
the complaint, even if presented not with the aim of enlightenment but
simply as a way of expressing a hostile attitude toward a social practice,
perhaps with the suggestion that the speaker’s finding the practice to be
objectionable suffices to condemn it. The speaker need not expect a col-
lective response nor act in the hope of prompting agreement, but instead
may be seen as providing information about how a feature of the culture
is not universally appreciated.
By fundamental criticism, I mean an explicit attack on the moral, reli-
gious, or philosophical foundations of someone’s political outlook—pre-
cisely the kind of thing that violates the duty of civility when it happens
in political discourse. But there are three reasons for pursuing fundamen-
tal criticism rather than adopting a general policy of reticence as a re-
sponse to fundamental disagreements.
The first reason is, once more, simply informational: I may want to let
my opponents know that others who share the public culture with them
find their public displays of religiosity creepy, or find their views false or
childish.
Second, I may entertain the more ambitious thought that an awareness
of deeper and persisting disagreements will result in some modification
of opposing views, particularly if fundamental criticism is coupled with
the political respect reflected in the duty of civility. This point has partic-
ular force when there are political disagreements that we think are proba-
bly rooted in deeper religious or moral disagreements: certainly true in
issues about sexuality, life, and death that provided the focus of my dis-
cussion of privacy rights. If a duty of civility restrains us from pursuing
these disagreements in the political arena, it will be important that there
be a setting in which they can be openly aired. Here I offer a speculation
about the cultural basis of political democracy: namely, that we are more
inclined to keep political deliberation within the confines of democracy’s
public reason and to respect the duty of civility if we think that opponents
have been subjected to more open-ended pressure elsewhere. A vibrant
cultural democracy may be both supportive of a liberal political democ-
racy and philosophically consistent with it.
Third and finally, I may think that the public culture itself is now one-
sided: that it is far easier to express conventional religious sentiment or
to deploy conventional religious symbols than to express skeptical or hos-
tile views about religion, or to embrace unconventional religious views.
Think again of Scanlon’s examples. Believing that the public culture is
324
important—that cultural norms do constrain conduct—and rejecting the

idea of regulating the culture politically, I think it important that it be
open to expressions of fundamental disagreement that do not have a
proper place in politics. The point of such expressions is neither to pro-
duce agreement nor to create a comfortably communitarian public cul-
ture. Instead, the point is to create a public culture more fully expressive
of the pluralism that democracy promises.
To be sure, there are dangers here. As Nagel says, “Revolution breeds
counterrevolution.” Nagel is surely right about the United States: the cul-
tural revolution of the 1960s eventually bred a cultural counterrevolu-
tion, filled with nonsense about how the civilization was in much better
shape in the 1950s when it was dominated by white Protestant men.
More fundamentally, though, my argument here suggests an answer to
worries about counterrevolution in response to cultural democracy: with
strong protections of privacy rights acting as limits on the exercise of po-
litical authority by a society of equals, the informal public sphere of that
same society of equals can more easily afford to reject cautious reti-
cence—really, deference to the status quo—as a response to persistent
disagreement.
325
10
R E F L E C T I O N S O N D E L I B E R AT I V E D E M O C R A C Y
1. Introduction
For more than two decades, egalitarian-democrats have sought to de-
scribe a “post-socialist” political project. The socialist project, including
its social democratic variant, comprised a set of political values and an in-
stitutional and political strategy for advancing those values. The values
were egalitarian and participatory. The institutional models and the po-
litical strategy focused on the state.
Contemporary debate among egalitarian-democrats begins in the con-
viction that this approach is misguided and moves along two paths.
The first, growing out of appreciation of the state’s limits as an eco-
nomic manager is sometimes called “asset-egalitarianism.” The idea is to
shift the distribution of income by changing the distribution of income-
generating assets. It is an important idea, particularly with pressures from
globalization on income security, but not my topic here.1
A second—and this will provide my focus here—is more political.
Building on what I will describe as “radical-democratic” ideas, it seeks to
construct models of political decision in which “local” players can be in-
volved more directly in regulation and in collective problem solving,
with a center that coordinates local efforts, rather than dictating the terms
of those efforts. Thus, local—or, more exactly, lower-level actors (nation-
states or national peak organizations of various kinds; regions, provinces,
I presented earlier versions of this essay at the University of Southern California Law School
and as a keynote address at a Princeton University conference on deliberative democracy. I am
grateful to the audiences on both occasions for their comments and criticisms.
1. For representative ideas, see John Roemer, Equal Shares: Making Market Socialism Work
(New York: Verso, 1996); Samuel Bowles and Herbert Gintis, Recasting Egalitarianism: New
Rules for Communities, States and Markets (New York: Verso, 1999); Stuart White, The Civic
Minimum (Oxford: Oxford University Press, 2003); Bruce Ackerman and Anne Alstott, The
Stakeholder Society (New Haven: Yale University Press, 2000); Richard Freeman, The New In-
equality (Boston: Beacon Press, 1998).
326
R E F L E C T I O N S O N D E L I B E R A T I V E D E M O C R A C Y
or sub-national associations within these, and so on down to whatever

neighborhood is relevant to the problem at hand)—are given autonomy
to experiment with their own solutions to broadly defined problems of
public policy. In return they furnish higher-level units with rich informa-
tion about the goals as well as the progress they are making towards
achieving them. The periodic pooling of results reveals the defects of pa-
rochial solutions and allows the elaboration of standards for comparing
local achievements, exposing poor performers to criticism, and making
good ones (temporary) models for emulation.
This “radical-democratic” project builds on two distinct strands of
thought.2
With Rousseau, radical democrats are committed to broader partici-
pation in public decision-making—though not, as Rousseau supposed,
through regular meetings of a legislative assembly open to all citizens.3
According to this first strand, citizens should have greater direct roles in
public choices or, in a less demanding formulation, should participate in
politics on the basis of substantive political judgments and should be as-
sured that officials will be responsive and accountable to their concerns
and judgments. Though more participatory democrats disagree on the
precise locus of expanded participation, they all are troubled when dem-
ocratic participation is largely confined to a choice between parties com-
peting for control of government, and particularly when that choice is
not founded on clearly articulated substantive programmatic-political
differences between and among parties and pursued with the expectation
that the parties will be held accountable to their announced programs.
The underlying participatory idea is that citizens in a democracy are to
2. I have written elsewhere on this radical-democratic project: with Joel Rogers (on associa-
tive democracy), Chuck Sabel (on deliberative polyarchy), and Archon Fung (on participation
and deliberation). In writing this essay, I have drawn freely on this joint work, and I am very
grateful to my co-authors for the collaborations that produced it. See Joshua Cohen and Joel
Rogers, Associations and Democracy (London: Verso, 1995); Joshua Cohen and Charles Sabel,
“Directly-Deliberative Polyarchy,” European Law Journal, vol. 3, no. 4 (December 1997): 313–
342 [reprinted as essay 6 in this collection]; Joshua Cohen and Joel Rogers, “Power and Rea-
son,” in Archon Fung and Erik Olin Wright, eds., Deepening Democracy: Institutional Innova-
tions in Empowered Participatory Governance (New York: Verso, 2003); Joshua Cohen and
Charles Sabel, “Global Democracy?,” New York University Journal of International Law and
Policy 37, 4 (2006): 763–797; Joshua Cohen and Archon Fung, “Radical Democracy,” Swiss
Journal of Political Science 10, 4 (2004).
3. Rousseau himself explored other forms of democratic participation, particularly in his
Constitution of Poland, where considerations of size precluded direct citizen participation in
lawmaking.
327
engage with the substance of law and policy and not simply delegate re-
sponsibility for such substantive engagement to representatives.
Along with participation, radical democrats emphasize deliberation. In-
stead of a politics of power and interest, radical democrats favor a deliber-
ative democracy in which citizens address public problems by reasoning
together about how best to solve them—in which, at the limit, no force is
at work, as Jürgen Habermas said, “except that of the better argument.”4
According to the deliberative interpretation, democracy is a political ar-
rangement that ties the exercise of collective power to reason-giving
among those subject to collective decisions. Once more, we see substan-
tial differences among different formulations of the deliberative-demo-
cratic ideal. Some see deliberative democracy as a matter of forming a
public opinion through dispersed and open public discussion and trans-
lating such opinion into legitimate law; others as a way to ensure that
elections—or legislative debates, or perhaps discussions within courts or
agencies—are themselves infused with information and reasoning; others
as a way to bring reasoning by citizens directly to bear on addressing regu-
latory issues. But in all cases, the large aim of a deliberative democracy is
to shift from bargaining, interest aggregation, and power to the common
reason of equal citizens—democracy’s public reason—as a guiding force
in democratic life.5
In this essay, I will explore these two distinct strands of the radical-dem-
ocratic project—participatory and deliberative—though I will focus on
the deliberative because that is the topic here. My central point is that
participation and deliberation are both important, but different, and they
are important for different reasons. Moreover, it is hard to achieve both,
but the project of advancing both is coherent, attractive, and worth our
attention. I begin by presenting an idea of deliberative democracy. Sec-
4. Jürgen Habermas, Legitimation Crisis, trans. Thomas McCarthy (Boston: Beacon Press,
1973), 108. In this passage, Habermas is not describing an idealized democracy but a hypotheti-
cal situation suited to the justification of norms.
5. We now have many statements of the deliberative conception. For my own, which I draw
on here, see Joshua Cohen, “Deliberation and Democratic Legitimacy,” Alan Hamlin and
Phillip Petit, eds., The Good Polity (Oxford: Blackwell, 1989); “Procedure and Substance in
Deliberative Democracy,” in Democracy and Difference: Changing Boundaries of the Political,
ed. Seyla Benhabib (Princeton: Princeton University Press, 1996); “Democracy and Liberty,”
in Jon Elster, ed., Deliberative Democracy (Cambridge: Cambridge University Press, 1998);
and “Privacy, Pluralism, and Democracy,” Law and Social Justice, eds. Joseph Keim Camp-
bell, Michael O’Rourke, and David Shier. (Cambridge, MA: MIT Press, 2005). [All these es-
says are reprinted here.]
328
ond, I will sketch three attractions of deliberative democracy. Third, I dis-

cuss four lines of skeptical argument. Fourth, I sketch three tensions be-
tween deliberation and participation. Fifth, I consider two political and
institutional strategies for blunting those tensions. I conclude by men-
tioning two large challenges.
2. Deliberation
Carl Schmitt said that deliberation belongs to the world of the par-
liament, where legislators reason together about how to address public
problems. It does not belong to the world of mass democracy, where
ethno-culturally homogeneous peoples find leaders who pick the peo-
ple’s friends and enemies. According to Schmitt, “The development of
modern mass democracy has made argumentative public discussion an
empty formality.”6 Rejecting Schmitt’s view, as well as its more benign
contemporary progeny, deliberative democrats explore possibilities of
combining deliberation with mass democracy. And not just explore: we
are hopeful about the possibilities of fostering a more deliberative de-
mocracy.
Deliberation, generically understood, is about weighing the reasons
relevant to a decision with a view to making a decision on the basis of
that weighing. So an individual can make decisions deliberatively, a jury
has a responsibility to deliberate, and a committee of oligarchs can delib-
erate: deliberation, in short, is not intrinsically democratic. The “democ-
racy” in “deliberative democracy” is not pleonastic.
Democracy is a way of making collective decisions that connects deci-
sions to the interests and the judgments of those whose conduct is to be
regulated by the decisions. The essential idea is that those governed by
the decisions are treated as equals by the processes of making the deci-
sions. Democracy, as Tocqueville emphasized, is also a kind of society—a
society of equals—but I will be confining myself as a general matter to
the more specifically political understanding of democracy. Of course,
even if we think of democracy politically, as a way to make binding col-
lective decisions, constructing a more deliberative democracy is not a
narrowly political project: deliberative democracy requires attention to
encouraging deliberative capacities, which is, inter alia, a matter of edu-
6. Carl Schmitt, The Crisis of Parliamentary Democracy, trans. Ellen Kennedy (Cambridge,
MA: MIT Press, 1985), 6.
329
cation, information, and organization.7 I will return briefly to this point

near the end.
Deliberative democracy, then, combines these two elements, neither
reducible to the other. It is about making collective decisions and exercis-
ing power in ways that trace in some way to the reasoning of the equals
who are subject to the decisions: not only to their preferences, interests,
and choices but also to their reasoning. Essentially, the point of delibera-
tive democracy is to subject the exercise of power to reason’s discipline, to
what Habermas famously described as “the force of the better argument,”
not the advantage of the better situated. Deliberative democracy does not
aim to do away with power, an idea that makes no sense, nor does it aim
simply to subject power to the discipline—such as it is—of talking, be-
cause talking is not the same as reasoning (consider verbal assaults, in-
sults, racial slurs, lies, blowing smoke, exchanging pleasantries, exploring
common experiences); nor is it simply to reason together, because rea-
soning together may be without effect on the exercise of power.
Moreover, the notion of reason’s discipline is not nearly definite
enough. Plato’s philosopher-guardians subject power to reason’s disci-
pline—that, at any rate, is what they say they are doing. But deliberative
democracy is a kind of democracy, so the reasoning must be in some rec-
ognizable way the reasoning of the equal persons who are subject to the
decisions. And not just the process of reasoning but also the content of
the reasons themselves must have a connection to the democratic con-
ception of people as equals. Deliberative democracy is about reasoning
together among equals, and that means not simply advancing consider-
ations that one judges to be reasons but also finding considerations that
others can reasonably be expected to acknowledge as reasons. That’s why
deliberation focuses, as a constitutive matter, on considerations of the
common good, and also why—or so I have argued elsewhere—basic
personal liberties are essential elements of a deliberative democracy. De-
liberative democracy is not majoritarian, and these substantive condi-
tions—the common good and the personal liberties—are essential to
7. In Is Democracy Possible Here? (Cambridge, MA: Harvard University Press, 2006), Ron-
ald Dworkin emphasizes the importance of education in a well-functioning, deliberative de-
mocracy, and he asserts that “the most daunting but also most urgent requirement is to make a
Contemporary Politics course part of every high school curriculum” (148). The idea of such a
course is sensible enough, but it hardly seems the most urgent issue about the reform of our
educational system, for the purposes of fostering the partnership in argument that is so central
to democracy.
330
democratic deliberation, under conditions of reasonable pluralism. In

short, the ideal of deliberative democracy is to discipline collective power
through the common reason of a democratic public: through democ-
racy’s public reason.8
To be sure, discussion, even when it is founded on reasons, may not—
and often does not—issue in consensus. No account of deliberative de-
mocracy has ever suggested otherwise. All complex practical problems—
from trade and security to organizing schools and transportation, provid-
ing clean water and public safety, allocating health care and ensuring fair
compensation—implicate a range of distinct values, and reasonable peo-
ple disagree about the precise content of and weights to be assigned to
those values. In any allocative decision, for example, there are likely to be
people who think that the worst-off person should have priority, others
who think there should be equal chances, others who think the person
who benefits most should get the good. In allocating medical resources,
some will think that priority goes to the worst off; others, to those who
would benefit most; others will think we should assist the largest number
of people; others may hold that we should ensure that all people have fair
chances at receiving help, regardless of the urgency of their situations
and of expected benefits from treatment. So no matter how deliberative
the democracy gets, collective decisions will always be made through vot-
ing, under some form of majority rule. Indeed, deliberation may work
best when participants do not (as in a jury setting) feel the pressure to ad-
just their views for the sake of consensus—as if attention to reasons en-
sured convergence, and disagreement revealed bias or incapacity or some
other failure.9
There may be some temptation to think that the prospect of majority
rule defeats deliberation: because collective decision-making concludes
with a vote, participants—anticipating that final stage of resolution—will
not have any incentive to deliberate in earlier stages so will focus instead
on counting heads rather than on weighing reasons: aggregation at the
end of the day, then aggregation all day. But that temptation should be
resisted. Even if everyone knows that, at the end of the day, heads may be
8. As Rawls observes, an idea of public reason is one of the “essential elements of delibera-
tive democracy.” See “The Idea of Public Reason Revisited,” in Law of Peoples (Cambridge,
MA: Harvard University Press, 1999), 139.
9. On problems with deliberation under a unanimity rule, see David Austen-Smith and
Timothy Feddersen, “Deliberation, Preference Uncertainty, and Voting Rules,” American Po-
litical Science Review 100 (2006): 209–218.
331
counted, they may still accept the idea of arriving at a collective judg-
ment based on considerations that others acknowledge as reasons. They
may, for example, believe that reason-giving is an important expression of
respect or that it is the right way to acknowledge the collective nature of
the decision. If they do, they will be willing to deliberate in the stages
leading up to the vote, even when they know a vote is coming.
Deliberative democracy, thus understood, is a distinctive interpretation
of democracy: democracy, no matter how fair, no matter how informed,
no matter how participatory, is not deliberative unless reasoning is central
to the process of collective decision-making. Nor is democracy delibera-
tive simply because the process and its results are reasonable: capable of
being given a rational defense, even a rational defense that would be rec-
ognized as such on reflection by those subject to the decisions. The con-
cern for reasonableness must play a role in the process. Thus the contrast
between deliberative and aggregative democracy. In an aggregative de-
mocracy, citizens aim to advance their individual and group interests. If
the process is fair, the results may well be reasonable. But unless the rea-
sonableness is aimed at by participants in the process, we do not have de-
liberation.
Of course, it might be argued that reasonable results must be aimed at
to be achieved and that democracy must therefore be deliberative to be
reasonable. So, for example, if we have a hypothetical test for the right-
ness of decisions, where the hypothetical process involves reasoning un-
der idealized conditions about what is best to do, then it might be said
that the actual process must look something like the hypothetical to pro-
vide a basis for confidence in the rightness of results.10 Still, it is best to
see this connection between reasonableness and deliberation as a broadly
empirical claim and to keep deliberation as a way of deciding—a way
that comprises both the nature of the process and the content of the rea-
sons—distinct from reasonableness as a property of decisions.
Aggregative and deliberative democracy do not exhaust the space of in-
terpretations of democracy. Consider a community of politically princi-
pled citizens, each of whom endorses a conception of justice. The con-
ceptions they endorse differ, but each person accepts some conception as
setting bounds on acceptable policy and decent institutions. Assume fur-
ther that they do not see much point in arguing about what justice re-
10. See, for example, Jürgen Habermas, Between Facts and Norms, trans. William Rehg
(Cambridge, MA: MIT Press, 1996), 296, 304.
332
quires, though they discuss issues with an eye to generating information,

and each conscientiously uses his or her own conception in reaching po-
litical decisions. No one in this political community thinks that politics
is simply about advancing interests, much less a Schmittian struggle be-
tween friends and enemies. But reasoning together plays a very restricted
role in public political life: the members accept that they owe one an-
other an exercise of conscientious judgment but not that they owe a
justification by references to reasons that others might reasonably be ex-
pected to accept. I will not develop this distinction further here. I men-
tion it to underscore that the case for deliberative democracy needs to
be made not simply in contrast with accounts of democracy that focus on
interests and power but also in contrast with views that assume a consci-
entious exercise of moral-political judgment by individual citizens, al-
though not deliberation.
This emphasis on subjecting power to reason’s discipline is a thread
that runs through much of the literature on deliberative democracy.11
Thus, Amy Gutmann and Dennis Thompson say that “deliberative de-
mocracy’s basic requirement is ‘reason-giving.’”12 Jon Elster also empha-
sizes that deliberation is about argument—in fact, arguments addressed
to people committed to rationality and impartiality.13 John Dryzek says
that a “defining feature of deliberative democracy is that individuals par-
ticipating in democratic processes are amenable to changing their minds
and their preferences as a result of the reflection induced by delibera-
tion.”14 Elsewhere he emphasizes “communication that encourages re-
flection upon preferences without coercion.”15 But Dryek’s characteriza-
tions of deliberative democracy are not literally defining: they follow from
the more fundamental characteristics of deliberative democracy. The
point of deliberative democracy is not for people to reflect on their pref-
11. The emphasis on deliberation as reason-giving is not captured in models of deliberation
as cheap talk signaling, where the point is to convey some piece of private information and
success depends on beliefs about the trustworthiness of the speaker (see Austen-Smith and
Feddersen, “Deliberation, Preference Uncertainty, and Voting Rules”). For an interesting ef-
fort to model deliberation as reasoning—arguing from premises to conclusions, where individ-
uals can check the quality of the reasoning themselves—see Catherine Hafer and Dimitri
Landa, “Deliberation as Self-Discovery and Institutions for Political Speech,” Journal of Theo-
retical Politics 19, 3 (2007).
12. Amy Gutmann and Dennis Thompson, Why Deliberative Democracy? (Princeton:
Princeton University Press, 2004), 3.
13. “Introduction,” Deliberative Democracy (Cambridge: Cambridge University Press, 1998), 8.
14. Deliberative Democracy and Beyond (Oxford: Oxford University Press, 2004), 31.
15. Ibid., 8.
333
erences but to decide, in light of reasons, what to do. Deciding what to do

in light of reasons requires, of course, a willingness to change your mind,
since you might begin the deliberative task with a view about what to do
that is not supported by good reasons. But the crucial point is that Dryzek
emphasizes that deliberation is basically about reasoning—about rational
argument—and that other kinds of communication need to be “held to
rational standards.”16
3. Reasons for Deliberative Democracy

