Copyright (c) 1994 Chicago-Kent College of Law
Chicago-Kent Law Review
69 Chi.-Kent. L. Rev. 589
LENGTH: 15855 words
PLURALISM AND PROCEDURALISM
Joshua Cohen *
* Professor of Philosophy and Political Science, Massachusetts Institute of
Technology. B.A., M.A., 1973 Yale; Ph.D. 1979, Harvard. I presented an earlier
draft of this paper at a Notre Dame University Symposium on
Rawls's Political Liberalism. I am grateful to members of the audience for their comments, and to Oliver
Gerstenberg and Uday Mehta for helpful suggestions.
... In his review of John
Rawls's Political Liberalism, Stuart Hampshire criticizes
Rawls's assertion that
"under reasonably favorable conditions that make democracy possible, political
values normally outweigh whatever nonpolitical values conflict with them." ... And it excludes the possibility of agreement on a conception of the
common good - for example,
Rawls's difference principle - that extends beyond the good of having a democratic
process. ... Here I will outline the substantive elements of
Rawls's view, trace his path to the idea of an overlapping consensus on a substantive
conception of justice, and locate the precise force of Hampshire's objection.
... More precisely, I will argue that if - I underscore the conditional - if
moral pluralism is consistent with procedural or constitutional consensus, then
it is consistent, too, with substantive, overlapping consensus. ... Of course
the members of the religious minority might say that those limits do raise
issues of fairness - urging that a fair procedure is a procedure that yields
fair outcomes, and that, since the outcome is not, the procedure is not. ...
Introduction: Democratic Pluralism
In his review of John
Rawls's Political Liberalism, Stuart Hampshire criticizes
Rawls's assertion that
"under reasonably favorable conditions that make democracy possible, political
values normally outweigh whatever nonpolitical values conflict with them."
n1 Hampshire finds this
"alarming," and says that it
political liberalism, supposedly not itself a comprehensive morality, into direct conflict with many
comprehensive moralities that are likely to flourish in a modern democracy."
n2 Proceeding from these observations about the plurality of
"comprehensive moralities," Hampshire arrives at pessimistic conclusions about the prospects for consensus
"substantive" conception of justice, whether its content be liberal or traditionalist.
This alarm and pessimism stand in striking contrast to Hampshire's assessment
of the consistency of moral pluralism with agreement on fair procedures:
If [the] distinction between procedural and substantive justice is understood,
between fairness in the process and fairness in the result, the claims for
political liberalism and public reason and for the duty of civility become plausible. Within
different moralities, liberal and conservative, the fairness of the actual
outcome of a conflict will be evaluated differently, even though both sides
recognize the fairness of the adversarial process. Outcomes are by their nature
open to dispute, but processes need not be....
Rawls is implicitly recognizing the distinction between fairness in the procedure
and fairness in the result when he makes democracy an essential element in the
political liberalism. Democracy is a procedure ....
[*590] The distinction between procedural and substantive justice and the associated
Rawls's surfeit of substance echo a complaint in Hampshire's Innocence and Experience:
"Rawls's political or procedural justice is, as he defines it, not narrowly procedural
These criticisms suggest three important claims about moral pluralism and its
implications for an account of justice.
n6 By way of background, let's say that a society is morally - as distinct from
ethnically, culturally or organizationally
n7 - pluralistic if it features a set of moral views whose members are both
comprehensive in scope and incompatible in at least some of their important
n8 Hampshire, then, makes the following points:
(1) Contrary to the extravagant predictions of an end of ideology associated
with theories of social and political modernization, any modern democracy will
be morally pluralistic.
n9 I assume this to be correct, and will not say anything more about it.
(2) Moral pluralism does not defeat hopes for consensus on political
procedures. It is compatible, in particular, with common acceptance that
justice requires fair, democratic procedures for resolving disagreements.
(3) Moral pluralism inevitably generates disagreements about nonprocedural
issues of political justice - for example, issues of fair distribution - making
it unrealistic to hope for agreement on a substantive conception. The relevant
disagreements are not simply matters of negotiable differences of preference or
interest; they are matters of intrinsically zero-sum conflict about what,
beyond fair procedures, justice demands.
n11 Put otherwise: any substantive account of justice - for example, any account
of fair distribution - de
[*591] pends for its justification on a comprehensive conception of the good.
I will use the term
"democratic pluralism" for views that endorse these three claims: in a nutshell, that moral pluralism
is compatible with agreement on rules of a democratic political game, but not
with more substantive agreement.
I disagree with the democratic pluralist view. Procedural and substantive
values come, I will argue, as parts of a package; contrary to Hampshire's
claim, moral pluralism does not drive a wedge between them. In making this
case, I will side with
Rawls against Hampshire. For
Rawls's view is centrally committed both to a substantive conception of justice
n13 and - in
Political Liberalism - to accommodating the moral pluralism that allegedly undermines agreement on
such a conception.
To clarify the terms of the argument and locate the issues on a wider
landscape, though, I need first to fill out the democratic pluralist position.
For democratic pluralism is a familiar view, suggested in conceptions of
American constitutional design advanced by a tradition extending from Holmes
n15 to Ely
n16 to Ackerman
n17 , and in the normative pluralism of democratic theorists from the early Harold
n18 to Robert Dahl.
Take the constitutional variant of democratic pluralism. Putting to the side
the many disagreements among its proponents, it states that the U.S.
Constitution is a design of democratic process. The Constitution establishes a
framework of democratic procedure; it falls to
[*592] the people, acting within that framework, to make substantive value choices,
which are legitimate so long as they have a permissible procedural pedigree.
n20 Rights, principles, and standards that are not ingredients in the ideal of
democratic process - whether guarantees of religious liberty or norms of fair
distribution - are not as constitutionally fundamental as those that are - for
example, rights of voting and political speech, or norms of fair
A similar outlook informs the tradition of democratic pluralism in political
n22 Dahl, for example, holds out the possibility of a consensus extending to the
democratic process, to a principle of equal consideration that provides the
normative foundation for that process, and to certain rights - e.g., the right
to a fair trial - rooted in the idea of equal consideration though not required
for a democratic process. But the range of that extension is uncertain: it is
unclear whether it covers nonpolitical speech, privacy, or the free exercise of
religion. And it excludes the possibility of agreement on a conception of the
common good - for example,
Rawls's difference principle - that extends beyond the good of having a democratic
n23 Given pluralism, Dahl argues, conceptions of the common good are inevitably
"too limited to be generally acceptable or too general to be very relevant and
Different versions of democratic pluralism are advanced for different
purposes, and we may find it compelling as constitutional the
[*593] ory without endorsing it as political philosophy.
n25 Still, the many variants of it are animated by a common thought about the
implications of moral pluralism for modern politics: that the only political
consensus we can reasonably hope for is confined to democratic political
"the right to vote and freedom of political speech and association, and whatever
else is required for the electoral and legislative procedures of democracy ...."
Borrowing terminology from
Rawls, democratic pluralism endorses (as consistent with moral pluralism) the idea of
constitutional consensus - an agreement on the
"political procedures of democratic government."
n27 But it rejects the possibility that
Rawls holds out - a deeper and broader overlapping consensus. A consensus is broader
if it extends beyond political procedures and the rights required for them, to
matters of liberty of conscience and freedom of thought, fair equality of
opportunity, and fair distribution; it is deeper if it reaches conceptions of
the person and such abstract values as fairness, rather than simply rules and
n28 Democratic pluralism, then, rejects depth and breadth not as intrinsically
unattractive, but as incompatible with moral pluralism.
My disagreements with democratic pluralism extend to issues of breadth and
depth. To present those disagreements, I will sketch (in Part I) some
Rawls's conception of justice as fairness, which will serve as an illustration of a
view committed to both procedural and substantive norms. Here I will outline
the substantive elements of
Rawls's view, trace his path to the idea of an overlapping consensus on a substantive
conception of justice, and locate the precise force of Hampshire's objection.
Then (in Part II) I will present a response to the democratic pluralist view,
explaining - to use
Rawls's terms - the connections between constitutional and overlapping consensus. I
will suggest that the distinction between procedure and substance is not a
fundamental distinction in political justification, and
[*594] that, contrary to Hampshire's assertion, democracy is a substantive, not
simply a procedural, ideal.
n30 More precisely, I will argue that if - I underscore the conditional - if moral
pluralism is consistent with procedural or constitutional consensus, then it is
consistent, too, with substantive, overlapping consensus. Procedural and
substantive concerns stand on a common footing in democratic thought.
I emphasize that my skepticism about the depth of the procedure-substance
distinction is equally friendly to three quite different political
philosophies, which might loosely be called
"democratic-egalitarian." The nihilist affirms that procedure and substance are on a par, accepts moral
pluralism, and argues that moral pluralism undermines both procedural and
n31 The communitarian agrees that they are on a par, is concerned that moral
pluralism undercuts both, but, with an eye to affirming both sorts of
consensus, resists moral pluralism - either denying it as a fact or urging
moral consensus as an attractive and reasonable ideal.
n32 The democratic-egalitarian thinks that they are on par, accepts moral
pluralism, but denies that such acceptance defeats the possibility for
substantive consensus. Despite their differences, however, these three views
converge in rejecting the democratic-pluralist claim that moral pluralism
drives a wedge between procedural and substantive values.
I. Substantive Justice and Moral Pluralism
To this point I have followed Stuart Hampshire's practice - the conventional
practice - of using the terms
"substance" without much explanation. I propose now to give some content to the
distinction, and to that end will make a few remarks about the conception of
justice in A Theory of Justice.
A. Reconciling Liberty and Equality
A Theory of Justice proposes a conception of justice for a constitutional
democracy that would allow, as
Rawls puts it,
"a reconciliation of liberty and equality."
Rawls's first principle calls for equal basic liberties, political and civil, and for
assurances of a fair value for political liberty - guarantees that inequalities
of income and wealth do not translate into unequal political influence.
n35 His second principle requires that life chances not be affected by social
class at birth - fair equality of opportunity - and that socioeconomic
inequalities work to the maximal advantage of the least advantaged, mitigating
the effects of the natural lottery - the difference principle.
n36 Assume that the value of the equal liberties to a person is an increasing
function of that person's absolute command of resources.
n37 Then since the first principle assures equal liberties, and the second
principle maximizes the minimum command of resources, the two together require
maximizing the minimum worth of liberty. Thus the reconciliation of liberty and
Let's return now to the distinction between procedure and substance. The best
way to draw that distinction is controversial; but on any way of drawing it,
Rawls's reconciliation of liberty and equality - his concern to present a conception
of justice that draws on the central political values of liberal and
egalitarian traditions in modern democratic thought
n38 - includes elements of substance as well as procedure.
Take the first principle. It includes guarantees of religious liberty, and
liberty of conscience more broadly - not matters of fair proce
n39 To be sure, some aspects of freedom of thought and expression arguably have
procedural value. Consider expression that is, in Cass Sunstein's sense,
"intended and received as a contribution to public deliberation about some issue."
n40 Even in the case of political expression, however, it is not clear that
procedural considerations provide a compelling rationale for stringent
n41 And in any case the procedural rationale is limited: much artistic and
scientific expression is neither intended nor received as a contribution to
n42 Same for rights of bodily integrity and for the liberties associated with the
rule of law. The right to a fair trial is, in a way, procedural, but the
procedure is not political democracy. Moreover, while the fair value of
political liberty is arguably procedural, it is not intrinsic to democracy,
understood simply as a system with universal suffrage and regular, competitive
Or consider the second principle. Fair equality of opportunity is a
substantive, egalitarian ideal, condemning differences in chances that trace to
differences in social position. It is not a matter of fair procedures of
conflict resolution or collective choice. And the egalitarian conception of the
common good expressed in the difference principle,
n43 too, is fundamentally a matter of substance, not procedure.
In short, the proposed reconciliation of liberty and equality is crystallized
in five elements of substantive justice:
(1) rights of conscience;
(2) rights of expressive liberty (not confined to political speech);
(3) fair equality of opportunity;
(4) the fair value of political liberty (perhaps); and
(5) the difference principle.
The first two are associated with liberalism; the next three with
egalitarianism. Put them together, and you have the promised reconciliation of
values of liberty and equality.
It might be said that all these elements of substantive justice are, in
Rawls's own theory, fundamentally procedural. The justification for them appeals to
the original position, a fair procedure of collective choice. I cannot explore
this in detail now. Suffice it to say that it is misleading to think of the
original position as procedural rather than substantive.
n46 The parties in the original position have, for example, various higher-order
interests - including the interest in forming and revising a conception of the
good - and those interests are the sources of some of the requirements of
substantive justice (e.g., liberty of conscience and expressive liberty).
n47 But it is not clear that the original position would be less procedurally fair
if the parties were not assumed to have those interests. Moreover, it would be
no objection to the view I am presenting if procedural ideals issued in
substantive requirements of justice: that would prove, not disprove, that
procedural and substantive values come as parts of a package, and that moral
pluralism does not drive a wedge between them.
B. Reconciliation and Moral Pluralism
I have drawn my sketch of the elements of substantive justice in
Rawls's view entirely from A Theory of Justice. According to
Political Liberalism, the more recent modifications of his views do not require any change in those
n48 Nor do they require any substantial change in the arguments for them, beyond
emphasizing that the arguments are political: that the principles are designed
for the basic structure, that the case for them is to be presented
independently of any comprehensive moral view, and that they express values and
ideals implicit in democratic practice.
The difference in
Political Liberalism is its emphasis on the moral diversity characteristic of a just society. With
liberties of conscience, expression, and association, and the resources needed
to exercise those liberties, citizens will endorse the axioms, so to speak, of
different religions and first philosophies
n50 (many people of course will have, as now, more loose-jointed views with no
n51 Practical reason, operating under the favorable conditions provided by the
basic liberties, does not produce convergence in evaluative conception.