Why is deliberative democracy a good thing? It is of course hard to deny
that the exercise of collective power should be supported by appropriate
reasons. But deliberative democracy is not simply the undisputed idea
that the exercise of power should be rationally defensible, thus non-
arbitrary. The question is why it is important to discipline the exercise of
power by actually reasoning together. I will mention three consider-
ations.
The first is about promoting justice. Thus suppose we think that re-
quirements of justice are fixed by idealized reasoning under conditions of
full information and equal standing. One argument for deliberative de-
mocracy is that actual deliberation is needed if collective decisions are to
meet the standards of political right that would be accepted under ideal-
ized conditions of information and equality. So if justice is fixed by im-
partial reasoning in hypothetical conditions in which agents aim to jus-
tify principles to others, then, arguably, we will only achieve justice if we
make collective decisions using reasoning of a similar kind. We cannot
trust the achievement of justice to the pursuit of individual and group in-
terests, even under fair conditions.
A second line of argument is that reason-giving is a distinctive form of
communication and it may have desirable consequences, apart from pro-
moting justice. Thus, the requirement that I defend my proposals with
reasons that are capable of being acknowledged as such by others, will—
whatever my own preferences—impose some desirable constraint on the
proposals I can advance and defend. Of course if every proposal can be
rationalized in an acceptable way, then the requirement of defending
proposals with acceptable reasons will not have much effect: but I am
skeptical about this claim. Moreover, the need to give reasons that are ac-
16. Ibid., 167; and, in general, chap. 3.
334
ceptable to others might produce desirable consequences if reason-giving

itself changes preferences, or at least saliences. So while I start preferring
most what is best for me or my group, the practice of defending proposals
with reasons may change my preferences, dampening the tension be-
tween my beliefs about what is right or politically legitimate and my pref-
erences: not because that is the point of deliberation but because that is
its effect. In addition, deliberation may improve results by eliciting infor-
mation: though there are certainly truth-telling equilibria for strategic ac-
tors, I assume that the informational effects of deliberation depend in
part on a commitment to truthfulness or sincerity in communication,
which may itself be reinforced through deliberation, although it is hard
to construct from nothing. But that is true about the entire account of de-
liberation: though deliberation may reinforce a prior commitment to ar-
gue on terms that others can acknowledge as reasons, some such prior
commitment must be in place if the enterprise of mutual reason-giving is
to get off the ground and be sustained.
A third case for deliberative democracy, not about consequences, is
that the deliberative view expresses the idea that relations among people
within a pluralistic, democratic order are relations of equals. It requires
that we offer considerations that others, despite fundamental differences
of outlook, can reasonably be expected to accept, not simply that we
count their interests, while keeping our fingers crossed that those inter-
ests are outweighed. The idea of collective authorization is reflected not
only in the processes of decision-making, but also—as I said earlier—in
the form and the content of democracy’s public reason.
This point about the attractions of the deliberative interpretation of
collective decisions can be stated in terms of an idea of self-government.
In a deliberative democracy, laws and policies result from processes in
which citizens defend solutions to common problems on the basis of
what are generally acknowledged as relevant reasons. To be sure, citizens
will, as I mentioned earlier, interpret the content of those considerations
differently and assign them different weights. The reasons relevant to par-
ticular domains are complex and often competing, and there often will
be no clear, principled basis for ranking them: reasonable people may
reasonably disagree on how they should be weighted, even after all the
reasons have been aired. Nevertheless, they may accept the results of the
deliberative process in part by virtue of the process having given due con-
sideration to reasons that all reasonably accept.
335
When citizens take these political values seriously, political decisions

are not simply a product of power and interest; even citizens whose views
do not win out can see that the decisions are supported by good reasons.
As a result, members can—despite disagreement—all regard their con-
duct as guided, in general terms, by their own reason. Establishing such
political deliberation would realize an ideal of self-government or politi-
cal autonomy under conditions of reasonable pluralism. It may be as
close as we can get to the Rousseauean ideal of giving the law to our-
selves.
4. Skepticism about Deliberation

I want now to consider four objections that have been raised against the
deliberative conception of democracy. The interest of exploring the ten-
sions between deliberation and participation will be greater if some of
these concerns can be dispelled.
1. The first is about inequality. It begins with the observation that rea-
soning is an acquired capacity, one that is not equally distributed among
all. So collective decision-making through reason-giving may not neutral-
ize power, but it may, instead, create a “logocracy,” in which political
power is effectively shifted to the rhetorically gifted (or at least to the ver-
bally uninhibited), which may well compound existing social inequali-
ties and deliver political power to the educated, or economically advan-
taged, or men, or those possessed of cultural capital and argumentative
confidence.17
While the concern is important and understandable, the evidence,
such as it is, suggests that this objection exaggerates the feared effect, in
part by “depoliticizing” it—more precisely, by underestimating the ca-
pacity to recognize and alleviate it, should it arise. Democracy, to bor-
row a phrase from Jane Mansbridge, is always a work in process, and
much can be done to address this concern. Thus Archon Fung finds
that citizen participation in Chicago policing efforts is greater in poorer
neighborhoods (not a very large surprise, given crime rates in different
neighborhoods), and that the city, cognizant of obvious concerns about
cultural and class bias, invested resources in training participants in po-
licing and schooling efforts.18 Studying the case of participatory budget-
17. I believe that Lynn Sanders was the first to raise this objection, in “Against Delibera-
tion,” Political Theory (1997): 347–376.
18. Archon Fung, Empowered Participation: Reinventing Urban Democracy (Princeton:
Princeton University Press, 2004), chap. 4.
336
ing in Porto Alegre, Abers and Baiocchi find high rates of involvement by
poorer, less-educated citizens and discover substantial rates of participa-
tion by women and Afro-Brazilians.19 The thread running through these
and related cases is that participation is not exogenously given. Delibera-
tive bodies can undertake affirmative measures to address participatory
biases. In particular, they can help to train participants in the issues de-
cided by the body and in how to frame arguments about the relevant pol-
icies.
Now it might be argued that in the favorable cases just noted, the de-
liberative bodies aim to solve relatively concrete problems—to improve
policy in relatively well-defined areas (say, pertaining to the provision
of local public goods)—not to have an open-ended public debate. In-
equalities of argumentative skills on broader matters may resist remedy.20
But evidence from deliberative polling suggests otherwise: deliberative
capacities seem reasonably widely shared, even when issues are more ab-
stract and less locally focused. Critics of deliberation, it seems, were too
quick to conclude that deliberative decision-making empowers the ver-
bally agile.
2. A second objection is about effectiveness. Thus it might be said that
a deliberative process does not mitigate the effects of power on outcomes
of collective decisions. In addressing this issue, we face a large method-
ological problem. As a general matter, and putting aside the issue of de-
liberative democracy, it is hard to make an empirically compelling case
that process changes produce outcome changes, because changes in pro-
cess and in result may well both be produced by some third factor: as, for
example, when a party with a redistributive project empowers the less ad-
vantaged and promotes a shift in economic resources as well, thus sug-
gesting (incorrectly) that the change in process produced the change in
result.
A few studies, though not of deliberative democracy, have forcefully
addressed these problems of spuriousness. Ansolabehere, Gerber, and
Snyder have shown that court-ordered reapportionment in the 1960s
19. See Rebecca Abers, “Reflections on What Makes Empowered Participatory Governance
Happen,” in Fung and Wright, Deepening Democracy, 206, and more generally her Inventing
Local Democracy (Boulder, CO: Lynne Riener, 2000); Gianpaolo Baiocchi, Militants and Cit-
izens: The Politics of Participatory Democracy in Porto Alegre (Stanford: Stanford University
Press, 2005).
20. In his critique of deliberative democracy, Richard Posner is less hostile to locally focused
discussion about the provision of public goods, perhaps for reasons of the kind noted in the
text. See Law, Pragmatism, and Democracy (Cambridge, MA: Harvard University Press, 2003).
337
shifted public goods spending in the states in the direction of previously

underrepresented districts: a special case because reapportionment was a
court-ordered exogenous shock.21 Similarly, Chattopadhyay and Duflo
have made the case that reserved seats for women on Indian village coun-
cils have led to shifts in public goods spending, with greater spending on
goods that are preferred by women when the head of the village council
is a woman.22 Here the problem of spuriousness is solved by randomness
in the process that determines which village councils will be headed by
women. We have no comparably compelling case that increased deliber-
ativeness leads to changes in the content of the decisions.
Still, we have some suggestive evidence. Thus, participatory budgeting
in Porto Alegre and the village councils in Kerala appear to have pro-
duced substantial shifts in the allocation of public resources to the poor:
in Porto Alegre, for example, there is now full coverage of water and sew-
ers and there is a three-fold increase in school attendance.23 Similarly,
Lucio Baccaro has argued that internal democratic reform in Italian
unions produced large shifts in union policy in directions more favorable
to the interests of outsiders (pensions, employment, and regional devel-
opment issues). To be sure, the results in these cases may come not from
deliberation but from broader participation or the dominance of a left
party. But deliberation seems to be part of the story, both because deliber-
ation shifts preferences and because it shifts collective decisions by mak-
ing some proposals harder to defend: namely, proposals that cannot be
defended in public on the basis of acceptable reasons. (Baccaro makes a
good case that deliberation made the difference).24
3. A third concern is about deliberative pathologies. A social-
psychological variant of this concern says that group discussion imposes
normative pressure on group members: a variance-reducing pressure not
to be less extreme than the group median, and a mean-shifting pressure
not to be less extreme than the group mean. A cognitive story claims that
group discussion in a relative homogenous group is dominated by argu-
21. Stephen Ansolabehere, Alan Gerber, James M. Snyder, “Equal Votes, Equal Money:
Court-Ordered Redistricting and the Distribution of Public Expenditure in the American
States,” American Political Science Review (August 2002).
22. Raghabendra Chattopadhyay and Esther Duflo, “Women As Policy Makers: Evidence
from a Randomized Policy Experiment in India,” Econometrica 72, 5 (2004): 1409–1443.
23. Gianpaolo Baiocchi, The Citizens of Porto Alegre, Boston Review 31, 2 (March-April
2006).
24. Lucio Baccaro, “The Construction of Democratic Corporatism in Italy,” Politics and So-
ciety 30, 2 (June 2002): 327–357.
338
ments embraced by the majority, so that when people update on a rela-

tively homogenous argument pool, they consolidate. In either case, it is
bad for outsiders.25
These are very serious concerns, but at least in principle, the remedies
seem straightforward, whatever the likelihood of their adoption. If delib-
eration under conditions of homogeneity drives polarization, then it is
important to ensure that deliberative settings in some way reflect the
wider diversity: in some deliberative settings, the competitive quality of
the decision—when the issue at stake is the allocation of scarce re-
sources—engenders such expression. In other settings, ensuring diversity
of opinion may be a matter of institutional principle or the responsibility
of a moderator. In settings of group discussion, this might mean ensuring
that some time is devoted to expressing beliefs or judgments that are as-
sumed not to be shared by others in the group: ensuring that this hap-
pens seems to be well within the reach of moderators or participants
themselves.
Put more generally, the point is that studies of deliberative pathologies
need to be treated with some care. Those pathologies may emerge from
group decision-making conducted without efforts to avoid the pathologi-
cal results. So such studies may often be interpretable as sources of cau-
tionary notes and recommendations for improvement rather than being
interpretable as undermining the case for deliberation. That said, it is
also true that the more fragile deliberation is, the more structure that
needs to be in place to move from discussion to good deliberation, the
less confidence we can reside in the project of building a more delibera-
tive democracy. A naïve version of the deliberative ideal supposes that
people are waiting to deliberate and need only to get competitive politi-
cal structures out of the way. Deliberation may be a more fragile accom-
plishment.
4. The final objection is about naïveté concerning power.26 Because
constraints on what counts as a reason are not well defined, the advan-
taged will find some way to defend self-serving proposals with consider-
25. Cass Sunstein, “Group Judgments: Statistical Means, Deliberation, and Information
Markets,” NYU Law Review 80 (June 2005): 962–1049; and Tali Mendelberg’s very instructive
discussion of deliberation and small-group decision-making, in “The Deliberative Citizen:
Theory and Evidence,” in Political Decision-Making, Deliberation, and Participation: Research
in Micropolitics, vol. 6, ed. Michael X. Delli Carpini, Leonie Huddy, and Robert Shapiro
(Greenwich, CT: JAI Press, 2002), pp. 151–193.
26. See Cohen and Rogers, “Power and Reason.”
339
ations that are arguably reasons. For example, they may make appeals to
ideas of the common advantage but press a conception of the common
advantage that assigns great weight to a deeply unequal status quo. Or if
they fail in this, the advantaged will simply refuse to accept the discipline
of deliberation.
If this objection is right, then proposals for deliberative democracy that
are inattentive to background relations of power will waste the time of
those who can least afford its loss: those now subordinate in power. The
time and energy they spend in argument, laboring under the illusion that
sweet reason will constrain the power that suppresses them, could have
been spent in self-organization, instrumental efforts to increase their own
power, or like efforts to impose costs on opponents.
The complaint that deliberative democracy is touchingly naïve about
power betrays vertiginously boundless confusion.
First, the importance of background differences in power is not a criti-
cism of the deliberative ideal per se, but a concern about its application.
Deliberative democracy is a normative model of collective decision-
making, not a universal political strategy. And commitment to the nor-
mative ideal does not require commitment to the belief that collective
decision-making through mutual reason-giving is always possible. So it
may indeed be the case that some rough background balance of power is
required before parties will listen to reason. But observing that does not
importantly lessen the attraction of the deliberative ideal; it simply states
a condition of its reasonable pursuit.
Thus, in Habermas’s account of the ideal speech situation, or in my
own account of an ideal deliberative procedure, inequalities in power
are stipulated away for the sake of presenting an idealized model of delib-
eration.27 These idealizations are intended to characterize the nature of
reasoned collective decision-making and, in turn, to provide models for
actual arrangements of collective decision-making. But actual arrange-
ments must provide some basis for confidence that joint reasoning will
actually prevail in shaping the exercise of collective power, and gross in-
equalities of power surely undermine any such confidence. So discussion
that expresses the deliberative ideal must, for example, operate against a
background of free expression and association, thus providing minimal
conditions for the availability of relevant information. Equally, if parties
27. See Cohen, “Deliberation and Democratic Legitimacy,” and Habermas, Between Facts
and Norms, chap. 7.
340
are not somehow constrained to accept the consequences of deliberation,

if “exit options” are not foreclosed, it seems implausible that they will ac-
cept the discipline of joint reasoning and, in particular, reasoning in-
formed by the democratic idea of persons as equals. Firms retaining a
more or less costless ability to move investment elsewhere are not, for ex-
ample, likely to accept the discipline of reasoned deliberation about la-
bor standards, with workers as their deliberative equals.
Saying, “If you don’t listen to reason, you will pay a high price” is not a
joke: it is sometimes necessary to resort to destabilization, threats, and
open conflict as answers to people who won’t reason in good faith. A
sucker may be born every minute, but deliberative democracy is not a
recommendation that we all join the club. But if the willingness to rea-
son does depend on the background distribution of power, doesn’t that
defeat the point of deliberative democracy by reducing deliberation to
bargaining under a balance of power? Not at all. Once people do listen to
reason, the results may reflect not only the balance of power that de-
feated their previous imperviousness but also their attentiveness to rea-
sons that can be shared. If I need to drink some espresso to concentrate
hard enough to prove a theorem, it does not reduce theorem-proving to a
caffeine high. So similarly, paying attention to power and threats to exer-
cise it doesn’t reduce deliberation to bargaining. To suppose otherwise is
like thinking that if you need to trust your math teacher in order to learn
how to do a proof then there is nothing more to proof than trust. It con-
fuses conditions that make an activity possible with that activity itself.
5. Some Tensions between Participation and Deliberation

I started by noting two strands in the radical democratic tradition: partici-
patory and deliberative. But I have not said much at all about political
participation: deliberative democracy is about political reasoning, not the
breadth and the depth of participation. To be very brief: participation is
particularly important in connection with achieving fair political equal-
ity, because shifting the basis of political contestation from organized
money to organized people is a promising alternative to the influence
conferred by wealth. Similarly, expanding and deepening citizen partici-
pation may be the most promising strategy for challenging political in-
equalities associated with traditional social and political hierarchies.
Moreover, it may be important in encouraging a sense of political re-
sponsibility.
341
But participation is one thing, and deliberation is another, and they

may pull in different directions. Consider three sources of tension.28
1. Improving the quality of deliberation may come at a cost to public
participation. Suppose, for example, that legislators, regulators, and judges
were to embrace a deliberative form of decision-making. Instead of
seeking to advance the interests of their constituents or single-mindedly
maximizing their prospects of re-election, for example, legislators would
engage in reasonable discussion and argument about policies. Judges
would, in turn, reinforce the legislators by requiring explicit attention to
reasons in legislative and administrative decision-making. But doing so
might require insulation from public pressures.
2. Expanding participation—either the numbers of people or the range
of issues under direct popular control—may diminish the quality of de-
liberation. Initiatives and referenda, for example, allow voters to exercise
more direct and precisely targeted influence over legislation, policy ques-
tions, and even elected officials. But far from improving deliberation,
such measures—in part because they ultimately focus on a yes/no deci-
sion on a well-defined proposition—may discourage reasoned discussion
in creating legislation. And even bringing people together to discuss spe-
cific laws and policies may—with a homogenous collection of people or
a lack of commitment to addressing a common problem—diminish de-
liberation, as discussion dissolves into posturing, recrimination, and ma-
nipulation.29
3. More fundamentally, social complexity and scale limit the extent to
which modern polities can be both deliberative and participatory. Delib-
eration depends on participants with sufficient knowledge and interest
about the substantive issues under consideration. But on any issue, the
number of individuals with such knowledge and interest is bound to be
28. Diana Mutz explores a different tension between deliberation and participation in her
important book, Hearing the Other Side: Deliberative Versus Participatory Democracy (Cam-
bridge: Cambridge University Press, 2006). Mutz argues that deliberation among the diverse
encourages greater toleration but dampens participation because of a desire to avoid conflict
with the people to whom one talks. Participation, in turn, is animated by a sense of passion that
is dampened by deliberation. I am not sure that Mutz’s results extend outside participation in
highly competitive political settings. But the challenge she raises is deep and needs to be ad-
dressed.
29. See, for example, Derek Bell, “The Referendum: Democracy’s Barrier to Racial Equal-
ity,” Washington Law Review 54, 1 (1978): 1–29; Yannis Papadopolous, “A Framework for Analy-
sis of Functions and Dysfunctions of Direct Democracy: Top-Down and Bottom-Up Perspec-
tives,” Politics and Society 23 (1995): 421–448.
342
relatively small, and so the quality of deliberation will decline with the
scope of participation. Of course, knowledge and interest are not fixed,
and deliberation may improve both. Still, time and resource constraints
make it undesirable for any particular area of public governance to be
both fully deliberative and inclusively participatory.
6. And So?
These three tensions notwithstanding, public decision-making in liberal
democracies could become both more participatory and deliberative.
The challenge is to devise practical projects that can incorporate both.
Radical democrats have two broad strategies for achieving that aim,
which I will sketch in very broad strokes.
The first aims to join deliberation with mass democracy by promot-
ing citizen deliberation on political matters, in what Habermas calls the
“informal public sphere,” constituted by networks of associations in civil
society.30 Because such informal discussion does not aim at a practical
decision but—insofar as it has an aim—at an informed public opinion, it
can pursue an unencumbered discussion about political values and pub-
lic goals. Moreover, these dispersed discussions—one element of a politi-
cal society’s process of collective decision-making—are potentially very
broadly participatory, for they take place through structures of numerous,
open secondary associations and social movements. For this mix of mass
democracy and deliberation, the essential ingredients, apart from ensur-
ing basic liberties, are diverse and independent media; vibrant, indepen-
dent civil associations; and political parties that operate independently
from concentrated wealth and help to focus public debate. All of this ar-
guably helps to foster deliberative capacities—a point I mentioned earlier
and promised to return to. The marriage of open communication in the
informal public sphere with a translation—through elections and legisla-
tive debate—of opinion formed there into law provides, on this view, the
best hope for achieving a greater mix of participation and deliberation
under conditions of mass democracy and a rule of law.
Much of the attractiveness of this view, then, hinges first upon the
deliberativeness of discourse in the public sphere and then upon the
strength of the links between such deliberation and the decisions of legis-
lative bodies and administrative agencies. But because dispersed, infor-
30. Habermas, Between Facts and Norms, chap. 8; John Rawls, Political Liberalism (New
York: Columbia University Press, 1996), 14, 382–383.
343
mal public deliberation and public policy are only loosely linked, a more
participatory and deliberative informal public sphere may have little im-
pact on decisions by formal institutions. Citizen participation in the in-
formal public sphere, then, may be of limited political relevance, and the
marriage of reason with mass democracy may proceed in splendid isola-
tion from the exercise of power. To be clear: I am not here objecting to
this first approach, only pointing to a concern and a possible limitation.
A second radical-democratic approach builds on the distinctive practi-
cal competence that citizens possess as users of public services, as sub-
jects of public policy and regulation, or as residents with contextual
knowledge of their circumstances. The idea is to draw on these compe-
tencies by bringing ordinary citizens into relatively focused deliberations
over public issues. Typically, such strategies create opportunities for lim-
ited numbers of citizens to deliberate with one another or with officials to
improve the quality of some public decision, perhaps by injecting local
knowledge, new perspectives, and excluded interests or by enhancing
public accountability.
One approach randomly selects small groups of citizens to deliberate
on general political issues, such as laws and public policies. Citizen ju-
ries in the United States and planning cells in Germany, for example,
empanel small groups (12–40) of randomly selected citizens to discuss
issues such as agriculture, health policy, and local development con-
cerns.31 Fishkin and his colleagues have sponsored larger gatherings of
300–500 citizens—with randomization—to deliberate upon such issues
as the adoption of the Euro in Denmark, the regulation of public utilities
in Texas, and the extent of foreign assistance.32 On an ambitious analyti-
cal interpretation, post-deliberation polls provide insight into what the
people think about a policy issue. Political impact is another matter. As
with citizen juries and planning cells, their political impact—to the ex-
31. Julia Abelson, Pierre-Gerlier Forest, John Eyles, Patricia Smith, Elisabeth Martin, and
Francois-Pierre Gauvin, “Deliberations about Deliberative Methods: Issues in the Design and
Evaluation of Public Participation Processes” in Social Science and Medicine No. 57 (2003):
239–251; Ned Crosby, “Citizens’ Juries: One Solution for Difficult Environmental Questions,”
in O. Renn, T. Webler, and P. Wiedelmann, eds. Fairness and Competence in Citizen Partici-
pation: Evaluating Models for Environmental Discourse (Boston: Kluwer Academic Press,
1995): 157–174; G. Smith and C. Wales, “The Theory and Practice of Citizens’ Juries,” in Policy
and Politics 27, 3 (1999): 295–308; John Gastil, By Popular Demand (Los Angeles: University of
California Press, 2000).
32. For a sketch of polls and implications, see Bruce Ackerman and James Fishkin, Deliber-
ation Day (New Haven: Yale University Press, 2005), esp. chap. 3.
344
tent that they have impact—comes from their capacity to serve in advi-
sory roles and to alter public opinion or to change the minds of public of-
ficials.
Another strategy convenes groups of citizens to deliberate and develop
solutions to particular problems of public concern. Such participatory-
deliberative arrangements—characteristic in different ways of associative
democracy and directly deliberative polyarchy—differ from political ju-
ries in two main ways. Whereas political juries usually consider general
issues, such as economic, health care, or crime policy, these delibera-
tions aim to address more specific problems, such as the management of
an ecosystem, the operation of a public school or a school district, the
prevention of crime in a neighborhood, or the allocation of a city’s re-
sources across projects and neighborhoods. Whereas political juries re-
cruit impartial and disinterested citizens by randomly selecting them,
participatory-deliberative arrangements recruit participants with strong
interests in the problems under deliberation.
Because of the specificity of these arrangements, citizens may well en-
joy advantages in knowledge and experience over officials. In Chicago,
for example, residents deliberate regularly with police officers in each
neighborhood to set priorities on addressing issues of public safety, using
their background knowledge as a basis for deliberation. And in Porto
Alegre, citizens meet regularly at the neighborhood level to agree on pri-
orities for public investment (for example, street paving, sanitation, and
housing); the capital portion of the city’s budget is produced by aggregat-
ing the priorities that emerge from those deliberations.
Participatory-deliberative arrangements—in areas such as education,
social services, ecosystems, community development, and health ser-
vices—show promising contributions to political equality by increasing
popular engagement in political decision-making. As I mentioned ear-
lier, in Chicago’s community policing program, for example, participa-
tion rates in low-income neighborhoods are much higher than those in
wealthy neighborhoods. Similarly, poor people are substantially over-
represented in both the budgeting institutions of Porto Alegre and the
local development and planning initiatives in Kerala, India. Directly
democratic arrangements that address problems of particular urgency to
disadvantaged citizens can invert the usual participation bias in favor of
wealth, education, and high status. They can also, however, create large
potential political inequalities. If systematic and enduring differences—
345
in deliberative capabilities, disposable resources, or demographic fac-