Hampshire's concerns enter here. Consider a society regulated by Rawlsian
principles, or any substantive conception. Why should we expect the moral views
that flourish in such a society to support those principles: why will the
principles be derivable from within or even consistent with those diverse
systems? We can ask a parallel question about the ideal of a fair system of
cooperation among free and equal moral persons.
n52 Citizens in a democratic society may be familiar with those political ideas
and values from the
"political institutions of a constitutional regime and the public traditions of
n53 associated with constitutional democracy. But familiarity famously breeds many
things. Why expect that those political values will be embraced by the moral
conceptions that find favor among the members of such a system? In sum: why
expect an overlapping con
[*599] sensus, broad enough to encompass substantive principles, deep enough to
include conceptions of person and society?
This is not, I emphasize, the same as asking whether our values endorse the
principles, or whether most moral views endorsed over the course of history
have endorsed the principles, or whether reasonable moral views, however
defined, converge in accepting the principles and the political values that
they articulate. The question instead is about the legitimacy of the basic
institutions in a society regulated by a conception of justice that embraces
liberal and egalitarian political values. Formulated as a test for an
acceptable conception of justice (a necessary but not sufficient condition),
the question is: would the comprehensive views that flourish within a society
regulated by a conception of justice that aims to reconcile values of liberty
and equality uphold that conception? Would they endorse a rationale for the
conception that draws on certain abstract ideas of fair cooperation and of
persons as free and equal? And would they accept the requirement that political
justification proceed on the ground made available by that conception and the
abstract ideas associated with it?
C. Revisiting Hampshire
According to Hampshire, the answer to each of these questions is: no. The fact
"moral and religious sentiments ... are in their essence exclusive and divisive" requires that we give up on the ideal of a consensus on substantive principles
n54 Different views - not unreasonable, except in a question-begging sense -
converge only in acknowledging such great evils as
"murder and the destruction of life, imprisonment, enslavement, starvation,
poverty, physical pain and torture, homelessness, friendlessness,"
n55 and in recognizing the need for fair procedures of negotiation to avoid those
evils and ensure a
"peaceful and coherent society."
n56 But in the face of disagreements on substantive justice that follow (he
supposes) from moral disagreement, not all views about the requirements of
justice can be simultaneously satisfied: someone must lose by the lights of
his/her own views
[*600] of justice. That, together with an acknowledgement of the great evils and the
need for fair procedures to avoid those evils, suggests that we can - in a
slogan that I will be returning to later - reasonably demand only to be heard.
By contrast, it is fanaticism to insist as a general matter on winning, to
place one's own substantive views - whether liberal or traditionalist - above
the requirements of fair bargaining.
n57 And it also betrays fanaticism, or perhaps an allegiance to the Platonic myth
"of a sovereign reason which can secure a consensus,"
n58 or maybe simple wishful thinking, to expect agreement on a conception of
justice that embraces and reconciles values of liberty and equality (and
community!). Egalitarian and liberal elements must both be dropped in favor of
the more minimal conception of justice as fair process. That is the force of
the democratic pluralist critique of substantive justice.
II. Connecting Procedure and Substance
I want now to explore this claim about the different relations that procedural
and substantive principles bear to pluralism. I will argue that democracy is,
in its preconditions, implications, and rationale, a substantive as well as
procedural ideal. So if - note again the conditional - pluralism does not
defeat a procedural consensus, then neither does it exclude a deeper and
broader overlapping consensus. By the end, I hope to have made the case that
the ideal of democracy provides a basis for substantive values of liberty and
equality of the kind that justice as fairness aims to reconcile
n59 - that democracy is not simply, as Hampshire says, a matter of procedure, and
that the prospects for overlapping consensus are on a par with the prospects
for constitutional consensus.
To begin with, then, consider some of the substantive preconditions of
democracy. I will distinguish three categories of such preconditions
(1) rights implicit in the idea of democratic process;
(2) rights and other conditions that are not implicit in the idea but required
as a supporting framework for an open democratic process; and
(3) rights and other conditions required to encourage consent to democratic
I emphasize that these categories do not correspond to particular rights and
conditions, but instead specify the abstract roles of rights and other
n61 A single right may, then, play several roles, and this will be an important
source of complexity in the right itself.
1. Implicit Rights and Conditions
Rights of suffrage are, in one way, transparently procedural. Without such
rights, we have no democratic procedure; limit them, and you limit the
inclusion (participation) and perhaps also the contestation (competition for
office) essential to democracy.
n63 But a variety of political liberties that are less transparently procedural
might also be described as implicit in the democratic process: they can be
provided with a rationale that connects them reasonably directly with
n64 Take, for example, the ideal of
"an open and effective democratic process,"
n65 and the associated conception of political legitimacy: that outcomes are
legitimate if they trace to such a process. Arguably, liberties of political
expression and association are implicit in this ideal; while their connection
with the democratic process is less transparent than the connection of suffrage
rights, they must be in place for a political process to be open and so for
outcomes to be legitimate.
This point is of course welcome to democratic pluralists: indeed I take it
from them, and do not suggest that rights of expression, for example, are
substantive. (They do not fall into the five categories noted earlier.
n67 ) Agreement on procedures is in part agreement on rights that are, as Dahl
"part of the very conception of the
[*602] democratic process."
n68 But it is important to note that if the democratic pluralist rests the case
for rights of political expression and association on the claim that they are
part of the
"very conception" of democracy, then that conception must be richer than the minimal idea of
democracy as a process characterized by regular elections, competing parties
(political contestation), and universal suffrage (political inclusion). The
conception must also include the idea that the political process is open, in
ways that supplement those requirements. To explore the implications of this
openness, consider the second category of preconditions: the rights and
conditions that provide a supporting framework for democratic process.
2. Supporting Framework
One reason for thinking that the democratic process ought to be open - in the
ways that are advanced by rights of political expression and association - is
rooted in the value of fairness. For example, a process may fail to be open
because it includes viewpoint-based restrictions on political expression. Such
restrictions, though they apply to everyone, are unfair because they
effectively disenfranchise those whose views they exclude.
n69 Recall a central idea of democratic pluralism: that in the face of moral
pluralism it is unreasonable to insist on winning, but reasonable to insist on
being heard. People are effectively disenfranchised - not heard - when their
views and interests do not count for anything because they are barred, despite
their political rights, from exerting influence on the results of the political
n70 Given effective disenfranchisement, it is not true that
"interested parties have reasonable access."
Suppose, then, that the rationale for an open democratic process - the kind of
process that democratic pluralists endorse - is based on a concern for fair
influence. Then the commitment to openness has implications that extend beyond
protections of the rights of political expression and association that are
required to give substance to rights of participation. Considerable controversy
surrounds the precise statement of a requirement of fair influence. Proposals
might range from a requirement of equal influence, to an assured minimum
[*603] value of political rights guaranteed, for example, by providing public support
for education, or decent welfare minima,
n72 to a ban - as with
Rawls's own fair value of political liberty - on inequalities of political influence
that derive from unequal resources, achieved by supplementing education and
welfare minima with floors under political expenditures or ceilings on them.
n73 What seems undeniable, though, is the case for moving from democratic process
to openness, then to fairness, and finally to fair influence - that is, for
broadening a constitutional consensus to include some assurances of a
supporting framework of rights and conditions that are neither transparently
procedural, as are suffrage rights, nor directly required for openness, as are
rights of political expression.
A different route will take us to a similar conclusion. I just suggested that
an open political process is required to ensure fairness. But it is also
important for informed and intelligent collective decisions, with
"genuine argument and counter-argument."
n74 Consider again the case against viewpoint-regulation in the area of political
expression. Such regulation is not simply unfair to those whose views are
restricted; it is also troubling from the point of view of reflective and
deliberative collective choice. In Meiklejohn's terms, it impairs
"the thinking process of the community."
This deliberative case for free expression again provides a rationale for
broadening the constitutional consensus, extending it to substantive assurances
of, for example, an educated citizenry and the public good of well-conducted
political debate. As the Massachusetts Supreme Judicial Court said in a recent
The Commonwealth has a duty to provide an education for all its children, rich
and poor [and] this duty is designed not only to serve the interests of the
children, but, more fundamentally, to prepare them to participate as free
citizens of a free State to meet the needs and interests of a republican
3. Conditions of Consent
The third category of preconditions I will call
"conditions of consent." Intuitively, the idea is that consent to the outcomes of a democratic process
requires that certain fundamental interests not be effectively
"up for grabs" in that process.
Consider a religiously pluralistic society, with a constitutional consensus.
Everyone agrees to resolve their disagreements through democratic processes,
and all the rights that are plausibly required for the democratic process are
in place and form part of the consensus. Assume, too, that there is no
consensus on the free exercise of religion, no overlapping consensus, as
Rawls defines it.
Suppose now that a constitutional amendment is passed requiring Sunday
worship. The amendment does not impose a religious test for office; that would
directly implicate the democratic process, inconsistent with the assumption of
constitutional consensus. Despite the amendment, then, universal political
liberties and liberties of political expression remain fully in place,
permitting opponents of the amendment to argue for its repeal.
So we have a case in which limits on rights of free exercise do not directly
implicate the fairness of the procedures of collective choice. Of course the
members of the religious minority might say that those limits do raise issues
of fairness - urging that a fair procedure is a procedure that yields fair
outcomes, and that, since the outcome is not, the procedure is not. Assessments
of the fairness of procedures - from race-conscious districting to requirements
of supermajorities - are, of course, commonly assessments of the merits of the
procedures at protecting important interests and values. But if a charge of
unfairness is raised, then there is not really the consensus on the fairness of
the procedure, which, for the purposes of the example, I have stipulated to
hold. (I will say more on this issue below.)
Now there is much to be said about what is wrong with the Sunday worship
amendment. I want here to confine myself to the following question: Is
consensus on a democratic process that permits this result really more
plausible than consensus on a democratic process that rules it out? Let me fill
out the example. I am assuming that members of the different religious
traditions take their religious views to impose obligations on them, including
obligations about the day of
[*605] worship; seen from the inside, so to speak, the requirements are not matters
of preference or choice. But how, then, could they possibly consent to a
process which might realistically issue in regulations of their day of worship?
The prospects of procedural agreement - of a shared willingness to accept a
constitutional scheme of collective choice - are increased, then, when that
constitution also works to
"reduce the stakes of political battles."
n78 And such reduction must proceed in part by ensuring that certain nonprocedural
rights and interests will not be subject to political interference.
One response to this, institutionally speaking, is a court with powers of
judicial review which has the responsibility to protect religious liberty.
n80 But I am not making a case here for judicial review, and other institutional
possibilities may do as well - including a less majoritarian, more consensual
(say, consociational or supermajoritarian) form of democracy that effectively
institutionalizes a requirement of general approval of legislation touching on
n81 Pointing to this latter possibility, the democratic pluralist might say that
the example does not really show a need for an agreement on something other
than a fair democratic procedure but rather shows what form of democracy is
fair, given circumstances of religious division.
But that gives away the game. Let me explain.
Democratic pluralism appeals to the idea of a procedural consensus. We can
think of such a consensus in either of two ways. According to the first, the
domain of the procedural is defined by certain practices, say the practice of
proportional representation, or of judicial review. According to the second, it
is defined by certain values. Let's say that a procedural value is a value used
for the assessment of procedures without regard to the results of those
procedures. For example,
"fair influence" is a procedural value if we understand it as requiring that differences of
influence not be explained by class position, and do not defend this value by
reference to its importance for equitable outcomes.
I am taking the idea of a procedural consensus in the second way, as a
consensus on procedural values. But - coming back to the point about giving
away the game - if we say that, for example, a more consensual (consociational
or supermajoritarian) democracy is procedurally more fair because it better
protects interests in the free exercise of religion, then we are agreeing that
procedural agreement depends in part on substantive agreement: not, to be sure,
on religious agreement, but agreement on the importance of the interest in free
exercise (the interest in fulfilling what a person takes to be his or her
religious obligations). Or suppose it is argued that a form of democratic
process that protects nonprocedural liberties is not more fair as a process,
but simply more likely to be stable. Even that suffices for the purposes of the
argument about the substantive preconditions of democracy. For that argument
aims to show that agreement on the value of fair democratic process is, morally
speaking, no more demanding than agreement on what are conventionally
understood as nonprocedural matters. So if agreement that justice requires fair
procedures is consistent with moral pluralism, then - given the connections
between procedure and substance - agreement on substantive requirements is also
consistent with moral pluralism: e.g., requirements ensuring basic liberties of
conscience and the fair value of political liberty.
I now turn to a second line of argument, focused not on the substantive
preconditions of democratic process but on its implications, that is, on the
appropriate outcomes of a democratic process. To introduce the second category,
I will start with a detour through a point about procedural and substantive
justice that Hampshire makes in Innocence and Experience.
1. Hampshire on Substantive Justice
Hampshire distinguishes procedural justice, which imposes
"absolute duties," from substantive moralities, including substantive conceptions of justice.
Duties to support procedural justice - by normally complying with fair
procedures - are absolute in two ways: they are not derived from or dependent
upon particular substantive conceptions and, though not always overriding, they
normally take precedence over the values of such conceptions.
n82 An acknowledgement of the values of procedural justice might arise from a
recognition that pain, suffering, cruelty, humiliation, starvation, and misery
are great evils;
n83 an awareness that fair procedures of dispute resolution are required to
prevent such evils; and a general unwillingness to subordinate a concern to
avoid those great evils to any of the substantive values of one's own moral
n84 The duties are absolute, then, in part because that recognition, awareness,
and unwillingness are available to people who endorse a wide range of
substantive moral doctrines - to the proponent of liberal justice, who also
attaches special weight to individual rights, perhaps for reasons of autonomy
or self-improvement; and to the proponent of a traditionalist conception,
troubled about the corrosive effects of rights on the community's way of life.
The conjunction of the fact of moral diversity
n86 with the absolute value of fair procedures might suggest that all questions of
substantive justice are, so to speak, up for grabs: the justice of a practice
would depend on the society one is asking about and what its members would
agree to. That, in turn, would rest on the substantive moralities prevalent in
But Hampshire rejects this relativism.
n87 He mentions five practices, not themselves matters of fair procedures for
dispute resolution, each of which is unjust under all circumstances. Slavery is
one, but so, too, are royal absolutism, unregulated factory labor, the absence
of fair opportunities for women, and the unlimited accumulation of vast
n88 Of course, people have not always regarded them as unjust. But the failure to
recognize their injustice must be understand as
"the blindness of reasonable persons."