tors—separate those who participate from those who do not, decisions
generated by participatory-deliberative arrangements will likely serve the
interests of participants at the expense of others.
The proliferation of directly deliberative institutions fosters democratic
self-government by subjecting the policies and actions of agencies such
as these to a rule of common reason. But these contributions to self-gov-
ernment are, however, limited by the scope of these institutions. Most
participatory-deliberative governance efforts aim to address local con-
cerns and do not extend to broader issues of policy and public priorities.
Moreover, there is the danger of administrative “capture”: that by enter-
ing the circuit of regulatory problem-solving with its pragmatic concern
about the effectiveness of policy, participating citizens and groups lose
their capacity for independent action and their sense of the importance
of open-ended reflection and morally motivated criticism and innova-
tion.33 They may become dependent on the state and its official recogni-
tion for power and resources, and their political horizon may come to be
undesirably confined by attention to policy constraints. If this is right,
then the alleged limitation of informal, society-wide deliberation—the
fact that its impact is so indirect—is really its virtue. The precondition of
the unconstrained discussion on which public deliberation depends re-
quires distance between a civil society’s associative life and the state’s de-
cision-making routines.
7. Final Reflections
So achieving both participation and deliberation is complicated. But be-
cause of their more direct bearing on the exercise of power, participatory-
deliberative arrangements have a particular promise as a strategy for
achieving the ends of radical democracy. Two large challenges, however,
lie on that path.
The first concerns the relationship between conventional institutions
of political representation and participatory-deliberative arrangements.34
Participatory-deliberative arrangements make it possible to address practi-
33. See Lucio Baccaro and Konstantinos Papadakis, “The Downside of Deliberative Public
Administration” (unpublished).
34. For discussion of the issues sketched here, see Cohen and Sabel, “Directly-Deliberative
Polyarchy”; Cohen and Sabel, “Global Democracy?”; and Joshua Cohen and Charles Sabel,
“Sovereignty and Solidarity in the EU,” in Jonathan Zeitlin and David Trubek, eds., Governing
Work and Welfare in a New Economy: European and American Experiments (Oxford: Oxford
University Press, 2003), 345–375.
346
cal problems that seem recalcitrant to treatment by conventional politi-

cal institutions. But those arrangements are not a wholesale replacement
of conventional political institutions: they have limited scope and limited
numbers of direct participants. To the extent that they are successful,
however, participatory-deliberative arrangements and conventional polit-
ical representation can be transformed and linked so that each strength-
ens the other. If such arrangements became a common form of local and
administrative problem-solving, the role of legislatures and public agen-
cies would shift from directly solving a range of social problems to sup-
porting the efforts of many participatory deliberations, maintaining their
democratic integrity, and ensuring their coordination. Conversely, those
who participated directly in these new deliberative arrangements would
form a highly informed, mobilized, and active base that would enhance
the mandate and legitimacy of elected representatives and other officials.
The second challenge is to extend the scope of radical democracy. Can
participatory deliberation help democratize large-scale decisions about
public priorities—war and peace, health insurance, public pensions, and
wealth distribution? One way to address these larger questions is to con-
nect the disciplined, practical, participatory deliberations about solving
particular problems—say, efforts to reduce asthma rates in a low-income
community, or efforts to provide decent medical care in New Orleans or
Los Angeles—to the wider public sphere of debate and opinion forma-
tion about the costs of and the access to health care and the importance
of health relative to other basic goods. Participants in direct deliberations
are informed by the dispersed discussions in the informal public sphere,
and those more focused deliberations in turn invest public discussion
with a practicality it might otherwise lack. The ambitious hope is that
citizens who participate in constructing solutions to concrete problems
in local public life may in turn engage more deeply in informal delibera-
tion in the wider public sphere and in formal political institutions as
well.
In the end, then, radical democracy—understood as an effort to com-
bine the values of both participation and deliberation—has promise to
be a distinctive form of democracy, in which the informal public sphere
and conventional democratic institutions are reshaped by their connec-
tions with participatory-deliberative arrangements for solving problems.
Whether it will deliver on that promise remains, of course, a very open
question.
347
11
TRUTH AND PUBLIC REASON
In 1958 I wrote the following:

There are no hard distinctions between what is real and what is
unreal, nor between what is true and what is false. A thing is
not necessarily either true or false; it can be both true and false.
I believe that these assertions still make sense and do still apply
to the exploration of reality through art. So as a writer I stand
by them but as a citizen I cannot. As a citizen I must ask: What
is true? What is false?
har old pinter , Nobel Prize Lecture (2005)
Political constructivism does not use (or deny) the concept of

truth; nor does it question that concept, nor could it say that
the concept of truth and its idea of the reasonable are the
same. Rather, within itself, the political conception does with-
out the concept of truth.
john r awls , Political Liberalism
D emocratic politics comprises, among other things, public discus-

sion about laws and policies on the basis of reasons of justice. How
large a part is not my concern here. I assume that such reasoning, mixed
I presented earlier versions at the Catholic University of Leuven; LUISS Guido Carli; the
Harvard Graduate Conference in Political Theory; the philosophy departments at Columbia
University, Union College, Northwestern University, and Cornell University; the September
Group; the Stanford Law School Legal Theory Colloquium; and as the 2008 Mala Kamm Lec-
ture at NYU. I am grateful to audiences on all these occasions, to the Editors of Philosophy &
Public Affairs, and to Bradley Armour-Garb, Alex Byrne, David Estlund, Erik Freeman, Sam-
uel Freeman, Barbara Fried, David Hills, Erin Kelly, Cristina Lafont, Jon Mandle, Sebastiano
Maffetone, Richard Miller, Ingrid Salvatore, Thomas Scanlon, Seana Shiffrin, and especially
to Paul Horwich for comments and suggestions. I also benefited greatly from generous and il-
luminating comments by Richard Rorty at the Stanford colloquium. I sketch his concerns be-
low in note 55.
348
T R U T H A N D P U B L I C R E A S O N
with bargaining and hectoring, confession and accusation, self-pity and

compulsive self-display, provides some part of democratic politics. Fo-
cusing on this deliberative part, I want to consider the role that the con-
cept of truth might properly play in it.
I will defend two conclusions about that role.
First, the concept of truth, and judgments and assertions deploying
that concept—including judgments and assertions that apply the concept
to basic principles of justice—have a legitimate role to play in public, po-
litical argument. Here I endorse Pinter’s Nobel lecture view about the
place of truth in the reflection of citizens (not his comments on the artis-
tic license to violate the law of noncontradiction) and disagree with
Rawls’s claim in Political Liberalism that a political conception of justice
“does without the concept of truth.”1
Second, the conception of truth that plays a role in public justification
should be political. A “conception” or “understanding” of truth is a set of
claims about truth, though not offered as an analysis of the concept.2 By a
political conception of truth, then, I mean (very roughly) a set of claims
about truth—for example, that truth is distinct from warrant, and that it is
important—that is suited for the purposes of political reflection and argu-
ment in a pluralistic democracy, characterized by doctrinal disagree-
ments. I will explain later (section 3) the special aim and distinctive con-
tent of a political conception of truth. Suffice it to say here that the aim is
to present a view of truth that suffices for public reasoning and that could
reasonably be endorsed by the adherents of conflicting doctrines, which
may themselves employ richer conceptions of truth, for example, the
view that truth consists in a correspondence of truth bearer and fact, or in
some sort of idealized justification.
A political conception of truth is thus a genuine conception of truth,3
1. To be sure, Pinter was thinking about the truth and the falsity of assertions about matters
of nonnormative fact: for example, about the absence of weapons of mass destruction from Iraq
in 2002. Nevertheless, because he states his point in more sweeping terms, just as he had ear-
lier stated a perfectly general claim about truth, I think it is appropriate to interpret him as
making a more general claim about the place of truth in the discourse of citizens.
2. I say “conception” or “understanding” rather than “theory,” because I am not supposing
that a conception needs to have the internal structure that we associate with a theory. See be-
low, 372–373 (final two paragraphs of sec. 3).
3. In the way that the political conception of objectivity, associated with political liberalism,
is a genuine conception of objectivity, though different from, and in a way less committal than,
rational intuitionist and Kantian conceptions of objectivity. See John Rawls, Political Liberal-
ism (New York: Columbia University Press, 1996), 110–116.
349
although less committal than the conceptions that have traditionally oc-
cupied philosophical attention and that have deep roots in our prac-
tices of making and defending assertions (including logically complex
assertions), in our reasoning, and in ordinary understandings about the
content and the correctness of thoughts. Because a political concep-
tion is less committal, it bears some similarity to antimetaphysical, “de-
flationary” theories of truth. Those theories are inspired by the redun-
dancy theories of Frege and Ramsey, according to which the content of
the claim that a proposition is true is the same as the content of the prop-
osition.4 Their main thrust is that truth is not a philosophically deep idea;
the concept is fully captured by the infinitely many instances of the truth
schema (T):
(T) The proposition that p is true if and only if p.5
Deflationary theories thus deny that truth is correspondence, or war-

ranted assertibility, or consensus in inquiry’s ideal limit, or a cheerfully
solidaristic backslap. In denying these claims, such theories are concep-
tually deflationary and thus to be distinguished from the evaluatively
deflationary view that truth is not something to which we should at-
tach great importance. A political conception cannot endorse an anti-
metaphysical, conceptually deflationary theory. That endorsement is not
needed for public reasoning and would put the political conception
needlessly at odds with religious or philosophical views that comprise
more ambitious conceptions of truth. Endorsing an antimetaphysical
deflationism would make the political conception of truth unacceptably
sectarian. Instead, a political conception aims to be nonmetaphysical
rather than antimetaphysical.
Two points of clarification before proceeding. First, while disagree-
ing with Rawls’s claims about truth, I endorse the ideal of public reason.6
According to that ideal, political justification, at least on certain funda-
4. See Gottlob Frege, “The Thought: A Logical Inquiry,” Mind 65, 259 (July 1956): 293.
5. There are many versions of the deflationary approach and not all treat propositions (as op-
posed to sentences) as truth bearers: I have made that assumption here, but nothing in my dis-
cussion turns on it. For discussion of some of the variety, see Scott Soames, Understanding
Truth (Oxford: Oxford University Press, 1999), chap. 8; Bradley Armour-Garb, “Deflationism:
A Brief Introduction” (unpublished). On propositions as truth bearers, see Soames, chap. 1;
Paul Horwich, Truth, 2d ed. (Oxford: Oxford University Press, 1998), 16–17.
6. On public reason, see Political Liberalism, lecture 6; John Rawls, “The Idea of Public
Reason Revisited,” in Collected Papers, ed. Samuel Freeman (Cambridge, MA: Harvard Uni-
versity Press, 1999): 573–615.
350
mental questions, should proceed on a terrain of argument that can be

shared. In the case of democracy’s public reason, the shared terrain com-
prises values that can reasonably be shared by people who regard them-
selves as free and equal, despite irreconcilable doctrinal disagreements
and disagreements about justice. Although I find the idea of public rea-
son compelling, I disagree with Rawls’s claim that the concept of truth
finds no place in it. That claim makes the idea of public reason unneces-
sarily contentious, as if it were committed to the view that the truth about
justice does not matter. And it makes the idea of pubic reason hard to un-
derstand, because it proposes to leave the concept of truth behind while
preserving notions of belief, assertion, judgment, reason, and objectivity,
all of which are essential to an idea of public reason. In making his case
for putting truth aside, Rawls suggests that we face a dilemma: that we
can have truth or public reason, but not both. My aim in presenting a po-
litical conception of truth is to show that this is a false dilemma and that
we can have both. The political conception of truth, then, is presented as
part of a defense of the idea of public reason.
Second, I emphasize that the political conception is not, to borrow a
phrase from Rawls, “political in the wrong way.” A political conception of
truth does not say that a proposition is true just in case it solves our prob-
lems or is part of a broad consensus. Still less does it say that a proposition
is true just in case authorities say that it is true or it confers advantage on
the powerful. Such Thrasymachean theses about truth and politics are
best understood as critical claims about how the term “true” and its cog-
nates are used. They derive their critical edge from the fact that they do
not express conceptions of truth. Instead, they present the charge—un-
derstandable to anyone who has the concept of truth, troubling to any-
one who cares about it—that “true” is systematically applied to proposi-
tions in virtue of their satisfying a condition that does nothing to make
them worthy of belief.7 The political conception is not political in virtue
of tracing cognitive appraisals to an exercise of authority or to the effects
of power. It belongs to a very different conception of the political, and it
is classified as political because it can reasonably be endorsed as com-
mon ground for the purposes of consequential collective decisions.
To defend my two claims, I start (section 1) by sketching the back-
7. For discussion of the charge, and the ways that it puzzlingly deploys a passion for truthful-
ness against the concept of truth, see Bernard Williams, Truth and Truthfulness (Princeton:
Princeton University Press, 2002).
351
ground of Rawls’s idea that public reason “does without the concept of
truth.” Second, then, I discuss (section 2) why it is hard to understand the
idea that public reason leaves truth aside. To clarify the puzzlement, I
distinguish four views that diminish the place of truth in political argu-
ment: the No Concept view (Rawls’s position), No Truth Bearers, No Sub-
stantive Judgments, and No Big Deal. Third, I sketch a political concep-
tion of truth and offer an account of what having such a conception
available in public political argument comes to (section 3). Fourth, I crit-
icize three arguments that might be offered in support of the No Con-
cept view that truth does not belong to public reason (section 4). Finally,
I distinguish the idea that public reason does without truth from an
evaluatively deflationary view, which I will call the “cultural proposal.”
Drawing on this cultural-political distinction, I suggest (section 5) that a
political world that does without the concept of truth is missing some-
thing important.
1. Truth and Public Reason

The intellectual context for the discussion is provided by Rawls’s idea of a
specifically political liberalism. The phrase “political liberalism” may
have the air of a pleonasm. Liberalism has always been, inter alia, a polit-
ical outlook, defined by an emphasis on personal freedom, religious tol-
erance, open inquiry, the rule of law, social mobility, and, in its modern
formulations, democratic politics.
The point of the phrase, then, is that the liberal political outlook has
also often been presented as the political department of a broader philo-
sophical position, mixing an ethic of self-direction, religious latitudi-
narianism, suspicion about the normative force of tradition and author-
ity, metaphysical and epistemological modesty. But as, for example, the
twentieth-century Catholic accommodation to modernism illustrates, a
liberal political perspective can derive support from other philosophi-
cal starting points, including religious views that downplay autonomous
choice in favor of substantively correct decisions, regard tradition as a
deposit of evolving insight, are metaphysically and epistemologically am-
bitious, and see religious commitments as providing less latitude and
greater density of demands.8 Presentations of liberal political commit-
8. See “Declaration on Religious Freedom,” The Teachings of the Second Vatican Council:
Complete Texts of the Constitutions, Decrees, and Declarations (Westminster, MD: Newman
Press, 1966), 366–367; John Courtney Murray, “The Problem of Religious Freedom,” in Reli-
gious Liberty: Catholic Struggles with Pluralism (Louisville, KY: Westminster, 1993).
352
ments that tie them to a general philosophical liberalism may, therefore,

impose unnecessary barriers to a broad embrace of liberalism. To the ex-
tent that such presentations take hold in the political culture, they may
work to intolerantly exclude people from the political public sphere. Po-
litical liberalism aims to free the formulation of liberalism as a political
outlook, so far as possible, from that wider set of philosophical and reli-
gious commitments, and thus to “put no unnecessary obstacles in the
way of . . . affirming the political conception.”9 Moreover, it would honor
the value of tolerance and the ideal of the political arena as a space of
public reasoning open to equal citizens.
Rawls’s concerns about truth emerge from this background. To explain
how, I need to say a little more about why political liberalism might seem a
forceful response to doctrinal conflict, particularly in a democratic society.
The trouble with truth begins, then, with doctrinal pluralism, the plu-
ralism of incompatible “comprehensive doctrines”—views about the
world, the place of human beings in it, the appropriate ways to acquire
an understanding of the world, and the way that human beings should
accordingly live their lives. Doctrinal pluralism is not the only political
problem: conflicts of interest—or of ethnicity, race, or nation—are not, at
root, doctrinal. Moreover, doctrinal conflict is not only a problem. As
Mill and others have observed, it can work as a basis of social learning, as
we pool information drawn from the varieties of human experience and
reflect in light of thoughts and patterns of conduct that we would never
otherwise have entertained. Too, doctrinal conflict may be seen as hav-
ing some intrinsic attraction, as an expression of the scope of human pos-
sibilities.
Doctrinal pluralism is also a problem, however, and presents a distinc-
tive challenge in a democracy. Democracy is not simply a matter of liv-
ing together but also, ideally at least, a society of equals, whose members
decide together how to live together. People who disagree fundamentally,
however, have trouble occupying a common ground on which they can
justify to one another those joint decisions.
One response to this trouble is to try to characterize such a common
ground of political justification. I will call this “democracy’s public rea-
son.” Democracy’s public reason is a terrain of political reflection and
judgment that equal persons, drawn to conflicting doctrines, can reason-
ably be expected to occupy and endorse as a basis for addressing public
9. John Rawls, Justice as Fairness: A Restatement, ed. Erin Kelly (Cambridge, MA: Harvard
University Press, 2001), 37.
353
issues. The essential point here is that common ground, if it is available