Hampshire's thought seems to be that the survival of certain practices - for
example, slavery - is contingent on the common perception of them as natural
and unchangeable and so not proper topics for a practical reason concerned with
what is attainable through action. But the scope of practical reason is not
permanently fixed. And once these practices are subjected to practical
challenge, they become proper topics of public discussion, depriving the
metaphysical rationale of its force. But no other rationale is available for
n90 Put together, then, the idea that justice is a matter of fair procedure with
the power of deliberative reason to dissolve false necessities, and you have
the conclusion that certain practices are unjust because they could not ever be
the outcome of a fair procedure of dispute resolution, in which the
participants are not blind to real possibilities.
It is not clear from Hampshire's account precisely why they could not emerge -
how the conjunction of an expanded sense of possibility with a procedural view
of justice yields the injustice of slavery, absolute monarchy, unregulated
factory labor, the exclusion of women, and the unlimited accumulation of
wealth. And that uncertainty reflects the fact that Hampshire - as is commonly
true with defenders of procedural views - does not explain what makes a
n92 If, for example, unconstrained majority rule is a fair procedure, then any of
the unjust practices could be approved of, given the right distribu
[*609] tion of preferences; similarly if bargaining from a Hobbesian state of nature
is a fair procedure.
I assume, then, that Hampshire has in mind a richer and more attractive
conception of a fair procedure: not simply as an institutionally tamed conflict
of interests, but as a procedure in which people advance reasons in favor of
outcomes, reasons reflecting their own comprehensive moral conceptions. His
point, then, is that certain practices, once subjected to challenge, are
revealed as wholly lacking in rational defense. However much some people may
like and benefit from them, no adequate reasons can be given for the burdens
This, then, gives us another basis for rejecting a fundamental
procedure-substance distinction: begin with a conception of fair procedure and
a consensus on that conception. Practices are unjust if they could not be the
outcome of that procedure because they could not be given a rational defense by
those who accept the fairness of the procedure, given a correct understanding
of real possibilities. Their injustice is, in principle, no more controversial
than the justice either of the procedures, or of the preconditions that must be
in place for those procedures themselves to obtain. In particular, it can be
established without relying on any specific conception of the good, if the
justice of the procedures can be so established.
2. Democratic Procedure and Depth
The claims of substantive justice that Hampshire thinks are not socially
relative are of very great historical importance; but they are also rather
limited: in particular, they are not sufficiently broad to cover the five
elements I identified earlier in my remarks on
Rawls's reconciliation of liberty and equality (liberty of conscience, etc.).
Moreover, on the dimension of depth, they do reach down to the conception of
citizens as free and equal. These limits on breadth and depth are, however,
partly a result of the very abstract description of the procedure that
constrains substantive justice: thus far, we have said only that it is an
institutionalized, fair process of reason giving. This abstractness can,
however, be remedied. For recall that we are assuming a constitutional
consensus. So our concern is not merely with the substantive implications of
fairness or reason giving generi
[*610] cally understood, but with the substantive implications of consensus on a
specifically democratic procedure of conflict resolution.
Think, then, of the democratic process as one kind of institutionalized
process of reason giving. What distinguishes it is the requirement of openness,
of universal and fair access to political institutions: a strong condition of
inclusion, which makes political access independent of social position or
n94 Democracy, in short, is a procedure that institutionalizes an idea of citizens
as equals. Agreement on the democratic process - a constitutional consensus -
must also, then, constrain what can count as an acceptable reason within that
process. 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 others or that the
interests of one group are to count for less than the interests of others. And
that constraint on reasons will limit the substantive outcomes of the process,
in ways that supplement the limits set by the generic idea of a fair procedure.
Of course people can and often do violate this constraint. My claim is that
moral pluralism causes no more trouble for the idea of a population who agree
with the deeper political ideal of treating people as free and equal, and see
their acceptance of the results of a fair democratic process as simply one
aspect of that more comprehensive ideal, than it does for the idea of a
population who accept the results of a democratic process because the process
is open. In short, moral pluralism does not divide a more superficial
constitutional consensus from a deeper overlapping consensus.
3. A Process of Deepening?
I have presented the point about equality and deepening as, in the first
instance, simply an analytical matter, a claim about the implicit commitments
of constitutional consensus. But reflection on the conduct of political
argument under conditions of constitutional consensus suggests that it may have
some practical force as well. A stable democratic process, in which individuals
and parties seek to win support for their views and projects, puts some
pressure on views to endorse the deeper idea of citizens as equal persons.
Dahl, for example, claims that the rationale for democracy lies in part in the
idea of the intrinsic equality of human beings and the asso
[*611] ciated requirement of giving equal consideration to the interests of citizens.
n95 And he emphasizes that this requirement limits acceptable outcomes of the
democratic process: it condemns discriminatory practices and mandates
protections for a variety of rights and interests that are not preconditions
for democratic process. Moreover - and this is the crucial point here - he
"rough pattern" in the evolution of public opinion in stable polyarchies:
"the idea of Intrinsic Equality ... has steadily gained strength as an element
in the constitutional consensus and political culture."
n96 Dahl urges that stable democracy requires constitutional consensus: a
widespread belief in the value of democracy and
"habits, practices, and culture" suited to that belief.
n97 But he suggests as well that constitutional consensus tends to deepen:
"public opinion in democratic countries tends to move toward an ever more
inclusive commitment to ideas like intrinsic equality and equal consideration."
n98 And the strengthening of those ideas as elements of the public culture leads
in turn - as an empirical matter, and not simply as a matter of analytical
argument - to constraints on substantive outcomes of the democratic process,
greater protections of nonpolitical rights and interests.
Dahl does not describe mechanisms that lead to this strengthening. But it is
not hard to see how the deepening he describes might be generated by norms and
practices of political argument under conditions of constitutional consensus.
Consider the evolution of the political rhetoric and project of socialist
parties in this century.
n99 They begin the century with a class project, presenting themselves as agents
of the industrial working class; they expect the maturation of capitalism to
turn the working class into a majority of the population; and they understand
that they can only sustain their claim to serve as an agent of the working
class if they participate in democratic politics, winning near term gains by
But the identification of the industrial working class with the emerging
majority is overturned by events. And that sharpens the electoral dilemma of
socialist parties: to win gains through elections they need to win elections.
But winning elections means extending
[*612] their political appeal beyond the working class.
n100 The result is that socialist parties - those that preserve electoral
commitments - universalize their appeal, addressing themselves to all citizens
as equals. They reinvent a universalistic politics, addressed to citizens
generally, regardless of social position.
n101 According to Adam Przeworski, from whom I have borrowed this sketch:
Differentiation of the class appeal ... reinstates a classless vision of
politics. When social democratic parties become parties of
"the entire nation," they reinforce the vision of politics as a process of defining the collective
"all members of the society." Politics once again is defined on the dimension individual-nation, not in
terms of class.
Perhaps, then, political argument under conditions of constitutional consensus
itself encourages the conception of citizens as equals, thus deepening the
terms of agreement. Moreover, to return now to the issue of fair process and
permissible outcomes, this deepening in turn sets constraints on what can be
agreed to in a fair procedure by limiting what can count as a reason within
such a procedure. I have not yet examined the precise implications of these
constraints. But that must wait, since those implications will be best
understood after we have set out the third link between procedure and
substance: the connection I have labelled
In motivating a substance-procedure distinction, democratic pluralists say -
as I noted earlier - that given the diversity of interests and values, it is
unreasonable to insist on winning, but acceptable to insist on being heard. But
why is it reasonable to insist on being heard, as that requirement is
understood under democratic conditions? Why is it reasonable to insist on
access to open arenas of authoritative, collective decision-making? Addressing
these questions - understanding the rationale of democratic procedure - will
carry us, once more, beyond process to substance.
1. Expression and Recognition
Without making any claim to completeness, I want to mention two especially
important considerations that support the reasonableness of a demand to be
heard. They arise out of basic interests in expression and recognition.
The interest in expression is an interest in articulating thoughts, attitudes,
and feelings on matters of personal or broader human concern, and perhaps
through that articulation influencing the thought and conduct of others.
n104 One aspect of the expressive interest is an interest in stating views about
public affairs because of strongly held moral, religious, or political
convictions about the proper conduct of politics, broadly understood. We do not
all hold such convictions with the same intensity, scope, or determinateness.
But most citizens have conscientiously held convictions that, in at least some
cases, provide them with compelling reasons for addressing public affairs. And,
I suggest, claims to equal political rights are in part rooted in the fact of
those convictions. To be sure, under conditions of moral pluralism, citizens
sometimes have opposing convictions and the reasons that some regard as
compelling, others reject as insubstantial. Still, the failure to acknowledge
the weight of those reasons for the agent - even if one does not accept them -
and to acknowledge the claims to participate that emerge from them reflects a
failure to accept the terms of an open democratic process, to endorse the
constitutional consensus and the deeper idea of citizens as equals.
Claims to equal political standing are fueled, too, by the connection between
such standing and a sense of self-worth, a connection rooted in the public
recognition associated with equal standing.
n106 Rousseau urged this point in his argument for direct democracy:
"Once the populace is legitimately assembled as a sovereign body, all
jurisdiction of the government ceases; the executive power is suspended, and
the person of the humblest citizen is as sacred and inviolable as that of the
n107 Rousseau's affection for a
[*614] democracy of popular assemblies is not widely shared. But even if we reject
his critique of representation, we can accept the force of his point: that
rights of participation, political expression, and association are important in
part because they provide public support for a sense of self-worth by providing
public recognition of the equality of citizens, whatever their differences in
human and social circumstances.
Suppose, then, that these expressive and recognitional interests contribute to
the rationale for an open democratic process - the object of constitutional
consensus. We need to ask now whether the implications of the expressive and
recognitional interests can be confined to political procedures, or if instead
they have substantive implications as well.
Take first the expressive interest. Recall that the expressive interest is
implicated in open democratic procedure because important reasons, as
identified by a person's moral, religious, or political convictions, require
that person to address public matters. But such convictions also provide
important reasons for expression on nonpublic matters: the expressive interest
is not confined to speech that is
"intended and received as a contribution to public deliberation about some issue."
n109 Because the reasons require expression on nonpolitical matters, the case they
provide for expressive liberties is not confined to discussion of issues of
public concern. Moreover, given the role of religious and other conscientious
convictions in setting a person's obligations, they provide a case for rights
of conscience and religious liberty in areas with no special bearing on public
In short, the expressive interest, which provides part of the case for an open
democratic process, also provides a rationale for substantive rights that are
not implicit in or required for democratic procedure, and it strengthens the
case noted earlier for basic liberties required to encourage consent to a
A parallel case can be made about the interest in recognition, particularly
when constitutional consensus deepens along the lines suggested earlier to
include a conception of citizens as equals and associated conditions on
acceptable political reasons.
n110 Beyond an open democratic process and the requirement of equal standing within
[*615] that process, the interest in recognition suggests the importance of
nonprocedural norms - of a public understanding about equal opportunity and the
fair distribution of resources that severs the fate of citizens and the worth
of their liberties from the differences of social position, natural endowment,
and good fortune that distinguish the free and equal members of a well-ordered
Take the difference principle as an illustration.
n111 Assume that it is a matter of public understanding that justice demands
maximizing the minimum, that inequalities can only be justified by the benefits
they confer on the least advantaged, that citizens must forgo advantages when
they would reduce expectations at the minimum. That understanding - part of a
public sense of justice - serves as a way to express respect for those at the
minimum and fully to affirm their worth as equals, supplementing the
affirmation that is provided by rights to political liberties and a fair value
of those rights. The difference principle does not simply require that
departures from the distributional status quo work to the advantage of the
least advantaged - that a rising tide lift all ships from their current
positions. Instead it treats equality itself as a baseline, requiring that
existing inequalities themselves are to the maximal advantage of the least
advantaged. In this way, the difference principle provides an especially strong
affirmation of the interest in recognition, suited to an understanding of the
equality of citizens that, I suggested, a commitment to constitutional
consensus itself encourages.
In short, start from a consensus on an open political process. Now constrain
the process leading to agreement and the reasons that can figure in it by an
idea of equality that emerges from the deepening of constitutional consensus.
Add in the interest in recognition, which is itself part of the rationale of
democratic process. The result is a case for the substantive norms of equal
opportunity and fair distribution - a case that is no more dependent on appeal
to a comprehensive conception of the good than is the case for the process
That completes the case for connecting constitutional and overlapping
consensus. I have sketched some links between procedural
[*616] values and important elements of substantive justice, and indicated as well
how a constitutional consensus might deepen. More precisely, I have argued that
an attractive conception of democratic process must include ideas of openness
and reason giving. But any conception of democratic process that includes those
ideas will also provide the resources for an account of substantive, and not
only procedural, justice.
To conclude, I want briefly to consider three objections. I address these
because they articulate a residual sense that there must be something right
about the democratic pluralist's claim that moral pluralism forces the
procedure-substance distinction on us. That claim can, I think, seem
intuitively obvious in ways that immunize it from complicated (nonintuitive)
considerations of the kind advanced in this paper. So I need to say something
about what prompts people to find the democratic-pluralist view so intuitively
plausible, and to respond to those promptings.
The first objection emphasizes that substantive agreement is more demanding
than procedural in the perfectly straightforward sense that the substantive
supplements the procedural; as comprehensive moral disagreement grows, then,
the range of political values that can be supported by an overlapping consensus
must correspondingly narrow. And as it narrows, the plausibility that we will
be able to preserve the
"surplus" of agreement on substance over and above a more minimal agreement on procedure
It is certainly true that, as comprehensive moral disagreement grows, the
range of political values that can be supported by an overlapping consensus
correspondingly narrows. But this observation fails as a defense of a
fundamental procedure-substance distinction. In fact, it assumes what needs to
be shown: that disagreement might plausibly drive a wedge between procedure and
substance, that procedural commitments are, in some relevant sense, more
minimal than substantive commitments. Consider an analogy. Suppose the proposal
is made to confine the term
"theorem" to propositions whose provability is not in dispute between constructivists
and realists in the philosophy of mathematics. Take someone who now endorses a
realist philosophy and accepts this proposal; he/she will start to use the term
"theorem" more narrowly. And as we extend the range of constructivisms that need to be
accommodated (as they endorse increasingly strict forms of finitism), the class
of theorems, according to the new usage, will narrow. But we can be sure of one
thing: it cannot happen that we will end up with theorems about even numbers
and not about odd. The underlying disagreements may grow; but they will never
produce a cut
[*617] between odd and even numbers. By analogy, moral pluralism may reduce the range
of possible agreement, and it will certainly exclude certain substantive ideas.