at all, requires that the content of public reason is restricted relative to
the doctrines endorsed by members.
For example, concepts of self-realization, associated with the view that
there is an essential human nature that consists in the possession of cer-
tain self-governing powers, and of salvation, associated with the idea of a
transcendent God, are not available to public reason. Moreover, Rawls
thinks that the concept of truth is unavailable.
This claim about truth is not founded on any general philosophical
doubts about truth, nor on general doubts about the place of normative
notions in public reason. Ideas of rightness, justice, reasonableness, cor-
rect judgment, and objectivity all have a proper place in public reason.
His concern is specifically about public reason and truth, and he locates
this concern at the heart of his account of political liberalism and public
reason.10 When Rawls first proposed the idea of a conception of jus-
tice that would be “political, not metaphysical,” he defined a political
conception as one that avoids claims about truth: “The aim of justice as
fairness as a political conception is practical, and not metaphysical or
epistemological. That is [emphasis added], it presents itself not as a con-
ception of justice that is true, but one that can serve as a basis of in-
formed and willing political agreement between citizens viewed as free
and equal persons.”11 Later formulations of political liberalism remained
emphatic about steering clear of the concept of truth.12 “Political liberal-
10. In emphasizing Rawls’s concern specifically with truth, I disagree with Joseph Raz’s
claim that Rawls is equally concerned with a variety of terms of normative appraisal, including
“reasonable.” See his “Facing Diversity: The Case of Epistemic Abstinence,” Philosophy &
Public Affairs 19, 1 (Winter 1990): 15 (and esp. note 34). Raz’s article was written before Rawls
had developed a number of the main ideas of political liberalism, so the (mis)interpretation is
understandable. More broadly, I agree with Raz in thinking that Rawls was mistaken in putting
the concept of truth aside. But there are several important points of difference between his
view and mine: (1) Raz does not have the idea of a political conception of truth, and thus does
not see that doctrinal disagreement imposes any conditions on the understanding of truth that
falls within public reason (or perhaps see any need for an account of public reason at all); (2) I
think that Rawls, in emphasizing the importance of the political and the practical, was never
concerned simply with what Raz calls “consensus-based social stability.” Because Raz miscon-
strues Rawls’s concern with the political and the practical, he misinterprets the reasons for ab-
staining from judgments about truth; (3) Raz appears to endorse what I will later call the
Sufficiency and the Full Display arguments, an endorsement that may be associated with skep-
ticism about the idea of public reason.
11. John Rawls, “Justice as Fairness: Political Not Metaphysical,” in Joshua Cohen, Collected
Papers, ed. Samuel Freeman (Cambridge, MA: Harvard University Press, 1999), 394.
12. The passages that follow are from Political Liberalism, xxii, 116, and 94. In “The Idea of
Public Reason Revisited,” Rawls does not address the issue of truth at all, though he does not
suggest (except perhaps by the sheer omission) any change of mind.
354
ism,” Rawls says, “rather than referring to its political conception of jus-
tice as true, refers to it as reasonable instead.” Moreover, political liberal-
ism, he says, “has an account of objectivity” that is connected to an
account of “reasonable judgment” and suffices for the purposes of public
justification and “may leave the concept of a true moral judgment to com-
prehensive doctrines” (emphasis added). And, perhaps most fundamen-
tally, political liberalism “does not use (or deny) the concept of truth; nor
does it question that concept, nor could it say that the concept of truth
and its idea of the reasonable are the same. Rather, within itself, the po-
litical conception does without the concept of truth” (emphasis added).
Reasonableness, not truth, is the “standard of correctness,”13 and the ob-
jectivity of judgments about justice is characterized without reference to
the notion of truth. In response to Jürgen Habermas’s claim that political
liberalism cannot avoid issues of truth, Rawls forcefully replies that peo-
ple will “[c]ertainly . . . continue to raise questions of truth and to tax po-
litical liberalism with not discussing them. In the absence of particulars,
these complaints fall short of being objections.”14
Rawls’s point, I emphasize, is that the concept of truth is unavailable
(later I will contrast this No Concept view with a few other views that are
skeptical about a role for truth in political argument). Thus we cannot,
while operating within democracy’s public reason, make claims about
the nature of truth (and whether it has a nature), about its importance, or
about its relationship to justification, objectivity, and reasonableness. Nor
can we make assertions about the truth of any elements of our views, in-
cluding our comprehensive doctrine, or our views about justice, or our
understanding of how the society works.
Consider, for example, justice as fairness, with its two principles of jus-
tice. Assume that justice as fairness is one of several reasonable concep-
tions of justice that would win support among citizens in a democratic
society. The idea, then, is that it would be appropriate in political argu-
ment to think or assert that justice as fairness is reasonable—indeed the
“most reasonable” conception of justice15—and thus to use its principles
13. Political Liberalism, 127.
14. “Reply to Habermas,” in Political Liberalism, 394–395.
15. Political officials and candidates for office have a “duty of civility” to explain their posi-
tions on fundamental issues by reference to the political conception that they take to be most
reasonable. See “The Idea of Public Reason Revisited,” in Collected Papers, 576. The point is
not that it is advisable or obligatory to assert that one’s view is the most reasonable. Instead,
there is a duty to argue by reference to the view that one judges to be most reasonable. Judg-
ments about the reasonableness of one’s views thus belong to the terrain of political reflection
and argument.
355
in judging political arrangements.16 But it would be an inappropriate de-

parture from the norms of public reason to assert its truth, and it would
be inappropriate, in one’s capacity as a citizen assessing laws and policies,
to think that justice as fairness is true and to act on the basis of that
thought. More particularly, it would be inappropriate to affirm the truth
even of the proposition—uncontested among the competing reasonable
conceptions of justice—that individuals have a right to liberty of con-
science. And this would be inappropriate even though it is entirely in or-
der to assert that people have a right to such liberty as a matter of justice,
that it is reasonable to believe that they have a right to liberty of con-
science as a matter of justice, and that that thesis about the right to liberty
of conscience is objectively correct in virtue of being reasonable.17
Advancing claims about truth is, then, needlessly divisive: it under-
mines public reason and conflicts with the equal standing in public, po-
16. Rawls states that he takes justice as fairness, with its two principles, to be the most reason-
able conception. See Political Liberalism, xlvi.
17. Samuel Freeman proposed in correspondence that Rawls wishes to avoid claims only
about the truth of principles of justice not about the truth of consequences drawn from those
principles. We cannot, then, affirm, within public reason, that the principle of equal basic lib-
erties is true, nor can we affirm that it is true that justice requires equal basic liberties. If that
principle is, however, part of the most reasonable conception of justice, then there is no objec-
tion to saying about a law infringing freedom of worship that it is true that it is unjust. The ac-
ceptable claim of truth is made from “within” the conception of justice, whereas the unaccept-
able claim is made about the conception itself. The same distinction, however, does not apply
to the notion of being reasonable (or most reasonable). I do not see a case for interpreting
Rawls this way, and do not, in any case, see the basis for the proposed restriction on the con-
cept of truth. First, if we cannot say that the principles are true, can we nevertheless affirm the
principles, or are we limited to asserting that the principles describe the most reasonable con-
ception? If we affirm the principles, assert them, and believe them, what is the force of with-
holding the judgment that they are true? Second, suppose it is permissible to judge, from
“within” a conception of justice, that it is true that certain policies are unjust. Let’s say that is
because we confine claims about truth to what we can argue for from common ground (not
that I wish to accept that thesis). But then why is it not permissible to judge, from “within pub-
lic reason”—which provides a terrain of argument that different conceptions of justice share—
that certain principles of justice are true because they can be argued for from common
ground? Third, the case for confining truth to judgments made “within” a conception of jus-
tice seems to turn on a sharp distinction, reminiscent of Carnap, between internal questions,
which arise within a linguistic framework, and external questions, about whether to adopt a
framework. But Carnap’s distinction did not permit judgments about which framework or
which language (say, a thing-language or a sense-datum language) is “most reasonable.” Our
attitude to a framework, on Carnap’s account, is a matter of acceptance, not of belief or of as-
sertion: it is, he says, “not of a cognitive nature.” So it is nonsense to describe a framework as
true, but that is because the framework is not a matter of belief at all. See “Empiricism, Se-
mantics, and Ontology,” reprinted in Rudolf Carnap, Meaning and Necessity, 2d ed. (Chicago:
University of Chicago Press, 1956), 205–221. The idea that cognitive appraisal is entirely inter-
nal is foreign to public reason.
356
litical argument that democracy promises. “Once we accept the fact that
reasonable pluralism is a permanent condition of public culture under
free institutions, the idea of the reasonable is more suitable as part of the
basis of public justification for a constitutional regime than the idea of
moral truth. Holding a political conception as true, and for that reason
alone the one suitable basis for public reason, is exclusive, even sectar-
ian, and so likely to foster political division.”18
Rawls does sometimes hint, if only indirectly, an alternative view. “Ra-
tional intuitionism,” he says, “conceives of truth in a traditional way [em-
phasis added] by viewing moral judgments as true when they are both
about and accurate to the independent order of moral values.”19 Rational
intuitionism, which Rawls associates with Cudworth, Clarke, Grotius,
Pufendorf, and Locke, is a species of nonreductive moral realism, treats
our moral thought as an exercise of theoretical reason, and is arguably an
ingredient in much natural law theory.20 His remark about a “traditional
way” of understanding truth and an independent order of values suggests
that it endorses a correspondence theory of truth, and the phrase “a tradi-
tional way” suggests that other conceptions of truth are available. If there
is such an alternative, then it might have a role in public deliberation.21
Moreover, such a conception might fit with the spirit of political liber-
alism. Political liberalism distinguishes comprehensive doctrines, which
include moral ideas that guide people in all aspects of their lives, from
political conceptions, which comprise moral ideas and values expressed
in political judgments. I will not try to characterize the distinction here.
Suffice it to say that ideas with a role in comprehensive doctrines can
also play a role in political conceptions, if those ideas are given a “politi-
cal” interpretation, as with political conceptions of justice, the person,
objectivity, reasonableness, autonomy, and liberalism itself. The political
18. Political Liberalism, 129. I will come back later to this passage, which ties together two ar-
guably very different ideas, one about holding a political conception as true, the other about
holding it as the one suitable basis for public reason in virtue of its being true.
19. Ibid., 92. See also 111 (“in the familiar way”) and 114 (“this idea of truth”).
20. On rational intuitionism and the idea of an independent order of moral values, see Polit-
ical Liberalism, 91–92; John Rawls, Lectures on the History of Moral Philosophy, ed. Barbara
Herman (Cambridge, MA: Harvard University Press, 2000), 69–83, esp. 70–72.
21. When Rawls states that political liberalism “does without the concept of truth,” he iden-
tifies this as one of four elements in the contrast between rational intuitionism and political
constructivism. See Political Liberalism, 91–94. But, logically speaking, the alternative to en-
dorsing a “traditional conception” of truth is to not endorse a traditional conception. There are
two ways to do that: do without the concept of truth altogether, or endorse a nontraditional
(that is, noncorrespondence) interpretation of truth.
357
conceptions are intended, roughly, to play a role when citizens reason to-
gether about political affairs. Thus, a political conception of persons as
free says, for example, that no particular ends are mandatory or obligatory
from a public point of view, that obligations a person has in virtue of his
or her moral or religious outlook do not have public standing as obliga-
tions, and that civil standing does not shift with shifts in fundamen-
tal aims, no matter how much or how deeply a person’s self-conception is
bound up with those aims.22 This political conception of persons as free
is meant to be available to adherents of views that endorse distinct and
incompatible philosophical conceptions of the free person, as an autono-
mous self-legislator with a will sensitive to reflective judgments, or as cre-
ated in God’s image, subject to His laws, and free when in willing com-
pliance with those laws. Similarly, a political conception of liberalism as
assigning equal liberties to all is meant to be available to both proponents
and critics of the moral-liberal view that individual, reflective self-direc-
tion is essential to a good life.
Why not, then, a political conception of truth, understood as a con-
ception of truth suited to play a role in democracy’s public reason? Rawls
did not pursue this path. Perhaps he thought that the concept of truth is
unavoidably metaphysical, originally owned by Platonists, and passed
along to their rational intuitionist descendants. As applied to moral
thought, he might have supposed that truth comes packaged with the ra-
tional intuitionist’s notion of an independent order of values. Or per-
haps he thought that the alternatives are a correspondence theory, a
coherence theory, an antimetaphysical deflationism, or some other unac-
ceptably philosophically demanding theory. In any case, he proposed to
“leave the concept of a true moral judgment to comprehensive moral
doctrines.”
I think that Rawls is mistaken and that affirmations of the truth of prop-
ositions about justice are a perfectly legitimate part of democracy’s public
reason. But to address the animating concerns about the tensions be-
tween truth and public reason, we need a political conception of truth.
To understand why, it will help to say more about Rawls’s proposed re-
sponse to these tensions and about why that response is deeply puzzling.
22. On the political conception of the person, see Political Liberalism, 18–20, 29–35, 48–54,
and 86–88.
358
2. Doing without Truth?

How could we “do without the concept of truth”? Truth is of course
a controversial idea. But however it is best understood, the concept of
truth—like concepts of cause, object, fact, reason, and evidence—is
deeply rooted in our thought and reasoning. Truth is intimately linked to
the notions of belief and meaning, both fundamental in an account of
thought. Thus, beliefs are said to “aim at” the truth, in that truth is their
standard of correctness; correspondingly, coming to believe that p is not
true is typically “fatal” to the belief that p. Moreover, because truth is the
standard of correctness for beliefs, while we may come to believe that p
without deliberating about whether p (is true), when we deliberate about
whether to believe that p, we try to determine whether p (is true).23 As for
meaning, Donald Davidson has observed that we often figure out what
someone’s utterances mean by assuming that they are saying something
true, at least when they are saying something about publicly observable
surroundings.24
The notion of truth is also fundamental in our understanding of rea-
soning. Thus, truth is tied to judging, in that judging whether p is closely
connected to judging whether p is true (judgment is the “acknowledge-
ment of the truth of a thought”);25 to assertion, inasmuch as asserting is
commonly understood to involve presenting the asserted content to oth-
ers as true (“What distinguishes [truth] from all other predicates is that it
is always asserted when anything at all is asserted”);26 to assuming that p
for the sake of argument, in that when we assume that p we assume that it
is true; to reasons for believing, which are reasons for believing true; and
to ideas of a logical consequence, as a proposition whose truth is assured
by the truth of other propositions, and of deductive argument, whose spe-
23. On the relevant kind of fatality, see Williams, Truth and Truthfulness, 67–68. On truth as
standard of correctness, see Nishi Shah, “How Truth Governs Belief,” Philosophical Review
112, 4 (October 2003): 447–482; and Nishi Shah and David Velleman, “Doxastic Deliberation,”
The Philosophical Review 114, 4 (2005): 497–534.
24. See Donald Davidson, Truth and Predication (Cambridge, MA: Harvard University
Press, 2005), chap. 3.
25. Gottlob Frege, Basic Laws of Arithmetic, sec. 5, in Translations from the Philosophical
Writings of Gottlob Frege, ed. Peter Geach and Max Black (Oxford: Blackwell, 1970), 156.
26. Frege, Posthumous Writings, ed. H. Hermes, F. Kambartel, and F. Kaulbach, trans. P.
Long and R. White (Chicago: University of Chicago Press, 1979), 129. Neither this remark of
Frege’s about assertion, nor the previous remark about judgment, depends on his redundancy
theory of truth, much less on his view that true sentences denote the same object.
359
cial virtue is to be truth preserving. Truth is connected as well to norms

of thought and interaction that call for accuracy in representation, sin-
cerity in expression, consistency, “getting it right,” and being attentive to
how things are and not simply how we wish them to be.27
These observations are all familiar. I register them to distinguish the
concept of truth from a variety of other concepts that public reason might
arguably do without. Suppose someone proposes that public reason should
avoid concepts of salvation, or of self-realization, soul, personal auton-
omy, purity, courage, or honor. Whatever the merits of such abstention,
we have some idea of what is being proposed. We can imagine what it
would mean to conduct political justification without recourse to such
concepts. But the idea of locating a common ground of political re-
flection and argument that does without the concept of truth—like doing
without the concept of an object, or a cause, or a thought, or a reason, or
an inference, or evidence—is hard to grasp. Truth is so closely connected
with intuitive notions of thinking, asserting, believing, judging, and rea-
soning that it is difficult to understand what leaving it behind amounts to.
Keep in mind that what is at stake here is the concept of truth, not a
theory of truth. It is easy to see reasons for avoiding philosophical contro-
versy about truth: whether the truth of a proposition consists in its corre-
spondence with a world of facts that exist determinately and indepen-
dently or, instead, it consists in the beliefs on which ideally conducted
inquiry would ultimately converge; or whether, as the minimalist theory
claims, grasping the concept of truth consists in being disposed to assert
instances of the truth schema, or the Heideggerian, that truth is the dis-
closure of being.28 Such disagreements may be seen as needlessly divi-
sive, because they invite controversies that are politically idle. What mat-
ters is what we think is right or just, not what we think the truth of
normative propositions consists in.
There are good reasons for steering democracy’s public reason clear of
such philosophical controversies and for leaving it to competing tradi-
tions of thought to add their own philosophical interpretation of truth to
public discourse (if they have such an interpretation). Similarly, they
may add different interpretations of the nature of justice to a public un-
derstanding of justice or may add different accounts of the competence
27. On the good of accuracy and sincerity, see Williams, Truth and Truthfulness, chap. 3.
28. For an illuminating discussion of Heidegger on truth, see Cristina Lafont, Heidegger,
Language, and World-Disclosure (Cambridge: Cambridge University Press, 2000), chap. 3.
360
of reason to a conception of the politically reasonable. But the fact that

there is more to be said about justice does not exclude considerations of
justice from public reason; disagreements between empiricists and ratio-
nalists about the nature and powers of reason do not exclude a concep-
tion of the politically reasonable; and disagreements between Kantians
and rational intuitionists about objectivity do not keep the idea of objec-
tivity out of public reason. So, too, the fact that there is more to be said
about truth does not provide a rationale for excluding truth.
What could it be for the shared public terrain of argument to do with-
out the concept of truth? I have expressed some puzzlement about this,
but skeptical attitudes about the place of truth in political argument are
familiar. Is it really so puzzling? To explain why it is, I want to distinguish
the idea under consideration—I will call it the No Concept view—from
three other skeptical views: that truth has no role because no truth-apt
claims are in play (No Truth Bearers); that truth-apt claims are in play
and the concept is available, but we should refrain from applying it to
substantive political claims (No Substantive Judgments); and that truth-
apt claims are in play, the concept is available, and there is no problem
applying it to substantive normative judgments, but it is of no real inter-
est (No Big Deal). The distinctions will help to clarify the view and the
puzzlement.
1. No Truth Bearers. The idea that democracy’s public reason does

without the concept of truth must be distinguished, first, from a non-
cognitivist view of political discourse, akin to the views taken by classical,
metaethical noncognitivists about evaluative discourse generally (though
rejected by many contemporary noncognitivists). According to the classi-
cally noncognitivist emotivist, for example, the concept of truth has no
application to normative discourse, because no truth-apt claims are being
made.29 Statements with the surface form of assertions that are made in
29. Allan Gibbard initially defended an expressivist account of normative discourse and ar-
gued that such discourse is not fact-stating or truth-apt. See Wise Choice, Apt Feelings (Cam-
bridge, MA: Harvard University Press, 1990), 8, 10. For criticisms, see Paul Horwich,
“Gibbard’s Theory of Norms,” Philosophy & Public Affairs 22, 1 (1993): 67–78. More recently,
Gibbard proposes an expressivist explanation of normative discourse but recognizes that the
expressivist strategy—which starts by asking what states of mind normative claims express—can
help itself to notions of truth and fact. See Thinking How to Live (Cambridge, MA: Harvard
University Press, 2003), esp. chap. 9.
361
normative discourse are devices for emotional appeal, rhetorical manipu-

lation, and badgering.
If we say, with a variant of classical noncognitivism, that political argu-
ment is a matter of decisions (say, deciding on friends and enemies and
expressing the decisions) or expressions of attitude (cheering for your
side, shouting at the other) or words used to provoke behaviors, then the
concept of truth would arguably have no hold. Despite the surface forms
of political discourse—apparently asserting logically simple propositions
and embedding them in logically complex ones (“If slavery is not unjust,
then nothing is unjust,” or “If the fetus is a person, then abortion is mor-
ally wrong”), seeming to reason to and from those apparent assertions—it
would not involve claims that could be either true or false. Saying that a
political conception does without the concept of truth, then, would be
like saying that cheerleading does without the concept of truth. You do
without the concept because you are not trafficking in anything to which
the concept applies. Of course, cheers often have the form of assertions.
But when cheerleaders say, “We are number one” in the final minutes,
with the team down 46–0, no one thinks that they are really expressing
the belief that theirs is the best team.30
Whatever the merits of this view, it is not public reason. Rawls proposes
to leave truth out, but not because he is endorsing an interpretation of
political argument in which no claims are being made that are apt to be
either true or false. While public reason is to do without the concept of
truth, it is an exercise of practical reason, of reflection, and of judgment:
“So if the idea of reasoning and judgment applies to our moral and politi-
cal statements, as opposed simply to our voicing our psychological state,
we must be able to make judgments and draw inferences on the basis of
mutually recognized criteria and evidence; and in that way, and not in
some other way, say by mere rhetoric or persuasion, reach agreement by
the free exercise of our powers of judgment.”31
My point about the truth-aptness of claims made in the domain of pub-
30. Anticipating what will come later: If public reason includes the concept of truth, does
this impose a barrier for noncognitivists, who traditionally have thought that truth is not in play
in normative discourse? The answer lies in the content of the political conception of truth.
Noncognitivists have no reason to object to that use, given the relatively minimal commit-
ments that come with it. See Gibbard, Thinking How to Live, chap. 9; Simon Blackburn, Truth
(Oxford: Oxford University Press, 2005).
31. Rawls, Political Liberalism, 110–111. For a forceful statement of the point, see John Rawls,
Lectures on the History of Political Philosophy, ed. Samuel Freeman (Cambridge, MA: Harvard
University Press, 2007), 7.
362
lic reason can be put more strongly. The reason that the concept of truth
has no place cannot be that the claims made by a political conception are
not truth-apt; those claims must be truth-apt, even if the political concep-
tion itself somehow abjures the concept of truth. They must be, if there is
to be a common ground of argument under conditions of doctrinal dis-
agreement. To deny the truth-aptness of the claims made on the terrain
of public reason would offend against the essential idea of public reason.
That is because the very propositions advanced in public political argu-
ment—even if not taken as or presented in that context as true—might
be judged true by the religious or moral doctrine affirmed by a citizen.
Consider an example. Suppose I endorse a Catholic natural law view. I
will say with Dignitatis Humanae that it is true that citizens have a right
to religious freedom. But also, with Veritatis Splendor, I will say that “the
splendour of truth” is the foundation of morality and political justice,
that there is a “moral obligation, and a grave one at that, to seek the truth
and to adhere to it once it is known,” and that there is an “inseparable
connection between truth and freedom—which expresses the essential
bond between God’s wisdom and will.”32 Moreover, with the conviction
that Christ is “the way, the truth, and the life,” and, resisting the tempta-
tion of “detaching human freedom from its essential and constitutive re-
lationship to truth,”33 I will hold that the truth of the (political) proposi-
tion—that individuals have a basic right to religious freedom—follows
from the truth of an account of human dignity founded on the doctrine
that human beings, created in God’s image, are bound by His laws. If that
is all correct, then the political claim expressed on the terrain of public
reason—that there is a right to religious freedom—must be truth-apt. It
must be capable of being true or false, because that very proposition can
be judged to be true from the standpoint of Catholic doctrine, and it is
said to be derivable from the underlying truths about human dignity.
It would conflict with the essentials of public reason to deny any of
this: to deny that the proposition that individuals have a right to liberty of
conscience is true or to deny that its truth follows from the basic truths
about the right way to live as fixed by natural law. Lacking the concept of
truth, a political conception used for public reasoning can issue no such
32. Veritatis Splendor, paragraphs 34, 99. This is an encyclical by Pope John Paul II on
moral theology. See http://www.vatican.va/holy_father/john_paul_ii/encyclicals/documents/
hf_jp-ii_enc_06081993_veritatis-splendor_en.html.
33. Ibid., paragraph 4.
363
denial. At the same time, the Rawlsian proposal is that the very proposi-
tion about religious liberty that the natural law adherent affirms as true
and as a consequence of more fundamental truths is available for asser-
tion in political reflection and argument, available to be used as a prem-
ise in reasoning. The natural law adherent is simply not permitted to say
everything he or she believes about that proposition—in particular, that
it is true and a consequence of more fundamental truths.
In short, Rawls’s proposal is to endorse a cognitivist understanding of
political conceptions of justice and political argument on which notions
of judgment, reasoning, and argument are fully in play, while denying
the availability of the concept of truth within such conceptions. The con-
cern is specifically with the concept of truth, and the reason for leaving it
out cannot be that political discourse traffics in something other than
truth bearers. Someone might offer that rationale for the view that truth
has no place in political argument, but not if they think of political argu-
ment as an exercise of public reason.
2. No Substantive Judgments. A second view is that the contents of politi-

cal conceptions are truth-apt, the concept of truth is available, and cer-
tain kinds of judgments using it are permissible, but in public reason we
are to abstain from making substantive moral or political judgments us-
ing that concept. David Estlund appears to attribute this combination of
views to Rawls. The attribution is, I think, mistaken, but characterizing
the view will help in clarifying the No Concept view.
Estlund summarizes Rawls’s view this way: “Truth,” Estlund says, “is
held to be neither necessary nor sufficient for a doctrine’s admissibility
[in public political argument].”34 But the claim that the truth of a propo-
34. David Estlund, “The Insularity of the Reasonable: Why Political Liberalism Must Admit
the Truth,” Ethics 108 (January 1998): 253. Habermas also misinterprets the claim that the con-
cept is unavailable. He attributes to Rawls the view that “a theory of justice cannot be true or
false” and mentions a “weak interpretation” of this thesis on which it asserts that “normative
statements do not describe an independent order of moral facts.” See “Reconciliation Through
the Public Use of Reason,” Journal of Philosophy 92, 3 (March 1995): 123. A political concep-
tion of justice cannot make either of these claims, on Rawls’s account. Lacking the concept of
truth, it cannot make any claims about the truth or the falsity of conceptions of justice, includ-
ing the claim that they cannot be true or false. Moreover, endorsing the moral constructivist
thesis that normative statements do not describe an independent order of moral facts (the so-
called weak interpretation) would put it directly in conflict with rational intuitionism. See Po-
litical Liberalism, 113. What political constructivism does say is that “once, if ever, reflective
equilibrium is attained, the principles of political justice (content) may be represented as the
outcome of a certain procedure of construction (structure)” (Rawls, Political Liberalism, 89–
364
sition is neither necessary nor sufficient in licensing appeal to it in public

argument—if understood an assertion within public reason—belongs to
a very different view from the one that Rawls endorses and that we are ex-
ploring. Asserting that truth is neither necessary nor sufficient requires
having the concept of truth available. It is thus quite different from nei-
ther holding truth to be necessary nor holding truth to be sufficient for a
doctrine’s admissibility, which does not require the concept.
Consider a parallel. It is one thing to say that a view holds that con-
duciveness to salvation is neither necessary nor sufficient for the recti-
tude of conduct. It is quite another to say that it neither holds that salva-
tion is necessary nor holds that it is sufficient. You can only endorse the
first if you have the concept of salvation. But you can, consistent with
holding neither, lack the concept. So a public understanding of justice
for use in public reason does not hold salvation to be necessary nor hold
it to be sufficient for rectitude. But it would be a serious misunderstand-
ing to say that it holds that salvation is neither necessary nor sufficient.
Lacking the concept of salvation, the public understanding makes no
judgments about it at all.
If public reason works without the concept of truth, then, it cannot
hold that truth is neither necessary nor sufficient for admissibility as a
premise in political argument, much less make any claims about the
truth or the falsity of substantive propositions of political justice. Estlund
in effect, then, urges a modified political liberalism, which has the con-
cept of truth available, affirms that truth is neither necessary nor suf-
ficient for admissibility in public justification, also affirms as true the
proposition that reasonable acceptance is necessary for admissibility, but
does not apply the concept to substantive moral and political claims.
Adding this proposition about reasonable acceptance would be a natural
extension of a view that has the concept of truth available and that holds
truth to be neither necessary nor sufficient for a doctrine’s admissibility.
But this is a very different view from the No Concept conception.35
Rawls’s No Concept thesis is that public reason lacks the concept. No
90). The italicized phrase is essential. It would defeat the purpose of political constructivism to
say that the principles of justice are the product of a procedure of construction and, therefore,
“do not describe an independent order of moral facts.” The rational intuitionist may well agree
with the political constructivist claim about how the principles “may be represented,” while re-
jecting the moral constructivist claim about what those principles are.
35. In saying that it is very different, I am not criticizing the view. My aim in this section is
simply to clarify the No Concept account.
365
Substantive Judgments says that public reason has the concept but ab-
jures substantive applications of it. I will not discuss this view further.
Suffice it to say that the motivations for it are not very clear. After all, it is
perfectly consistent to say that reasonableness is necessary and sufficient
for the admissibility of a consideration into public reason—for its use in
public justification—and to say that it is permissible to present substan-
tive moral and political claims views as true. Once truth is available, it is
not clear why it should be cabined in the way suggested.
3. No Big Deal. A third view—normative-political positivism—accepts

a role for assertions of truth and falsity in political argument. It says, as
distinct from No Truth Bearers, that the contents presented in political ar-
gument are (at least sometimes) truth-apt. It says, as distinct from No
Substantive Judgments, that the application of the concept of truth to sub-
stantive moral and political judgments is legitimate. But because of the
normative positivism, truth does not play a substantial role in political re-
flection and argument.
To appreciate the point, consider Hobbes’s claim that “the authority of
writers, without the authority of the commonwealth, maketh not their
opinions law, be they never so true”: auctoritas non veritas facit legem.36
Hobbes’s thesis is about legal validity, not about justice. The idea is that
legal validity is fixed entirely by an act of authority and not at all by moral
rectitude. I think it is a useful, deflationism-inspired anachronism to
think that Hobbes expresses a point about the independence of legal va-
lidity from moral rectitude in convenient shorthand when he says that
truth does not make the laws. He might have said: If lying is right, that is
irrelevant to legal validity; if lying is wrong, that is irrelevant; if stealing is
right, that is irrelevant; if stealing is wrong, that is irrelevant; if kindness is
nice, that is irrelevant; if kindness if not nice, that is irrelevant; and so on,
ad infinitum. Instead, taking semantic flight, he generalizes over all prop-
ositions of any content and says that truth does not make the laws.
That position is legal-positivism, which is consistent with the view that
there are natural standards of rectitude to be used in evaluating laws.37
But Hobbes was arguably (only arguably) led to his legal positivism from
positivism about justice itself: auctoritas non veritas facit justitiam. As
36. Thomas Hobbes, Leviathan (Indianapolis: Hackett, 1994), 180.

37. See John Austin, The Province of Jurisprudence Determined (Amherst, NY: Prometheus
Books, 2000). According to Austin, the moral rectitude of laws depends on conformity with nat-
ural law, understood as God’s law for humankind.
366
Hobbes says, there are no unjust laws, because, antecedent to the sover-
eign’s law-making activity, there is no just or unjust distinction for laws to
be answerable to. So when Hobbes says that truth does not make law, he
means that legal validity does not depend on truths about rightness. But
that is in part because there are no normative truths available prior to au-
thority that might enter into determinations of legal validity. If that is in-
deed the rationale for legal positivism, then it follows as well that the
truth—that is, truths about what is just and unjust, right and wrong—
cannot figure in assessing valid laws as just or unjust, because the justice-
making facts, too, are exercises of sovereign legislative authority.
Once more, in semantic ascent, we might report this view by saying
that truth does not make justice. Nevertheless, the concept of truth can
be used in assertions about propositions concerning the justice of valid
laws. There is no trouble affirming that the proposition that theft is unjust
is true, no more trouble than in affirming simply that theft is unjust or
that theft violates the law. What makes theft unjust is that authorities
make laws defining and enforcing property rights. Nothing beyond that
illegality is needed to make theft unjust and therefore to make the propo-
sition that theft is unjust true. As a result, claims about the truth of propo-
sitions concerning justice are available in political argument, even
though truth does not make the justice of the laws.
Truth is available for the normative-positivist, then, but its relevance is
limited. The limits come not from the conception of truth as a device for
expressing generalizations but from the positivist view that justice is fixed
by the social facts of sovereign enactment. That view limits the force of
worries about whether our views about justice are true, about whether we
have it right about justice. The concern is intelligible. I might think that
I have been misinformed about the laws in a jurisdiction and so wonder
whether my beliefs are true. But once I know what the social facts are—
in particular, what the authority has decided—I have no basis for further
concern about whether justice is what I suppose it to be.
Drawing these points together, we have four views in play, each inter-
preting the general idea that, roughly stated, truth is not important in po-
litical argument:
1. No Concept: Though public, political justification is an exercise of

reason, and the contents presented are truth-apt, the concept of
truth is not available.
2. No Truth Bearers: The concept of truth has no application, be-
367
cause the contents presented in normative political discourse are

not truth-apt.
3. No Substantive Judgments: the contents are truth-apt, and the con-
cept is available, but the concept is to be left on idle when it
comes to substantive issues of justice.
4. No Big Deal: The concept is available and exercisable but does no
significant work, because it is simply available to report the results
of exercises of authority that are not constrained by norms of jus-
tice (or any other norms).
Rawls, I said, in effect suggests that we face a dilemma: that we can

have truth or public reason, but not both. He embraces the idea of public
reason as common ground under conditions of doctrinal disagreement,
and concludes that we should leave truth aside. But for the reasons I have
been suggesting, it is hard to see how we can. It is difficult to understand
what leaving truth aside comes to. Moreover, if political argument in-
volves (as the idea of public reason indicates) beliefs, assertions, judg-
ments, and reasoning, then truth also seems to be in play. And if justice is
not—as the normative-positivist holds—fixed by authoritative decision,
then there is plausibly something important about getting justice right. Is
there some way to reconcile these competing pulls?
To see how we might free ourselves from the dilemma, we need first to
have an account of what is involved in the idea that the concept of truth
is available to political reflection and argument.
3. Including Truth
I said at the beginning that political argument in democracies is in part a
matter of reasoning and judging on the basis of considerations of justice.
Such reasoning and judging might appeal, for example, to the principle
that all citizens are entitled to the same basic liberties, to the abstract idea
that a just society must treat its members as being of equal importance,
and to the arguments showing that a just society must secure equal basic
liberties in part because securing those liberties is essential to treating
members as being of equal importance. All of this—the principles as well
as the reasoning—lies, I will assume, on the terrain of democracy’s pub-
lic reason.
What, then, would it mean to say that a conception of justice—say, a
conception that includes the proposition that justice requires equal basic
368
liberties—includes the concept of truth? The issue is not about a word,

but about what it is for the concept of truth to be part of the conception.
Rawls says that a political conception of justice does without the concept
of truth, and leaves the concept to comprehensive moral doctrines. We
want to know what this “doing without” consists in, or, more precisely,
what it would be to “do with” the concept.
Think of a political conception, used to formulate arguments in public
reason, as a set of propositions. Let’s say it includes, among others, the
proposition that persons are entitled to a fully adequate scheme of equal
basic liberties, that fairness is a fundamental political value, and that the
original position models the ideal of fair cooperation among persons un-
derstood as free and equal. Consider a formulation of the conception in
English, including the word “true” and the biconditional. Speakers—by
virtue of their mastery of English—can say what propositions are ex-
pressed by various sentences; they can also say that the sentences express-
ing those propositions are true; and they can formulate biconditionals
connecting the propositions and their truth conditions. So, for example,
the sentence “Justice requires assurances of equal fundamental liberties”
expresses the proposition that justice requires assurances of equal funda-
mental liberties, and the proposition that justice requires assurances of
equal fundamental liberties is true if and only if justice requires assur-
ances of equal fundamental liberties. So the formulation of the concep-
tion enables them to state and endorse instances of the truth schema: the
proposition that p is true if and only if p. For example, they can say that
the proposition that justice requires equal basic liberties is true if and
only if justice requires equal basic liberties.
But more than this is needed to have truth available in public reason.
To see why, consider a deflationary theory of truth, which tells us that
truth is not a substantial property—not, for example, a relation of corre-
spondence between a truth bearer and an independent order of facts—
and that the truth schema tells us all there is to be said about the con-
cept of truth. I assume that that schema is not itself in dispute between
deflationary theories and their opponents. The point of contention is
whether the schema exhausts the concept, or if there is anything more to
the concept or nature of truth than is captured by the schema (or by a
truth-definition that entails instances of the schema).
Consider, in particular, a version of the deflationary theory that says,
among other things, that the truth schema, itself unexplained by any
369
deeper theory, fully explains everything about the property of being true;
that a grasp of the concept of truth consists in a disposition to affirm all
the instances of the truth schema; and that truth is not a substantial prop-
erty and lacks a nature. Moreover, it says that the sole point of having a
truth predicate is to be able to express generalizations: for example, to say
that everything that Einstein said is true, or that all propositions of a cer-
tain form are true, or that the truth of normative propositions is not rele-
vant to legal validity.38 If you hold such a minimalist theory, then you will
think that having the phrase “is true,” and all the sentences that can be
used to express instances of the truth schema are—assuming competent
speakers, who are disposed to affirm all instances of the schema—what
having the concept of truth in the political conception consists in.
But neither minimalism of this kind, nor any other form of defla-
tionism, can provide the account of what it is to have the concept of truth
in the political conception. Although the instances of the schema and
the inferences deploying the schema can be part of democracy’s public
reason, the minimalist theory of truth cannot be, and for two reasons
(parallel considerations apply to other forms of deflationism).
First, minimalism does not say, on its face, anything about the fact that
truth is important or about why it might be important: that our beliefs
and assertions ought to be true; that it is good that they be true. Under-
standing the equivalence schema and being disposed to assert its in-
stances while not knowing anything about the point of classifying propo-
sitions as true and as false is like knowing the rules of a game without
knowing that the point of the game is to win. Grasping the significance of
truth is arguably (I come back below to the force of this qualification) not
something you understand simply by virtue of understanding that the
use of the concept of truth is captured by the instances of the schema or
that mastery of the concept consists in a disposition to affirm those in-
stances.39 A conception of justice, even when its formulation is under-
stood to include the term “is true” and all the instances of the schema,
38. See Paul Horwich, Truth.
39. See Michael Dummett, “Truth,” in Truth and Other Enigmas (Cambridge, MA: Har-
vard University Press, 1978), 2–3; on the importance of giving an account of the point of the
concept of truth, see as well, Davidson, Truth and Predication; Williams, Truth and Truthful-
ness; and Crispin Wright, Truth and Objectivity (Cambridge, MA: Harvard University Press,
1992). Dummett focuses on Frege’s redundancy theory. “It is part of the concept of truth,”
Dummett says, “that we aim at making true statements,” but Frege’s account of truth “leaves
this quite out of account.” For criticisms, see Tyler Burge, “Frege on Truth,” in Truth, Thought,
Reason (Oxford: Oxford University Press, 2005), 130, and more generally, 88–93 (esp. note 3).
370
may nevertheless arguably be said not yet to include the concept of truth.
Including the concept as part of public reason also requires incorporating
some account of the point of judging claims to be true and of the classi-
fications that employ the concept of truth. I will return to this point in a
moment.
More fundamentally, second, minimalism is a contested theory about
truth. In contrast to correspondence, coherence, and pragmatic theories,
it says that truth has no nature and is not a substantial property. It thus
puts public reason needlessly at odds with philosophical doctrines—say,
the rational intuitionism associated with some formulations of natural
law theory—that embrace metaphysically more demanding theories of
truth. It says more than is necessary for the purposes of including the
concept of truth within democracy’s public reason. Minimalism is an
antimetaphysical theory of truth. In contrast, a political conception of
truth cannot—as I said earlier—be antimetaphysical. It should be, how-
ever, nonmetaphysical, which will give it significant overlap with an
antimetaphysical theory: it will not make claims about the real nature of
truth or affirm that it is substantial property, nor will it make claims about
what substantial property it is. But it needs to say less than minimalism
does so that it does not impose unnecessary barriers to entry onto the ter-
rain of public reason (it will not deny claims about the nature of truth
either).
In short, minimalism says too much (denies too much) and perhaps
says too little to serve as a conception of truth in public reason. For the
concept of truth to have a place in public reason, we need both more and
less. The political conception of truth needs, first, to avoid asserting any
theory about the nature of truth or its lack of a nature. Such assertions
impose unnecessary hurdles and seem to serve no purpose within public
reasoning.
Then, second, the political conception needs to include at least four
commonplaces about truth:40
• Attitudes: Believing (asserting, judging) is believing (asserting, judg-
ing) true, where this slogan is understood to mean that truth is the
norm governing beliefs, assertions, and judgments;
40. See Crispin Wright, “Truth: A Traditional Debate Reviewed,” in Truth, ed. Simon
Blackburn and Keith Simmons (Oxford: Oxford University Press, 1999), 203–238; and Wright,
Truth and Objectivity, chap. 2.
371
• Correspondence: True beliefs present things as they are (they “say of

what is that it is and of what is not that it is not”), and in that un-
controversial sense correspond to how things are, although it will
not add (or deny) that such beliefs present things as they really are
in themselves, determinately and mind-independently;
• Contrast: There is a distinction between truth and warrant or jus-
tification, so that an account of justice, for example, may be war-
ranted—we may have grounds for endorsing it—but may not be
true:41 although, to reiterate, it will not say that truth is a substantive
property different from warrant, nor will it offer an account of what
property truth is that distinguishes it from warrant; but it will avoid
claiming that nothing informative can be said on this issue;
• Value: Truth is important; and, given that truth is different from
warrant, that truth is important in a way that is different from the
way that warrant is important.
I mentioned before that minimalism may say too little to serve as an ac-
ceptable political conception of truth, and I need here to clarify the
point. Minimalists do not dispute these four claims but argue that the
truth predicate is a device of generalization, that our grasp of the concept
consists in our disposition to affirm instances of the truth schema, and
that these additional claims—for example, about the value of truth—are
derivative. They are explained in terms of the truth schema: according to
minimalism, everything about truth can be so explained.42 The political
conception of truth, in contrast, takes no position on whether the truth
schema is “explanatorily basic” in this way: it neither affirms nor denies
the minimalist’s explanation. It is agreed that an adequate account of
truth must in some way include the commonplaces about truth. The po-
litical conception of truth can simply treat these commonplace claims as
elements in the shared understanding of truth without claiming that they
41. A distinction of some kind between truth and warrant seems important to understanding
the practice of deductive argument. Deductive argument is both compelling and useful. It is
compelling because it is truth-preserving. But it is useful because often we are not anteced-
ently warranted in endorsing the conclusion of a deductive argument or warranted in endors-
ing the conclusion independently of its issuing from that argument. See Michael Dummett,
The Logical Basis of Metaphysics (Cambridge, MA: Harvard University Press, 1991), chap. 7.
42. Horwich, Truth; idem, “Norms of Truth and Meaning,” in Reflections on Meaning (Ox-
ford: Oxford University Press, 2005), 104–133; and “The Value of Truth,” Noûs 40 (2006): 347–
360, esp. sec. 10, on how to explain the desirability of truth without supposing that the concept
is “constitutionally normative.”
372
are part of the concept or without aiming at an explanation (or denying

the plausibility of some proposed explanation). In contrast, minimalism
either lacks an account of the value of truth, in which case it says too lit-
tle, or offers a theory about the value of truth, in which case it says too
much.
If our aim is to provide a philosophical theory of truth, then abstaining
from investigating the structure of this whole cluster of claims is bound
to seem unsatisfying, a dereliction of intellectual duty, or simple laziness.
But that is not the point of an account of public reason. Here, the aim is
simply to show that the concept, understood in a certain way, is available
as part of a shared ground of argument. We should not exclude the con-
cept, but we need not incorporate a philosophical theory of truth, any
more than we need to incorporate a philosophical theory of reason when
we include an account of reasonableness. We have an understanding that
serves the purposes of public political argument, but it does not go be-
yond those purposes.
4. Three Reasons for Keeping Truth Out of Public Reason