But it will not yield a fundamental cut between procedure and substance. That
simply is not a deep distinction.
The thought that it is - and here I come to the second objection - derives
from the fact that people say, truthfully, that there are outcomes they find
objectionable but which they would be willing to tolerate if they were given a
fair chance to make their case. They do not regard the fact that a fair
political process produces a bad legislative outcome - even a deeply morally
objectionable outcome - as sufficient reason for condemning the process as
unjust and for urging its replacement with an alternative.
But that is simply to assert what is not in dispute: that assuring fair
procedures is an important political value, and that outcomes can sometimes be
justified by being shown to result from a fair procedure. It does not follow
from this undisputed truth that the only fundamental political values are
A final reason for thinking that the view I have advanced cannot be right - or
anyway, cannot capture what
Rawls is now arguing - is that I have not really taken moral pluralism seriously. I
have not started from divergent moral conceptions, aimed to find points of
convergence among them, and sought to show in particular that they converge in
accepting substantive norms of justice.
This third objection reflects an unnecessarily narrow picture of how political
argument is to accommodate moral pluralism, and an important confusion about
Rawls's idea of a political conception of justice. Such accommodation need not deny
the autonomy of political argument and turn political philosophy into a search
for points of convergence among comprehensive moral views. Put otherwise,
Rawls's idea of
political liberalism is not that reasonable moral views converge on a common understanding of
justice. Instead, the idea is to present, in the first instance, a complete
political conception of justice, without drawing on or referring to
comprehensive moral views. Then, with such a complete conception on hand, we
can consider whether it would be supported by the range of reasonable moral
conceptions that we expect to arise in a society governed by it. This paper has
operated solely at the first stage - the stage of freestanding political
argument that articulates and works out the implications of a set of ideas
without presenting them as dependent on or rooted in any com
[*618] prehensive moral view.
n112 Whether these ideas can be supported by an overlapping consensus is another
matter. But for my purposes the freestanding argument suffices. For my point,
as I indicated at the outset, is that moral pluralism causes no more trouble
for agreement on substance than for agreement on democratic procedure. Since
the democratic pluralist does not exclude such a consensus - Hampshire finds
nothing alarming in it - this suffices as a response to the democratic
n1. Stuart Hampshire, Liberalism: The New Twist, N.Y. Rev. Books, Aug. 12, 1993,
at 44 [hereinafter Hampshire, The New Twist].
Rawls says roughly what is attributed to him by Hampshire in the text in John
Rawls, Political Liberalism 139, 155-56, 168-69, 209 (1993) [hereinafter
n2. Hampshire, The New Twist, supra note 1, at 44.
n4. Id. at 46.
n5. Stuart Hampshire, Innocence and Experience 187 (1989) [hereinafter Innocence
n6. Each of these themes is developed in detail in id.
n7. On the role of organizational pluralism in a substantive conception of
& Joel Rogers, Secondary Associations and Democratic Governance, 20 Pol.
& Soc'y 393 (1992).
n8. A moral conception is comprehensive
"when it includes conceptions of what is of value in human life, and ideals of
personal character, as well as ideals of friendship and of familial and
associational relationships, and much else that is to inform our conduct, and
in the limit to our life as a whole."
Political Liberalism, supra note 1, at 13.
n9. I do not want to suggest that Hampshire thinks - or that I think - that moral
pluralism is either peculiarly modern, distinctively democratic, or a bare,
unexplained fact. In Innocence and Experience he traces pluralism to the
imaginative powers that are essential to human nature. See in particular
chapters one and four. I believe that Hampshire's discussion (especially in
chapter four) ties moral pluralism too closely to the good of individuality,
but I cannot pursue this point here.
"Universal agreement can be expected, in the name of rationality, only on the
methods of fair argument which will arbitrate between the different answers to
these questions [of substantive justice], when an answer is needed for public
purposes and social arrangements." Innocence and Experience, supra note 5, at 108. For an important defense of
this assertion, see id. at 142-46.
n11. See Innocence and Experience, supra note 5, at 61, 141.
n12. In discussion, John
Rawls has emphasized this formulation of Hampshire's position.
n13. See infra text accompanying notes 34-37.
Political Liberalism, supra note 1, at xvi-xviii.
n15. See for example Holmes's dissent in
Lochner v. New York, 198 U.S. 45, 74 (1905).
n16. See John Hart Ely, Democracy and Distrust (1980).
n17. See 1 Bruce Ackerman, We The People (1991) [hereinafter 1 Ackerman, We the
People] (especially see chapter one). Ackerman's important distinction between
monist and dualist democratic interpretations of the U.S. Constitution does not
affect my discussion here.
n18. See the essays by Laski in The Pluralist Theory of the State (Paul Q. Hirst
ed., 1989) (especially see Law and the State). Laski was principally concerned
to reject the modern conception of sovereignty, and to encourage administrative
decentralization. But along the way he suggests a view of the kind described in
n19. See Robert A. Dahl, Democracy and Its Critics (1989) [hereinafter Dahl,
Democracy]; Robert A. Dahl, Preface to Democratic Theory (1956) [hereinafter
Dahl, Preface]. Preface to Democratic Theory emphasizes the importance of
"on policy" in a pluralistic democracy.
Id. at 132. But it is not clear what
"the underlying consensus on policy" includes, and in what ways it is procedural. The connections of, distinctions
between, and institutional consequences of procedural and substantive agreement
are central themes in Democracy and Its Critics. Dahl, Democracy, supra, at
163-92, 280-310. On the relationship between the earlier normative pluralism
and modern pluralism, see Earl Latham, The Group Basis of Politics: Notes for a
Theory, 46 Am. Pol. Sci. Rev. 376 (1952); Robert A. Dahl, Pluralism, Polyarchy,
and Scale, in Democracy, Liberty, and Equality 226 (1988).
n20. More precisely: for Ely, the government, acting in the name of the people, is
to make the substantive value choices; the Supreme Court, acting in the name of
the Constitution, is to ensure that those choices are made through a genuinely
representative process. Ely, supra note 16. For Ackerman, the people themselves
make substantive value choices in periods of heightened political engagement
(exemplified by the New Deal); the role of the Court is to police the processes
of normal politics and to ensure that the government acts within the bounds set
by the people in periods of constitutional politics. 1 Ackerman, We the People,
supra note 17. But for both Ely and Ackerman, the Constitution is fundamentally
a design of democratic process. So there could, for example, be no doubt about
the constitutional validity of a partial repeal of the First Amendment
establishing Christianity as an official religion. See 1 Ackerman, We the
People, supra note 17, at 3-16; Ely, supra note 16. In defending his account of
the Constitution, Ackerman emphasizes that the U.S. Constitution - unlike the
German - does not include entrenchment clauses protecting substantive rights
from constitutional amendment. 1 Ackerman, We the People, supra note 17. This
point is of uncertain significance, since the Constitution has no clauses
entrenching procedures either.
n21. For example, we ought not to appeal to them in interpreting the open-ended
clauses of the Constitution.
n22. This paragraph summarizes themes in Dahl, Democracy, supra note 19, at 163-92,
n23. For criticisms of pluralist skepticism about the common good, see Theodore J.