Having sketched a political conception of truth, I want now to consider
in more detail some considerations that might be advanced for excluding
truth from public reason, for what I have called the No Concept view. I
presented an intuitive case earlier (section 1), but I want now to discuss
three lines of argument, each of which exemplifies some general reason
for thinking that concerns about truth do not belong to politics, at least
not to the kind of deliberative politics on common ground under condi-
tions of doctrinal disagreement associated with the idea of public reason.
The first is that the “singularity of truth” makes it practically divisive, per-
haps intolerant, thus at odds with the animating concerns of public rea-
son; the second is that the inclusion of truth encourages the idea that
truth is sufficient, which in turn encourages sectarianism; and the third,
that truth introduces a concern with depth that is inappropriate to public
reason. Rawls and others suggest each of these three, and each has some
force. But with the political conception of truth available, we can accom-
modate the force while also preserving public reason.
Singularity. Assertions about the truth of a view may be seen as need-

lessly divisive, perhaps intolerant, because of a fundamental logical prop-
erty of truth, namely, that truth is singular, whereas reasonableness is
373
plural. Thus, inconsistent propositions cannot both be true, and their ad-
herents cannot therefore all believe the truth, whereas inconsistent prop-
ositions can both be reasonable to believe, and their adherents can all
hold reasonable beliefs. It is not a philosophical theory, but a common-
place about the concept of reasonableness that reasonable people dis-
agree about certain matters. The singularity argument turns this logical
distinction between truth and reasonableness into a rationale for drop-
ping truth while keeping reasonableness.
Rawls suggests the singularity argument when he says, “Holding a po-
litical conception as true, and for that reason alone the one suitable basis
of public reason, is exclusive, even sectarian, and so likely to foster politi-
cal division.”43 I will return later to the phrase “and for that reason alone.”
Putting it aside for now, the essential point is that if I assert that my ac-
count of justice is true, and if yours conflicts with mine, then I am com-
mitted to denying that yours is true. But I may assert that my view is rea-
sonable, while accepting that yours is, too. Why bring truth in, with this
potentially troubling exclusiveness? Hannah Arendt gives forceful expres-
sion to this concern, suggesting that truth claims are divisive in ways that
are hostile to political life: “Factual truth, like all other truth, perempto-
rily claims to be acknowledged and precludes debate, and debate consti-
tutes the very essence of political life. The modes of thought and com-
munication that deal with truth, if seen from a political perspective, are
necessarily domineering; they don’t take into account other people’s
opinions, and taking these into account is the hallmark of all strictly po-
litical thinking.”44
A first trouble with the singularity argument is that truth has close
competitors in the divisive singularity market. Even if truth is not in play,
other standards of appraisal seem not only divisively singular but also un-
avoidable, at least on the broadly cognitivist understanding of political
discourse associated with public reason. So, for example, while there are
a variety of different political conceptions of justice, and though reason-
ableness is plural, only one can be the most reasonable conception; “most
reasonable” is as singular as “true.”45 Indeed, even “more reasonable”
43. Political Liberalism, 129.
44. Hannah Arendt, “Truth and Politics,” in Between Past and Future (New York: Penguin,
1977), 241. For an instructive discussion of the complexities of Arendt’s views, see Linda Zerilli,
“Truth and Politics,” Theory & Event 9, 4 (2006).
45. Rawls says that citizens fulfill their duty of civility when they explain their positions to
other citizens by reference to “the political conception of justice they regard as the most rea-
374
might be seen as divisive, since conflicting views can both be reasonable,

but each cannot be more reasonable than the other. When, in political
reflection and argument, I rely on a particular conception of justice, I
must distinguish mine from others on some relevant dimension of ap-
praisal. Saying it is the most reasonable conception (or at least more rea-
sonable than others)—that it is more strongly supported by the range of
relevant reasons—is one such distinction.
Of course, it might be argued that the fluidly continuous more/less rea-
sonable distinction is less needlessly divisive than the rigid true/false bi-
nary. But closer inspection reveals that this observation has little force.
After all, I can say that someone has it more or less right, or is close to the
truth, or that what he thinks is approximately true, or more or less true,
but not quite. Indeed, in the spirit of nondivisiveness, I can say that it is
very reasonable, even if it is not true.
A second problem with the singularity argument lies in its supposition
that “is true” causes the trouble. We should resist shooting the semantic
messenger. I think that justice requires equal basic liberties; you think it
requires maximin basic liberties; I think that privacy is among the basic
liberties required by justice, while you do not. We agree that these views
come to the same thing in most circumstances, and I accept that what
you think is reasonable. Still, we disagree, and we seem to disagree not
simply about which view is “most reasonable” but also about what justice
requires. I think that unequal liberties are unjust, even when the inequal-
ity is associated with greater liberty for those with lesser liberty. We can-
not both be right, though our views may both be reasonable. I want,
then, to reject the claim that the concept of truth (or the deployment of
that concept) provokes disagreement. Its use, insofar as it is used in pub-
lic reason, expresses the disagreement that we have.
The first two responses to the singularity argument both emphasize
that we can have divisiveness and perhaps intolerance without truth. A
third observation is that we can have truth without divisiveness. Consider
again the passage in which Rawls suggests what I am calling the singular-
ity argument: “Holding a political conception as true, and for that reason
alone the one suitable basis of public reason, is exclusive, even sectarian,
sonable” (“Idea of Public Reason Revisited,” 576). To be sure, as Philip Kitcher reminded me,
several views may be tied for the most reasonable. The relevance of this is uncertain, because
you are still left with the judgment that some are less reasonable. Public reason is animated by
concerns about tolerance, not by an unwillingness to make judgments.
375
and so likely to foster political division” (emphasis added). The point is

that divisiveness does not come from the concept of truth, with its singu-
larity, but from the thought that there is one suitable basis for public rea-
son. If that is objectionable, then it is wrong to think that because a con-
ception is true, it alone has a role to play in public argument. But that
inference can be resisted, while preserving a place for the concept of
truth. It makes perfectly good sense to say, “My view is true, but other
views, while not true, are reasonable to believe, and what matters for de-
mocracy’s public reason is reasonableness not truth.” We need not drop
the concept of truth in order to drop the thesis—suggested in this pas-
sage—that the truth of a proposition is a necessary condition for its play-
ing a permissible role in public reason, or the thesis—to be investigated
next—that its truth suffices to license appeal to it in political reflection
and argument.
Sufficiency Argument. The sufficiency argument, with its concern about

fostering sectarianism, begins from the suspicion that the concept of
truth would serve in public reason as a general license for introducing
considerations that might otherwise be of suspect appropriateness. I call
this the “sufficiency argument,” because the claim is that a proposition’s
truth suffices to make an appeal to it appropriate in political justification.
If truth did work as a license, there would be troubles for including
truth within democracy’s public reason. But this provides a rationale for
leaving truth behind only if we think that the concept of truth, once
available, will end up serving as a rationale for treating any true proposi-
tion as a relevant consideration in political argument. Thus Rawls wor-
ries about a “zeal for the whole truth” and about people who think of pol-
itics as a “relentless struggle to win the world for the whole truth,” and he
emphasizes “that politics in a democratic society can never be guided by
what we see as the whole truth.”46 But should concerns about the whole
truth provide a rationale for excluding truth? Only if its inclusion leads to
an endorsement of its sufficiency.
Suppose someone asserts that abortion after eighteen weeks should be
stopped because, as Aquinas argued, God ensouls the fetus at quickening
thus transforming it into a living human being. Or consider the view that
46. Political Liberalism, 42; “Idea of Public Reason Revisited,” 766; Political Liberalism, 243.
Emphases added. A number of other passages in Political Liberalism also focus their concern
on appeals to the “whole truth.” See 216, 218–219, 225, 242–243.
376
gay marriage should not be permitted because it violates the duality es-
sential to God’s creation, or that social insurance ought to be eliminated
because the best human life is a life of personal independence at odds
with social insurance that protects against life’s risks. In each of these
three cases, an argument can be made for excluding the consideration
from democracy’s public reason, because proponents cannot reasonably
expect others to endorse the consideration. (I am not defending this idea
here: it is part of the background to the discussion.) But when presented
with that case, the proponent might say, for example, “But it is true that
God ensouls the fetus at quickening, and therefore true that abortion af-
ter 18 weeks is murder. How could these truths—universal and objec-
tive—not matter in deciding what to do?”
This appeal to truth carries no weight, and the availability of the con-
cept changes nothing. If the proposition that God ensouls the fetus at
quickening is not relevant, then the truth of that proposition is not rele-
vant. After all, the case for its being irrelevant in public justification
could not depend on the claim that it is false, since that claim would de-
feat the point of public reason. Correspondingly, the assertion that it is
true does not add anything to the case for its relevance.
This point—that if the proposition that p is not relevant, then it is true
that p is also not relevant—may seem to depend on endorsing a redun-
dancy theory of truth, according to which the attributing truth to a propo-
sition adds nothing to an assertion of the proposition itself. That is, the ar-
gument against truth as a license may be seen as semantic: If a person
who asserts the truth of some proposition says nothing more than a per-
son who asserts the proposition, then how could the former be relevant if
the latter is not? Consider a person who rejects the redundancy theory
and says that when he claims that the proposition about the fetus is true
he is adding something. He is asserting, he might say, that the proposition
corresponds to the facts about the fetus—to how things really are—and
that it is true that abortion is murder because God decided the issue and
our belief fits the facts as they are, independently of our decisions and
judgments. How, he might say, could that not be of decisive relevance?
These claims indicate that the proponent rejects the redundancy the-
ory but add no politically relevant argument. They tell us what makes it
wrong to have an abortion, from which it follows that it is true. If some-
one, then, argues for the relevance of truth along the lines just suggested,
the answer is that if the wrong-making facts did not suffice to establish
377
the relevance of the consideration, then the truth-making facts cannot,

that the understanding of truth as correspondence does not yield any ad-
ditional facts of moral importance. If God’s making abortion murder was
not relevant, given the constraints of public reason, then how could the
additional fact of correspondence with the divine-instituted wrongness
make the wrongness relevant? Truth, in short, has no power to lift an oth-
erwise irrelevant consideration into relevance.
A similar point is familiar from other settings and does not have to do
with the theory of truth or with any judgment about the importance of
the truth. Suppose I am on a jury, deliberating on a charge of theft. If I
mention in the deliberations that the defendant stole my wallet, I will be
ruled out of order, and it will not make any difference to insist, “But I am
telling you the truth.” The objection by the other jurors was to the perti-
nence of my claim in this setting, not to its veracity. And denying the rel-
evance of my claim to be speaking the truth does not depend on a partic-
ular theory of truth.
But these reasons for rejecting the claim that truth operates as a gen-
eral license are not reasons for steering clear of the concept of truth, un-
less the concept itself—or the practice of using it in political argument—
invites us to slide to an assumption of its sufficiency. There appears to be
some temptation to make this slide. For example, in explaining why
Rawls refrains from presenting his conception of justice as true, Raz says:
“Asserting the truth of the doctrine of justice, or rather claiming that its
truth is the reason for accepting it, would negate the very spirit of Rawls’s
enterprise.”47 By “accepting it,” Raz must mean “accepting it as an appro-
priate basis for public justification.” Thus interpreted, Raz is certainly
right that the claim that the truth of a proposition is “the reason for ac-
cepting it”—and contrapositively, that doubts about acceptability require
denial of truth—is at odds with the fundamentals of political liberalism
and of public reason. But if we were not tempted to slide from truth to
sufficiency, why would we think that this observation provided a good
reason to “refrain from claiming” that a view is true—and correspond-
ingly, that if we have good reason to think that assertions about truth are
permissible, then we ought to accept that truth is sufficient for accept-
ability?
The way to resist the slide from affirming the truth of concededly rele-
47. Raz, “Facing Diversity,” 9.
378
vant normative claims to the view that all truths are relevant, the slide
from nothing but the truth to the whole truth, is by pointing to the error
in that inference, not by denying a role for truth. The inference may be
tempting because of the thought that there would be no good reason to
affirm the truth of a proposition in the practical setting of a political dis-
cussion if the sufficiency thesis were not true, that it would be practically
idle to affirm the truth of a claim unless an affirmation of truth sufficed
to establish the relevance of a claim and the use of it as a conversation-
stopping rebuttal to all objections. But familiar reasons about the use of
truth as a device of generalization show that this is wrong.48 Consider
again the jury setting. Imagine that the jury hears from a particularly con-
vincing witness, and one of the jurors says: “I think that everything she
said was true.” (Assume that nothing the juror said was ruled out of or-
der.) Alternatively, one of the jurors might say that the essential thing the
jury needs to settle on is whether “Jones is speaking the truth.” The other
jurors agree, and that common conviction is a working premise in their
subsequent deliberations. But the fact that that working premise is ex-
pressed using the concept of truth does nothing to establish the relevance
of my claim to be speaking the truth when I say that the defendant
robbed me.
In short, truth may be relevant without being sufficient. Our (intellec-
tual) response to those who aim to win the world for the whole truth is
not to yield the concept of truth. Instead, the response is to underscore
the phenomena of reasonable doctrinal disagreement, explain the value
of a shared ground of argument among equals, point out that the case for
a shared ground is not founded on a skeptical or a relativist outlook, and
clarify the appropriate but limited role of judgments about truth on that
shared ground—thus denying the sufficiency of truth.
Limited Display. The “limited display” argument suggests that we should

leave the concept of truth out of public reason because its inclusion leads
to a concern with philosophical depth that is unsuited to public reason.
Suppose that everyone in a group accepts a principle of equal basic liber-
ties and accepts it in part because this principle would be agreed to by
48. For a sketch of the familiar point, see Horwich, Truth, 2–5. Horwich begins his discus-
sion by saying that “the truth predicate exists solely for the sake of a certain logical need.” The
political conception does not include any such story about the sole reason, but it can incorpo-
rate the rest of what is said.
379
the members as a way to live together as equals with conflicting funda-

mental religious and moral convictions. They all see the principle as sup-
ported by a set of reasons—about living together as equals, about permit-
ting conduct guided by basic convictions, about finding terms that others
can reasonably be expected to accept—that, we are assuming, are ac-
cepted by all.
At the same time, the common acceptance of this argument is founded
on different and conflicting doctrines about the bases of those shared rea-
sons. Now someone might say: “I see that the principle of equal basic lib-
erties is reasonable, and I see how the case for its being reasonable can be
presented by reference to a procedure of construction that brings to-
gether the relevant reasons. But is the principle of equal basic liberties
true?” This question can be interpreted in two ways. It might be under-
stood as asking whether justice requires some other principle, even if
equal basic liberties are an acceptable accommodation. Or it might be
understood as asking for a “full display” of the case for equal basic liber-
ties: not simply for a case that operates on common ground but for an ar-
gument issuing from more fundamental values, principles, and reasons.
The argument might, for example, be founded on the idea of human dig-
nity associated with creation in God’s image or on an idea of human au-
tonomy associated with the powers of reflective thought. The line of
thought underlying full display might go like this (suggested by general
norms governing assertion): in asserting a principle of justice to be true,
the speaker communicates to the listener that he believes that he has
some grounds for making that assertion. Moreover, the speaker commits
him or herself to presenting a justification that gives the case for the truth
of the principle. But such a justification must derive the principle (or the
conception of which it is a part) from true moral principles: that is, a full
display of the case for it. But such a derivation—a full display—cannot be
part of public reasoning.49 Since the assertion about truth commits the
speaker to the full display, and only a limited display is permissible,
claims about truth should be excluded.
The second interpretation is the relevant one, and it invites four re-
sponses. First, even if we agree that asserting that p, or asserting that it is
true that p, commits the speaker to offering a justification for the claim
that it is true that p, it does not follow that assertions of truth invite or
commit the speaker to a full display of the speaker’s entire ethical view
49. Habermas suggests that Rawls endorses the full display argument. See “Reconciliation
Through the Public Use of Reason,” 124–125.
380
(assuming that he or she has one)—that the rule, in the space of reasons,
is in for a penny, in for a pound. Recall that democracy’s public reason is
a terrain of argument, and it may suffice, when challenged, to present a
case for liberty of conscience, for example, that lies on common ground,
with shared premises. If the premises are assumed to be true, then noth-
ing more is needed to make a case for the assertion that that proposition
is true. An assertion about truth need not be understood as committing
the speaker to presenting a full display, any more than the assertion that
justice requires liberty of conscience needs to be understood that way.
Even if assertions commit a speaker to presenting a justification, they
need not express commitments to depth. Truth is not depth. Sometimes,
as Rawls observed, the point of philosophy is “to extend the range of some
existing consensus.”50
Second, whatever commitments about presenting reasons may follow
from asserting that a proposition is true presumably also follow from as-
serting the proposition itself. If we are required to get into excessively
deep waters when we assert that it is true that justice requires liberty of
conscience, we also get into those waters when we assert that justice re-
quires liberty of conscience. If there is a culprit, it is assertion (and the
norms governing it), not truth.
Third, assertions about truth may not call for any display of argument
at all. Suppose Smith says that he agrees with Mill’s harm principle but
not with Mill’s views about equality and democracy. If Jones says in re-
sponse that she thinks that more or less everything Mill said was true, I
cannot see that she has advanced any claim at all about the kind of argu-
ment that can or needs to be given for Mill’s conclusions—in particular,
and following on the second point, that she now has a justificatory bur-
den greater than the one she would have had if she had simply said that
she agreed with various claims Mill made about equality and democracy
and had not said anything about truth.
Fourth, even if claims about truth do invite a full display, there may be
nothing objectionable about presenting a full display of a doctrine that
leads one onto common ground. I might say that I have a case, founded
on a broader doctrine that underwrites my view that a principle of equal
basic liberties is true. At the same time, I am fully aware that others dis-
agree with the doctrine and have their own reasons for endorsing the lib-
erty principle. What we share is an understanding of what justice re-
50. John Rawls, A Theory of Justice, revised edition (Cambridge, MA: Harvard University
Press, 1999), 509.
381
quires and a conviction about the truth of that requirement. I understand

that others endorse that requirement and assert it to be true for reasons
different from mine. A full display need not be divisive, then, but it may
involve my presenting my reasons for accepting the common ground we
are all assumed to occupy.51 Of course, the full display will not provide a
public justification, but it may nevertheless play a constructive role.
5. Getting It Right
The political conception of truth, I have argued, spares us from the horns
of a dilemma. It enables us to respect the limits of public reason and to
preserve a place for truth. The political conception thus addresses what
might be seen as a fundamental objection to the idea of public reason. I
want to conclude now with some comments that will set the discussion
in a wider context and will clarify what is at stake.
Consider the following remarks:
1. “[Fania Pascal’s] statement [‘I feel just like a dog that has been run
over’] is grounded neither in a belief that it is true nor, as a lie
must be, in a belief that it is not true. It is just this lack of connec-
tion to a concern with truth—this indifference to how things really
are—that I regard as the essence of bullshit.”52
2. “The truth of an opinion is part of its utility. If we would know

whether or not it is desirable that a proposition should be be-
lieved, is it possible to exclude the consideration of whether or not
it is true? In the opinion, not of bad men, but of the best men, no
belief which is contrary to truth can really be useful: and can you
prevent such men from urging that plea, when they are charged
with denying some doctrine which they are told is useful, but
which they believe to be false.”53
3. “[We] need to take seriously the idea that to the extent that we
lose a sense of the value of truth, we shall certainly lose something
and may well lose everything.”54
51. Here I follow Rawls’s discussion of “declaration,” in “Idea of Public Reason Revisited,”
594.
52. Harry Frankfurt, On Bullshit (Princeton: Princeton University Press, 2005), 33–34.
53. John Stuart Mill, On Liberty (Indianapolis: Hackett, 1978), chap. 2, paragraph 10.
54. Williams, Truth and Truthfulness, 7.
382
Set against the background of these remarks by Frankfurt, Mill, and

Williams, Rawls’s claim that the concept of truth has no place in public
reason may seem startling. Is political liberalism an invitation to the
bullshitter’s indifference to truth and falsity; to tying the hands of the
“best men” from pleading the truth of unconventional beliefs; or, by los-
ing the distinction between belief with its characteristic discipline and
undisciplined wishful thinking, losing everything? Certainly not. Rawls’s
point about truth and its place in public reason is less starkly at odds with
the claims of Frankfurt, Mill, and Williams than these passages might
suggest and is so in at least two ways that by now should be clear.
First, in his defense of truth and truthfulness, Williams was especially
concerned about a kind of infantilization of discourse consequent on los-
ing the distinction between belief and wishful thinking. But there may be
barriers other than truth to the encroachment of wishful thinking—per-
haps the Rawlsian standard of reasonableness suffices.
Second, and more fundamentally, the idea that democracy’s public
reason should do without the concept of truth is not the idea that “we”
can do without the concept of truth and, therefore, need not, in any
straightforward way, provoke Williams’s concern, or Frankfurt’s about a
culture of bullshitting, or Mill’s about indifference to truth. To explain, I
will call the proposal that we do without the concept of truth “the cul-
tural proposal.” The thought is that the culture, broadly speaking, should
do without the concept of truth. More particularly, suppose we accept (as
the political conception says) that truth is a norm for belief distinct from
the norm of justification and that a concern for believing the truth is a
concern for getting things right, for how things really are, and not simply
for having warranted beliefs, even ideally warranted. The proposal, then,
is that the culture would be better off—and we would be better off—if we
lacked the notion of truth and the associated distinct norm in public dis-
course. We would be better off if concerns about the correctness of belief
were correspondingly understood to be exhausted by concerns about hav-
ing beliefs that are supported by the best available reasons and that help
us to navigate our way in the world.
Richard Rorty’s view of truth is, I think, best understood along the
lines of the cultural proposal.55 Although Rorty sometimes presented a
55. See in particular Richard Rorty, “Is Truth a Goal of Inquiry? Donald Davidson versus
Crispin Wright,” in Truth and Progress: Philosophical Papers, vol. 3 (Cambridge: Cambridge
University Press, 1998), esp. pp. 41–42. In his comments on my paper at a Stanford Law School
383
reductive theory, on which truth is understood as acceptance in a com-

munity, or as warranted assertibility, or as warranted assertibility at the
idealized end of inquiry (or as a jocular slap on the back for ideas we
agree with), he agreed that none of these reductive proposals gives an ad-
equate account of truth.56 The problem with the concept of truth, on the
cultural proposal, is precisely that it is irreducibly distinct from justifica-
tion, that it makes perfectly good sense to ask, relative to any account of
what justification is, whether a proposition that meets that standard of
justification, and that we are, therefore, warranted in believing and as-
serting, is true.57
Because of this irreducible difference from justification, the adherent
of the cultural proposal urges that we drop the concept of truth from our
repertoire and aspire to a cultural world in which people regard the con-
cern for truth—for getting it right, latching onto things as they really
are—as a thankfully transcended preoccupation, like the concern about
whether ghosts are in the cellar or an incubus is upstairs. “[A]ttaining
truth as distinct from making justified statements is a goal for metaphysi-
cally active inquirers. We metaphysical quietists deplore the fact that
most people in our culture can be incited to this sort of activity. . . . [W]e
pragmatists hope our culture will eventually replace itself with the cul-
ture that James and Dewey foresaw.” Pragmatists, Rorty says, “should see
themselves involved in a long-term attempt to change the rhetoric, the
common sense, and the self-image of the community”58—all deposits for
the detritus left by decaying metaphysical doctrine. Pragmatism, thus un-
derstood, is not a thesis that truth is what works in the long-run, or any
other run, but a recommendation that we figure out what works, get
down to the business of doing it, and stop worrying about truth.
Of course, that counsel is hard to follow. Foucault says that “the ques-
tion for the West” is: “How did it come about that all of Western culture be-
gan to revolve around this obligation of truth which has so far taken a lot
of different forms? Things being as they are, nothing so far has shown that
colloquium in October 2006, Rorty disagreed that truth is a norm distinct from warrant, for rea-
sons that struck me as assuming some form of pragmatism, and resisted my attribution to him
of the cultural proposal. But once we accept that truth and warrant are distinct norms, and dis-
tinguish the word “true” from the concept of truth, then we must conclude that Rorty is en-
dorsing something along the lines of the cultural proposal. Or so it still seems to me.
56. Rorty notes the wavering in ibid., 21–22.
57. See Hilary Putnam, “Does the Disquotational Theory Solve All Problems,” in Words
and Life (Cambridge, MA: Harvard University Press, 1994), 264–278.
58. Rorty, “Is Truth a Goal,” 29, 41.
384
it is possible to define a strategy outside of this concern [emphasis added].

It is within the field of the obligation to truth that it is possible to move
one way or another, sometimes against the effects of domination which
may be linked to structures of truth or institutions entrusted with truth.”59
These hardships notwithstanding, the pragmatist urges that we soldier on.
According to the cultural proposal, concerns with truth foster anxiety
about whether we have things really right, or skepticism because we
might not, or despair because we do not or cannot. Those concerns re-
flect a failure to take a proper sense of responsibility for our convictions
as our own. We would live better, freer, happier, less fraught lives if we
gave up worrying about getting things really right, stopped fretting that
the world might be other than what we justifiably take it to be, focused
on solving problems, and—with cheerful irony—embraced our convic-
tions and our culture wholeheartedly and without embarrassment as . . .
ours. And we should unburden ourselves not because we finally have
found the philosopher’s stone—a way to answer the skeptic, or a grip on
the idea of correspondence with reality, or an understanding of precisely
what substantial property distinct from warrant truth really is—and are
entitled to some rest after an intellectual job well done. The critique of
truth is a cultural intervention, not a philosophical argument—a recom-
mendation for living, not a theory. We should drop the concept of truth
because it does not help us get on with life’s business.
Williams and Frankfurt are, I think, worried about the cultural pro-
posal. But the cultural proposal is different in three important ways from
the idea that democracy’s public reason lacks the concept of truth. First,
public reason is about political justification, not about the culture gener-
ally. Second, in presenting a political conception as available on com-
mon ground, the idea is precisely not to take a view on the correct theory
of truth or the proper attitude to it in the culture but to assume a range of
views and to avoid unnecessary controversy: the political proposal begins
from the assumption, suggested by the idea of doctrinal conflict, that “the
public culture” is unalterably divided in its understanding of, inter alia,
the nature of truth. Third, the point of finding a conception that can
serve as common ground is to provide a basis for cooperation among
equals on a basis of mutual respect; the cultural proposal, in contrast, ap-
pears to express a specifically romantic ideal of self-creation and rejects
the concept of truth because it burdens such creation.
59. Michel Foucault, “The Ethics of Concern for Self as a Practice of Freedom,” in Ethics:
Subjectivity and Truth, ed. Paul Rabinow (New York: New Press, 1997), 295.
385
Suppose, then, we put aside the cultural proposal about dropping truth
from the culture, not because the cultural proposal is wrong, but because
it is not our topic. The disagreement at issue here is not between Rorty
and Williams (or Frankfurt or Mill), but between Rawls and Pinter. As-
sume, then, a culture in which the concept of truth is available: some
people think we have already lost this culture; others regret that they are
wrong. And suppose that we are tempted to reject the Rawlsian view
about keeping truth out of public reason: we agree that it is hard to un-
derstand what it means to drop the concept (unless we are also prepared
to drop, among others, the concepts of belief, assertion, judgment), that a
nonmetaphysical (not antimetaphysical) understanding of truth is avail-
able, and that the reasons for dropping it are not compelling. So we are
prepared to keep the concept of truth (interpreted along the political
lines I described earlier) in political justification and to acknowledge
that truth is different from (even ideal) warrant. In rejecting the Rawlsian
proposal, aren’t we still recommending the public anxiety that the cul-
tural proposal condemns? The objection to keeping the concept of truth
raised by the cultural proposal may still have force even when our focus
is narrowed to the arena of public political discourse.
It may. But should the alleged anxiety be laid at the doorstep of truth?
Why not say instead that it comes with the territory of justice, thus with
the territory of public reason?
We are concerned to do what justice requires: anyway, that is what we
say; that is what we want other people to think we are committed to; it is
what we want to take ourselves to be committed to. Not simply what we
think justice requires, or what we warrantedly believe it to require, but
what justice requires. But caring about justice, as the political concep-
tion indicates, requires caring about the truth about justice. If a concern
about justice has—as I said at the outset—a place in democracy’s public
reason, then so, too, does a concern about getting it right, that is, about
the truth about justice. We of course should, while keeping a concern for
the truth, steer clear of needless controversy about the nature of truth: the
political conception of truth suffices to meet that aim; nothing either so
extreme or so fugitive as leaving the concept of truth out is needed.
As for the anxiety that comes with the concern to get justice right: that
comes with the territory of taking justice seriously. We can live with it,
should not live without it, and should not enlist philosophy to provide
therapy for that anxiety.
386
INDEX
Abers, Rebecca, 207n28, 337 Austin, John, 366n37

abortion, and privacy rights, 18–19, 313–316, Austin v. Michigan Chamber of Commerce, 290,
314–315n13, 317n16 291
Acker, Kathy, 116, 118, 252–253 autonomy: deliberative democracy and, 17, 25,
Ackerman, Bruce, 293n44 27–28; directly deliberative polyarchy and,
aggregative conception of democracy: delibera- 215; freedom of expression and, 120; liberty
tive democracy and, 157–160, 163–164, 168; di- and, 257, 257n54, 259
rectly deliberative polyarchy and, 191–192; lib-
erty and, 223–224, 224n3, 264–266; privacy Baccaro, Lucio, 338
rights and, 307–308, 308n6 “background culture,” 271, 303, 318
Amar, Akhil Reed, 148n131 Baiocchi, Gianpaolo, 337
Anarchy, State, and Utopia (Nozick), 5 Baker, C. Edwin, 127n81
Ansolabehere, Stephen, 282n22, 296, 337–338 Barnes v. Glen Theatre, 254, 255
Aquinas, Thomas, 376 Beitz, Charles, 138n110, 171nn32–33, 270n4
Arendt, Hannah, 204, 220, 220n42, 221, 221n45, Bellotti decision, 289
374, 374n44 Benhabib, Seyla, 220n42
“Associations in Democratic Governance” (Co- Black, Hugo L., 127n80, 129
hen and Rogers), 13, 36n40, 61 Blackmun, Harry, 262, 317
associative democracy: alternative governance Blasi, Vincent, 283
and, 69–70; comparative experience lessons Bollinger, Lee C., 114n31, 122n70, 124n76
and, 70–74, 72n16; directly deliberative Bowers v. Hardwick, 254, 255, 262, 317
polyarchy and, 200–201; education of citizens Bowman v. The United Kingdom, 301–302
and, 69; egalitarianism and, 63–66, 64nn5–7, Brady, Henry E., 280
67; equalizing representation and, 68–69; Brandeis, Louis D., 102, 118n60, 123, 134, 136,
groups, and potential contributions to, 67– 139, 140, 250, 301
68; ideal, 16, 21–22, 37; impossibility argu- Brandenburg v. Ohio, 126
ment and, 75–76; participation/deliberation Brazil, and participatory democracy, 207n28,
tensions and resolution using, 175–180, 337, 338, 345
177nn46–47, 179nn49–50; response to prob- Breyer, Stephen G., 270, 291, 302n56
lems with, 74–75; summary of, 5, 10, 12–13, British law, and electoral finance, 301–302
61–63, 62nn3–4, 96–97; undesirability argu- Bruce, Lenny, 116
ment and, 76–83, 77n21, 82n25; unions and, Buckley v. Valeo: democracy and, 11–12, 302;
80–81; vague delegations of power, and risks electoral finance issues and, 173–174, 174n38,
of abused discretion in, 80, 82–83, 82n25 268–269, 285–286, 288, 290–291, 293,
387
I N D E X
Buckley v. Valeo (continued) constitutional issues: directly deliberative

293nn43–44, 297–299; equal opportunities polyarchy, and effects of, 201, 212–216; elec-
and, 275; fair access and, 138, 138n111 toral finance and, 268–269, 269n1, 270, 285–
292, 292–293, 293nn43–44, 301–302; judicial
Cain, Bruce, 277n14 review and, 212–216, 242, 301–302. See also
campaign finance. See electoral finance specific cases and decisions
Capitalism, Socialism, and Democracy costs of reticence, and privacy conventions,
(Schumpeter), 12, 31, 301 319–321
Carnap, Rudolf, 356n17 cultural democracy, 305, 319, 320, 322, 324–325
Carter, Stephen, 247n36 cultural proposal, and truth, 352, 383–384nn55–
Casey decision, 314–315, 317n18 56, 383–386
Catholic doctrines: Dignitatis Humanae, 363;
Evangelium Vitae, 235, 315; on homosexual- Dahl, Robert, 272n8; on citizen control over po-
ity, 256; on human dignity, 363; on religious litical agenda, 272n8; on common good, 168;
freedom, 229n12, 352; on toleration, 229n12; on consensus, 48n21; on courts and judicial
Veritatis Splendor, 363 review, 242; on democracy, 158n10, 189n12,
Chaplinksy v. New Hampshire, 121n67 224n3; on directly deliberative polyarchy, 187;
Chattopadhyay, R., 338 on equal opportunities, 224n3, 228n11, 241–
child pornography, 104, 105, 126, 143 242; on liberty, 241–242, 242n27, 243n29; on
citizens: education of, 69, 109, 330; elections, polyarchy, 187
and role of, 299, 299n51; participatory/delib- Davidson, Donald, 359
eration tensions, and role of, 343–346; politi- Davis v. Bandemer, 170n29, 271n6
cal agenda, and control by, 272n8 deflationary theories of truth, 350, 350n5, 352,
civic consciousness, 19–20, 19n13, 19n15 369–370
civility, duty of, 305, 321–322 de Grazia, Edward, 118n59
Cohen, G. A., 40n6 deliberation: directly deliberative polyarchy
collective authorization: deliberative democ- and, 201–207, 202n24; liberty and, 238; partici-
racy and, 154, 155, 305, 308, 308n7, 311, 335– pation tensions with, 175–180, 177nn46–47,
336; democracy and, 154–155, 154n1, 156, 223; 179nn49–50, 343–346; reflections on, 329–
directly deliberative polyarchy and, 191–192, 334, 330nn7–8, 333n11, 336–341
191n17; privacy rights and, 305–309, 308n7; deliberative democracy: aggregative democracy
Rawls on, 308n7 and, 157–160, 163–164, 168; associative de-
collective decision-making. See collective au- mocracy, and resolution of tensions between
thorization participatory democracy and, 175–180,
Colorado Republican Federal Campaign Com- 177nn46–47, 179nn49–50; autonomy and, 17,
mittee v. FEC, 292 25, 27–28; civic consciousness and, 19–20,
commercial speech, 127n81 19n13, 19n15; collective authorization and,
common good, 25–28, 168–170, 169n26 154, 155, 305, 308, 308n7, 311, 335–336; com-
comprehensive doctrines: democracy and, 3; mon good and, 25–28; decision-making and,
Rawls on, 8, 225–226, 300, 306; reasonable 6–7, 9, 11, 12, 16, 21, 28, 29; deliberation, and
pluralism and, 225–226, 231–232; truth and, reflections on, 329–334, 330nn7–8, 333n11,
353, 355, 357 336–341; deliberative pathologies, and skepti-
consensus: Dahl on, 48n21; moral pluralism cism about, 338–339; described, 16, 17, 17n3,
and, 38, 38n3, 39, 41, 50–51, 55–56; overlap- 20–21, 180; effectiveness, and skepticism
ping, 39, 41, 50–51; Rawls on, 39, 40, 41–42, about, 337–338; egalitarianism and, 18–19,
59 18n9, 18n11, 306, 309, 336–337; freedom of ex-
388
I N D E X
pression and, 32–34, 32nn33–34, 34nn35–36; democracy and, 200–201; collective authori-
ideal, 23–25, 23nn22–23, 31–32, 160–164; insti- zation and, 191–192, 191n17; constitutional
tutional structures issues and, 17, 25, 28, 29, courts and, 212–216; constitutional issues and,
31, 34–37, 34n38, 36n40, 346–347; introduc- 201, 213–214, 215; deliberation and, 201–207,
tion to, 5, 8–11, 14, 16–17, 326–329, 328n4; 202n24; described, 181, 187–188, 189, 196, 199,
irrelevance objection to, 34–35, 34–35nn38– 221–222; egalitarianism, and effects of, 216;
39; justice and, 334; majority rule and, 25, 31, introduction to, 5, 12–13, 181–186, 183nn4–5;
68, 162, 191–192, 204, 237, 256, 259–260, legislatures and, 211–212; membership issues
263n70, 331–332; naiveté about power, and and, 209–210; political institutions and, 200–
skepticism about, 339–341; participatory de- 201, 211–216; problem-solving effectiveness
mocracy, and tensions with, 175–180, and, 210–211; public goods and, 13, 18n5, 208;
177nn46–47, 179nn49–50, 341–343, 342n28; radical-democratic concepts of the public
public goods and, 17–18, 35, 337–338, 337n20; and, 2, 216–221, 217–218nn34–35, 219n37,
public sphere large-scale decisions, and chal- 219n39, 220n42, 221n45; state and market is-
lenges to, 347; Rawls on, 16, 17–21; reasonable sues and, 207–209, 208nn29–30; virtues of de-
pluralism and, 305, 306–307, 309; reason- mocracy as political ideal and, 188–191, 188–
giving argument for, 334–335; sectarianism 189nn10–12, 191n17
and, 29–30, 45, 53–54, 112–113, 120–121, 373– disclosure of personal matters, and privacy
376 rights, 312–313, 312n10
deliberative inclusion, 164–166, 166nn17–18, discrimination: collective authorization, and
225, 241–243, 243nn29–30 protection against, 159, 159nn11–12, 164; con-
deliberative polyarchy. See directly deliberative tent-discriminatory regulations and, 125–126,
polyarchy 142, 144–145, 144n125, 147n129, 150–151; hate
democracy: aggregative concept of, 191–192, speech and, 99, 100, 152; sex, 144n125
223–224, 224n3, 264–265; authorization and, doctrinal pluralism, and truth, 353
157–160, 157n5, 158–159nn8–12, 223, 223n1; Doe v. Michigan, 100, 100n8
Buckley and, 11–12, 302; collective authoriza- “Domain of the Political, The” (Rawls), 41
tion and, 154–155, 154n1, 156, 223; conception Dryzek, John, 333–334
of, 5–8, 6n4; Dahl on, 158n10, 189n12, 224n3; Duflo, Esther, 338
deliberative-direct, 193–194, 193n20; egalitari- Dummett, Michael, 370n39
anism and, 189, 189n12, 189n14; electoral duty of civility, and privacy conventions, 305,
finance system and, 292–302, 293n43, 302n56; 321–322
institutional structures issues and, 194–199, Dworkin, Ronald: on abortion, 315; on citizen’s
196n22, 198n24; majority rule and, 191, 191n17; role in elections, 299n51; on consumption of
philosophy, and role in ideal, 15, 386; privacy pornography, 109; on education of citizens,
rights and, 305; procedure of, 156–158, 161– 109, 330; on egalitarianism, 274n10; on free-
163, 170; public reason and, 309–310; reli- dom of speech, 109n32; on liberal tolerance,
gious freedom and, 154–155; theory of, 61; vir- 255; Life’s Dominion, 315; on objection to de-
tues of, 188–189nn10–12, 189nn14–15, 188–191, liberative democracy, 32n34; on reflective
191n17 endorsement of morality, 257–258, 257n58
“Democracy and Morality” (Devlin), 256
Devlin, Sir Patrick, 255–256, 257, 260, 321 education of citizens, 69, 109, 330
difference principle, 169–170 egalitarianism: associative democracy and, 63–
Dignitatis Humanae, 363 66, 64nn5–7, 67; deliberative democracy
directly deliberative polyarchy: aggregative con- and, 18–19, 18n9, 18n11, 306, 309, 336–337;
cept of democracy and, 191–192; associative democracy and, 189, 189n12, 189n14; directly
389
I N D E X
egalitarianism (continued) fair access, 20–21, 20n16, 106–108, 135–140,

deliberative polyarchy, and effects on, 216; 137nn106–107, 138n109, 138n111
Dworkin on, 274n10; freedom of expression Falwell, Jerry, 106, 119n62, 128n84, 135,
and, 102–103; moral pluralism and, 41, 47–49; 135n103
participatory democracy and, 336–337, Federal Election Campaign Act (FECA), 285,
337n21; privacy rights and, 306, 309. See also 288, 289
equal opportunities finance, electoral. See electoral finance
electoral finance: British law and, 301–302; First National Bank of Boston v. Bellotti. See
Buckley and, 173–174, 174n38, 268–269, 285– Bellotti decision
286, 288, 290–291, 293, 293nn43–44, 297–299; Fish, Stanley, 128–129
citizen’s role in elections and, 299, 299n51; Fishkin, James, 344
conflicts in realizing ideal, 173–174, 173nn34– Flint, Larry, 106, 119n62, 128n84, 135
35; constitutional limits on reforms for, 268– Foucault, Michel, 384
269, 269n1, 270, 285–293, 293nn43–44, 301– Frankfurt, Harry, 382–383, 385, 386
302; current system of, 268, 279–283, 279n16, freedom of expression: artistic expression and,
280n18, 280n20, 292–302, 293n43, 302n56; 250; autonomy and, 120; Brandeis’s “more
equal opportunities and, 268, 270–279, 271– speech” and, 102, 118n60, 123, 134, 139, 140,
272nn6–8, 277nn14–15, 284–285, 300–302; 250, 301; child pornography and, 104, 105,
FECA and, 285, 288, 289; First Amendment 126, 143; commercial speech and, 127n81;
protection and, 286, 289, 291, 293, 297; floors/ content regulation and, 125–126; deliberative
ceilings on, 294–295; freedom of expression democracy and, 32–34, 32nn33–34, 34nn35–
and, 295–300; informal public sphere and, 36, 119, 119nn61–62, 120, 120n63; Dworkin on,
277, 287; summary of, 4–5, 10, 11–12, 268–270, 109n32; egalitarianism and, 102–103; electoral
302; voluntary public financing and, 175, finance and, 295–300; expressive interest and,
269–270, 270n3, 282–283n22, 288–289, 291– 114–118, 114nn56–57, 120, 120n63; fair access
292, 293n44, 294 and, 106–108, 135–140, 137nn106–107,
Elster, Jon, 21n20, 27, 333 138n109, 138n111; hate speech and, 98–100,
Ely, John Hart, 32n33, 104n18, 157n6, 214n33, 99nn6–7, 104–105, 140–153, 142nn120–121,
242n28, 243n30, 256n53 144n125, 146–148nn128–131, 153n137; historical
“Epistemic Conception of Democracy, An” background for, 4, 122–124, 122nn69–70,
(Cohen), 7 124nn74–76, 141–144; informational interest
equal opportunities: Dahl on, 224n3, 228n11, and, 119, 120, 120n63; interests and, 101–102,
241–242; directly deliberative polyarchy and, 113–120, 114n43, 115nn45–47, 120n63; intro-
212; electoral finance and, 268, 270–279, 271– duction to, 10–11, 12–13, 98–103; nihilism and,
272nn6–8, 274–275nn10–11, 277nn14–15, 284– 101, 101n10, 128–131; objections to deliberative
285, 300–302; liberty and, 235. See also egali- democracy and, 32–34, 32nn33–34, 34nn35–
tarianism 36; offensive expression and, 100, 118, 121, 131–
equal representation, 68–69, 86–90, 86n31 132, 134n102, 135, 135n103; political speech
Estlund, David, 158n9, 364, 365 and, 114nn59–60, 117–118, 250–251; pornogra-
European Court of Human Rights, 301–302 phy and, 10, 103, 107, 109, 111n40, 132–134,
Evangelium Vitae, 235, 315 133nn98–99, 134n102; stringent protections
exclusion of views, and moral pluralism, 56–58, for, 103–108, 104n18; Sunstein on, 102n15, 117,
58n30 117n53, 118n60, 248, 249
Freeman, Samuel, 356n17
facts of pluralism, 43–44, 44n12, 52, 52n25, 55– Frege, Gottlob, 350, 359n26, 370n39
56, 59–60 Fung, Archon, 327n2, 336
390
I N D E X
Gerber, Alan, 337–338 198n24; directly deliberative polyarchy and,

Germany, and groups’ potential contributions, 200–201, 211–216
71, 73–74, 81n24, 344 Isaac, Jeffrey, 221n45
Gibbard, Allan, 361–362nn29–30 Italian unions, and participatory democracy, 338
Gramsci, Antonio, 3, 3n3 Iyengar, Shanto, 296
groups (associations), potential contributions of,
67–68, 70–74, 72n16, 80–81, 81n24, 344 John Paul II, 235
Guinier, Lani, 175n44, 271n6 Johnston, Mark, 52n26, 114n44
Gutmann, Amy, 202n25, 333 justice, 334, 356n16, 374–375n45. See also in-
justice
Habermas, Jürgen, 7, 21n20, 24, 26n26; on delib- justification: Rawls on, 47; religious freedom,
erative democracy, 328, 328n4, 330; on ideal and deliberative, 11; truth conception, and
speech, 340; on informal public sphere, 271, political, 14, 349–350, 349nn2–3, 350n5, 351,
318, 343; on political concept of truth, 364– 368–373, 370n39
365n34; the public, and radical-democratic
views of, 217–220, 217–218nn34–35, 219n37, Kant, Immanuel, 49, 49n20, 228n11, 230, 235,
219n39; on Rawls’s full display argument, 349n3, 361
380n49 Kennedy, Anthony McLeod, 291
harm principle: freedom of expression and, Kitcher, Philip, 374–375n45
119n61, 122, 126, 127, 140; Mill on, 255, 257;
pornography and, 140, 252 labor unions, 80–81, 338
Hart, H. L. A., 255, 260 Lawrence v. Texas, 317, 317n16, 317n18
hate speech, 98–100, 99nn6–7, 104–105, liberalism, 307, 307n5, 352, 357n21; truth and,
140–153, 144n125, 146–148nn128–131, 352, 357, 357n21
153n137 Liberalism Against Populism (Riker), 7
Hayek, Friedrich, 197–198 liberty: aggregative concept of democracy and,
Hegel, G. W. F., 45–46, 45–46nn13–15, 59, 223–224, 224n3, 264–266; autonomy and, 257,
59n31 257n54, 259; Dahl on, 241–242, 242n27,
Heidegger, Martin, 360 243n29; deliberation and, 238; deliberative in-
Hobbes, Thomas, 366–367 clusion and, 225, 241–243, 243nn29–30; dis-
Holmes, O. W., 118n60, 121, 128 cussion and, 237–240, 238n20; freedom of ex-
homosexuality, 254–255, 256, 260–263, 261n64, pression and, 102–103, 248–254; introduction
261n66, 262n69, 320 to, 11, 223–225, 223n1, 224n3; moral, 124n75,
Horwich, Paul, 372n42, 379n48 254–264, 257n54, 257nn56–58, 261n64,
Hustler v. Falwell, 106, 119n62, 128n84, 135 261n66; motivation and, 239, 239nn22–23,
240–241; political legitimacy and, 266–267;
independence of judgment in decisions, and public reason and, 232–237, 235–236nn16–17;
privacy rights, 312–313, 312n10 Rawls on, 115n47, 117n53, 137n106; reasonable
India, and participatory democracy, 338, 345 pluralism and, 225–232, 227n8, 228–229nn10–
inequality. See egalitarianism 12, 231n14; reflectiveness and, 249, 251, 257–
informal public sphere, 12, 271, 277, 287, 305, 259, 257nn56–58; religious freedom and, 234–
318, 343 235, 235n16, 243–248, 243nn29–30, 245n32,
injustice, 32–34 247nn35–36; sexual intimacy and, 254–255,
institutional structures issues: deliberative de- 256, 260–262, 261n64, 261n66, 262n69,320
mocracy and, 17, 25, 28–37, 34n38, 36n40, life, death, and sex issues: abortion and, 18–19,
346–347; democracy and, 194–199, 196n22, 313–316, 314–315n13, 317n16; privacy rights
391
I N D E X
life, death, and sex issues (continued) natural law theory, 49n20, 166n18, 229n12, 357,
and, 305, 311–312; right to die and, 313–316, 363, 364, 366n37
314–315n13; sexual intimacy and, 313–316, nihilism, and freedom of expression, 101,
314–315n13, 316–317, 317n16, 317nn18–19 101n10, 128–131
Life’s Dominion (Dworkin), 315 Nixon v. Shrink, 270, 290, 298, 302n56
Lincoln, Abraham, 2, 15, 155n3 Noble, William, 118n59
Locke, John, 7, 49n20, 188n10, 258, 259, 357 Nozick, Robert, 5, 6, 239n22
Lowi, Theodore, 197
O’Connor, Sandra Day, 291
MacKinnon, Catharine, 107, 139–140, 251, On Democracy (Cohen and Rogers), 6, 10,
253 36n40
Madison, James, 32n34, 35n39, 77, 77n21, 123
majority rule, 25, 31, 68, 191, 191n17 Panglossian outlook, 15
Mansbridge, Jane, 336 participatory democracy: associative democracy,
market issues, state and, 207–209. See also pub- and resolution of tensions between delibera-
lic goods tive democracy and, 175–180, 177nn46–47,
Marx, Karl, 46 179nn49–50; deliberative democracy, and
Mathews, John, 62n4 tensions with, 175–180, 177nn46–47,
McClosky, Herbert, 228n10 179nn49–50, 341–343, 342n28; described, 328–
McConnell, Michael, 231n14 329, 341; egalitarianism and, 336–337, 337n21
Meiklejohn, Alexander, 125, 248–249 philosophy, and role in ideal democracy, 1–5,
Metro Broadcasting, Inc. v. FCC, 137n107 3n3, 6, 6n4, 15, 386
Mill, J. S.: on censorship, 119, 119n61, 137, 289– Pinter, Harold, 348, 349, 349n1
290; on doctrinal pluralism, 353; harm princi- Planned Parenthood of Southeastern Pennsylva-
ple of, 255, 257; on moral liberty, 124n75, 254, nia et al. v. Casey. See Casey decision
259 Plato, 330, 358
Miller v. California, 148 police enforcement, 109, 184, 184n5, 263,
Mora, Alberto, 2 263n70, 336, 345
moral liberty, 124n75, 254–264, 257n54, political legitimacy, and liberty, 266–267
257nn56–58, 261n64, 261n66, 310–311 Political Liberalism (Rawls), 348, 349, 376n46
moral pluralism: common ground for, 49–50, political parties, 17n3, 18, 24, 36, 36n40, 170n30,
49nn20–21; consensus and, 38, 38n3, 39, 41, 174–175, 175n42, 192n19, 218, 279–281, 286,
50–51, 55–56; egalitarianism and, 41, 47–49; 291, 293, 327–328, 343
exclusion of views and, 56–58, 58n30; as fun- political speech, and freedom of expression,
damental good, 43–44, 44n11; historical back- 114nn59–60, 117–118, 250–251
ground for, 45–46nn13–15, 45–47, 59, 59n31; polyarchy. See directly deliberative polyarchy
overview of objections to, 41–42; Rawls on, 8– pornography issues, 10, 103, 107, 109, 111n40,
9, 9n10, 38, 38n2, 41, 43–44, 44n12, 52, 52n25, 132–134, 133nn98–99, 134n102, 140–141, 148–
55–56, 59–60; reasonable pluralism and, 52– 151, 149–150nn134–135
55, 52nn24–25, 54n28, 155, 228–229nn10–12; Porto Alegre, Brazil, 207n28, 337, 338, 345
summary of, 8–9, 9n10, 38–41, 38nn2–3, 58– Posner, Richard, 337n20
60, 59n31 Post, Robert, 119n62
Mutz, Diana, 342n28 Prison Notebooks (Gramsci), 3
privacy conventions: costs of reticence and,
Nagel, Thomas, 54n28, 263n70, 304, 317–318, 319–321; duty of civility and, 305, 321–322;
320, 325 freedom of expression restraint and, 321, 323–
national identity, 155n3 325; introduction to, 305; pressure to conform
392
I N D E X
and, 321–323; public and private boundaries ference principle, 169–170; “The Domain of
and, 317–319 the Political,” 41; on justice, 356n16, 374–
privacy rights: abortion and, 18–19, 313–316, 375n45; on justification, 47; on liberalism,
314–315n13, 317n16; aggregative concept of 307, 352, 357n21; on liberty, 115n47, 117n53,
democracy and, 307–308, 308n6; cultural de- 137n106; on moral pluralism, 8–9, 9n10, 38,
mocracy system, 305, 319, 320, 322, 324–325; 38n2, 41, 43–44, 44n12, 52, 52n25, 55–56, 59–
deliberative democracy and, 11; democracy 60; on philosophy of democracy, 5, 6, 6n4;
and, 305; disclosure of personal matters and, on pluralism, 8–9, 9n10, 38, 38n2, 41; Politi-
312–313, 312n10; egalitarianism and, 306, 309; cal Liberalism, 348, 349, 376n46; on private
freedom of expression and, 310–311; indepen- sphere, 303; on public reason, 351; A Theory
dence in decisions and, 312–313, 312n10; in- of Justice, 5, 38, 46–48, 59, 61; on truth, 348,
formal public sphere and, 12, 305; liberalism 349, 351–358, 354n10, 354n12, 356nn16–17,
and, 307, 307n5; life, death, and sex issues 357n21, 364–365n34, 373–382, 374–375n45,
and, 305, 311–312; moral liberty and, 310–311; 380n49
reasonable pluralism and, 11, 305, 306–307, Raz, Joseph, 54n28, 120n63, 257n54, 354n10, 378
309; right to die and, 313–316, 314–315n13; sex- reasonable pluralism: comprehensive doctrine
ual intimacy and, 313–315, 314–315n13, 316– concept and, 225–226, 231–232; deliberative
317, 317n16, 317nn18–19; summary of, 12, 303– democracy and, 305, 306–307, 309; democ-
305 racy and, 155–156, 180; liberty and, 225–232,
proscribable expression, 143 227n8, 228–229nn10–12, 231n14; moral plural-
Przeworski, Adam, 196n22, 237 ism and, 52–55, 52nn24–25, 54n28, 155, 228–
public and private, 317–319 229nn10–12; privacy rights and, 11, 305, 306–
public goods, 337, 337n20, 338; deliberative de- 307, 309; religious freedom and, 8, 38n2,
mocracy and, 17–18, 35, 337–338, 337n20; di- 115n47, 156
rectly deliberative polyarchy and, 13, 208; pri- reflectiveness, and liberty, 249, 251, 257–259,
vacy rights and, 337, 337n20, 338 257nn56–58
public reason: democracy and, 309–310; liberty Rehnquist, William H., 254, 291
and, 232–237, 235–236nn16–17; Rawls on, 351; religious freedom: Catholic doctrine on,
religious freedom and, 234–235, 235n16; truth 229n12, 352; deliberative inclusion and, 164–
and, 14, 350–351, 352–358, 354n10, 354n12, 166, 166nn17–18; deliberative justification
356nn16–17, 357n21 and, 11; democracy and, 154–155; historical
public sphere, 12, 271, 277, 287, 305, 318, 343, context for, 10, 157, 159; liberty and, 234–235,
347 235n16, 243–248, 243nn29–30, 245n32,
247nn35–36; privacy rights and, 310–311; pub-
radical-democratic concepts, 2, 216–221, 217– lic reason and, 234–235, 235n16; reasonable
218nn34–35, 219n37, 219n39, 220n42, 221n45, pluralism and, 8, 38n2, 115n47, 156
326–327. See also deliberative democracy; di- right to die, 313–316, 314–315n13
rectly deliberative polyarchy; participatory Riker, William, 7, 31, 64
democracy Roe v. Wade, 314
rationalist fundamentalists, and denial of free- Roe v. Whalen, 312
dom of expression, 131, 131nn92–94 Rogers, Joel, 6, 10, 12–13, 36n40, 61, 327n2
R. A. V. v. St. Paul, 103, 141–142, 147n129 Rorty, Richard, 383–384, 383–384nn55–56
Rawls, John: on “background culture,” 271, 303, Rosati, Connie, 111n37
318; on “burdens of reason,” 52n25; on com- Rousseau, Jean-Jacques, 7, 34n38, 327, 327n3
prehensive doctrine concept, 8, 225–226, 300,
306; on consensus, 39, 40, 41–42, 59; on de- Sabel, Charles, 13, 181, 327n2
liberative democracy, 16, 17–21, 331n8; on dif- Sandel, Michael, 183n50, 244, 268
393
I N D E X
Scalia, Antonin, 103, 141–142, 145n126, 148n130, 384nn55–56; deflationary theories and, 350,
255, 290, 291, 314–315n13 350n5, 352, 369–370; doctrinal pluralism and,
Scanlon, T. M., 110n34, 111n38, 122n69, 162, 353; importance of political conception of,
162n16, 245n32, 322–323 382–386; justification, and conception of, 14,
Schlozman, Kay Lehman, 280 349–350, 349nn2–3, 350n5, 351, 368–373,
Schmitt, Carl, 170n31, 329, 333 370n39; minimalism, and concept of, 370–
Schumpeter, Joseph, 12, 31, 269, 269n1, 299, 301 371, 370n39, 372; political conception of, 358,
Scruton, Roger, 261n64 382–386; political liberalism and, 352, 357,
sectarianism, and deliberative democracy, 29– 357n21; public reason and, 14, 350–351, 352–
30, 45, 53–54, 112–113, 120–121, 373–376 358, 354n10, 354n12, 356nn16–17, 357n21;
sex discrimination, 144n125 Rawls on, 348, 349, 351–358, 354n10, 354n12,
sexual intimacy, 313–315, 314–315n13, 316–317, 356nn16–17, 357n21, 364–365n34, 373–382,
317n16, 317nn18–19; liberty and, 254–255, 256, 374–375n45, 380n49; value, and concept of,
260–263, 261n64, 261n66, 262n69, 320 372
slavery, 1, 27, 49, 99, 230, 362 Tu Wei-ming, 229n12
Snyder, James M., 337–338
sodomy, consensual, 254–255, 261 Unger, Roberto, 166n19, 243n29, 248
Souter, David, 270, 291 Universal Declaration of Human Rights, 2
state and market, 207–209, 208nn29–30 University of Michigan hate speech case, 98–
Stevens, John Paul, 312 99, 100
Strauss, David, 111n38, 111n40
Sullivan, Kathleen, 277n14, 297–298 Verba, Sidney, 280
Sunstein, Cass: on congressional control, 82n25; Veritatis Splendor, 363
on constitutionalism, 197; on deliberative de- Voice and Equality (Verba, Schlozman, and
mocracy, 16n1; on freedom of expression, Brady), 280
102n15, 117, 117n53, 118n60, 248, 249; on moral voluntary public financing, and electoral
pluralism, 18n5; on pornographic expression finance, 175, 269–270, 270n3, 282–283n22,
defined, 134n102 288–289, 291–292, 293n44, 294
voucher scheme, and electoral finance, 175,
Theory of Justice, A (Rawls), 5, 38, 46–48, 59, 61 293, 293n44
Thomas, Clarence, 291
Thompson, Dennis, 333 Wenar, Leif, 227n8
Title VII, and freedom of expression, 144n125, Whitney v. California, 118n60, 136
145nn126–127 Williams, Bernard, 14, 382–383, 385, 386
Tocqueville, Alexis de, 69, 329 Wright, Crispin, 371n40
truth: comprehensive doctrine concept and, Wright, Erik Olin, Politics and Society, 13
353, 355, 357; concept of, 359–368, 359n26;
correspondence, and concept of, 372; the cul- Zaller, John, 228n10
tural proposal and, 352, 383–386, 383–
394

Dokumen - Pub - Philosophy Politics Democracy Selected Essays 9780674271579

Uploaded by

Copyright:

Available Formats

Dokumen - Pub - Philosophy Politics Democracy Selected Essays 9780674271579

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dokumen - Pub - Philosophy Politics Democracy Selected Essays 9780674271579

Uploaded by

Copyright:

Available Formats

P H I L O S O P H Y, P O L I T I C S , D E M O C R A C Y

all rights reserved

Printed in the United States of America

Library of Congress Cataloging-in-Publication Data

Cohen, Joshua, 1951–

2 Moral Pluralism and Political Consensus 38

6 Directly Deliberative Polyarchy 181

7 Democracy and Liberty 223

8 Money, Politics, Political Equality 268

11 Truth and Public Reason 348

T he essays collected here were written over a period of twenty-two

Christiano and John Christman, eds., Contemporary Debates in Po-

T he essays collected here are about an ideal of democracy. It is the

is also a matter of expressing convictions that belong to political morality

sexual orientation; or whether the rights that many of us suppose people

lar on efforts to secure public ﬁnancing, which led to considerations

senses of that term, including a more democratic economy: to extend

were insufﬁcient and perhaps misguided. Third, I was troubled by the

philosophical.8 A conception of justice suited to such societies needed,

(1976), and preserved more or less consistently thereafter, reﬂected, I ar-

for representing interests. But we were also concerned, perhaps more

I n this essay I explore the ideal of a “deliberative democracy.”1 By a de-

interactions, both in its explicit attention to considerations of the com-

scheme, in which democratic politics consists of fair bargaining among

is manifest and to declare a commitment to that equality “as the public

12. Rawls, Theory, 233.

tion from the original position to the constitution . . . [thus] preserv[ing]

D1 A deliberative democracy is an ongoing and independent associa-

A theory of deliberative democracy aims to give substance to this for-

I1 Ideal deliberation is free in that it satisﬁes two conditions. First, the

prior norms or requirements. Second, the participants suppose

24. Habermas, Legitimation Crisis, 108.

mitted to acting on the results of a free and reasoned assessment of

Drawing on this characterization of ideal deliberation, can we say any-

common good a nd a u t onomy

even ideal deliberation would not consist in efforts to disguise personal or

cannot appeal to this desire itself in defending policies. The motivational

28. Whitney v. California, 274 U.S. 357 (1927).

natural objections to the conception I have been discussing, objections to

secta ria nism

rangements are not “exogenous constraints” on the aggregation of prefer-

irrel eva nce

questions to deliberative resolution, even if a direct democracy were a

liberation required to institutionalize a deliberative procedure. Since

T he idea of normative consensus plays a central role in John Rawls’s

can be restated under more realistic assumptions, Rawls introduced the

The pluralistic consensus test asks us to evaluate a conception

by the diverse moralities in a well-ordered society. But why

Responding to this objection, Rawls argues that consensus on justice is

generic statement of the problem of moral consensus and pluralism.

The Place of Consensus

in other respects attractive is also realistic—in particular, that it is stable.

Consensus and Moral Pluralism

mony, supports social peace, reduces the complexity of decision making,

ated with an acceptance of substantially inegalitarian forms of civic diver-

the importance of accommodating moral diversity lead to a thinner con-

The Case of Equality

The egalitarian content of A Theory of Justice is encapsulated in three

19. See Theory of Justice, 65.