Lowi, The End of Liberalism: The Second Republic of the United States (2d ed.
n24. Dahl, Democracy, supra note 19, at 283.
n25. Ely's case for democratic pluralism is based centrally on its alleged fit with
the Constitution, not on general considerations about justice or political
legitimacy. See Ely, supra note 16, at 87-101. For criticism of the case, see
Laurence H. Tribe, The Puzzling Persistence of Process-Based Constitutional
89 Yale L.J 1063 (1980).
Political Liberalism, supra note 1, at 159.
n28. On breadth and depth, see id. at 149-50, 158-59, 164-67.
Rawls presents his discussion as a response to Kurt Baier, and does not connect the
distinction between constitutional and overlapping consensus with the broader
tradition of democratic pluralism. These connections are suggested, however, by
his remarks on Ackerman and the idea of constitutional breakdown in id. at
n29. This statement requires certain qualifications in the case of Dahl's view as
presented above, though the qualifications do not undercut the central point of
& Joel Rogers, On Democracy 146-83 (1983);
Joshua Cohen, Deliberation and Democratic Legitimacy, in The Good Polity: Normative Analysis
of the State 17 (Alan Hamlin
& Philip Pettit eds., 1989);
Joshua Cohen, The Economic Basis of Deliberative Democracy, 6 Soc. Phil.
& Pol'y 25 (1989).
n31. See, for example, Duncan Kennedy, Form and Substance in Private Law
89 Harv. L. Rev. 1685 (1976).
n32. Michael Walzer's emphasis on the role of shared understandings in procedural
and substantive political argument might be understood as denying moral
pluralism as a fact; Michael Sandel's account of the limits of justice, and
Alasdair MacIntyre's concerns about interminable moral disagreement, suggest
that moral consensus is an attractive and reasonable ideal. See Michael Walzer,
Spheres of Justice: A Defense of Pluralism and Equality (1983); Michael J.
Sandel, Liberalism and the Limits of Justice (1982); Alasdair MacIntyre, After
Virtue: A Study in Moral Theory (2d ed. 1984).
Rawls, A Theory of Justice 204 (1971) [hereinafter A Theory of Justice].
n34. See id. at 204.
n35. On the basic liberties, see
Political Liberalism, supra note 1, at 294-99; on the fair value of political liberty, see id. at
Rawls sharply criticizes the Supreme Court's claim in
Buckley v. Valeo, 424 U.S. 1, 48-49 (1976), that regulations to ensure fair value represent impermissible abridgements of
expressive liberty, that, in the Court's words,
"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." See
Political Liberalism, supra note 1, at 359-63. The Court has now taken an important first step in
rejecting the claim in
Buckley. Austin v. Michigan Chamber of Commerce, 494 U.S. 652 (1990) (upholding a Michigan law prohibiting corporations from using general treasury
funds for independent expenditures in connection with state candidate
elections) represents a fundamental break with the Buckley framework. It
expands the conception of
"corruption" beyond quid pro quo to concerns about unfair influence.
n36. A Theory of Justice, supra note 33, at 56-83.
n37. That is, not dependent on their relative position in the distribution. This is
not true for the political liberties. See the discussion of the
"limited space" of the political process in
Political Liberalism, supra note 1, at 324-31 (especially see 328-29). Thus the special requirement
of fair value. For a more general discussion of relative positions, see A
Theory of Justice, supra note 33, 530-41.
n38. As is suggested in
Political Liberalism, supra note 1, at 9, 22, 34-35.
n39. See 1 Ackerman, We the People, supra note 17, at 14; Ely, supra note 16, at
94. According to Ely, attributions of nonpolitical functions to the First
Amendment's free speech guarantee suffer from the
"smell of the lamp." But he also says that
"the obvious cannot be blinked: part of the explanation of the Free Exercise
Clause has to be that for the framers religion was an important substantive
value ...." Id. This combination of views is puzzling. Why insist that the Free Speech
Clause must be interpreted procedurally, given the substantive basis of the
Free Exercise Clause?
n40. This is Sunstein's definition of political speech in Cass R. Sunstein,
Democracy and the Problem of Free Speech 130 (1993). Variations on the idea
that the Constitution's free speech guarantee is about democracy have been
advanced in Alexander Meiklejohn, Political Freedom (Oxford Univ. Press 1968)
(1948); Robert H. Bork, Neutral Principles and Some First Amendment Problems,
47 Ind. L.J. 1 (1971); Owen M. Fiss, Why The State?,
100 Harv. L. Rev. 781 (1987); Ely, supra note 16.
n41. For criticism of the procedural rationale, see Ronald Dworkin, The Forum of
Principle, in A Matter of Principle 33, 61-65 (1985).
n42. I believe that it is more limited than
Rawls thinks appropriate, but his discussion of freedom of expression in
Political Liberalism, supra note 1, at 331-40 is not clear on this issue. In places, he emphasizes
the connection between speech, the capacity for a sense of justice, and
politics. E.g., id at 335, 337. But the general framework of argument about
basic liberties draws on both the capacity for a sense of justice and the
capacity for a conception of the good, id. at 331-40, and this provides the
basis for wider guarantees. Aspects of expression that are not connected to
politics will sometimes be linked to the capacity for a conception of the good.
See my discussion of the deliberative and expressive interests in
Joshua Cohen, Freedom of Expression, 22 Phil.
& Pub. Aff. 207 (1993).
n43. On the egalitarianism of the difference principle, see A Theory of Justice,
supra note 33, at 100-08. For criticisms, see G.A. Cohen, Incentives,
Inequality, and Community, in 13 The Tanner Lectures on Human Values 263
(Grethe B. Peterson ed., 1992).
n44. I say
"fundamentally" because of the role of the idea of pure procedural justice in the account of
the difference principle: a distribution satisfies the difference principle
just in case it is the outcome of playing the political-economic game according
to fair rules. So fairness in the rules translates into fairness in the result.
See A Theory of Justice, supra note 33, at 83-90. But the rules are not
standards of collective choice, and - more to the point - the case for the
fairness of the rules is made by reference to expected outcomes, in particular
the expectations under the rules for the least advantaged. Because the
justification appeals to expected outcomes, it is misleading to think of the
difference principle itself as procedural.
n45. Moreover, this reconciliation is supposed to proceed in a way that also
"good of community." I have in mind the ideal of a well-ordered society as a social union of social
unions with a consensus of principles of justice and the shared end of
upholding justice, as defined by those principles. On the connection between a
well-ordered society and the values of community, see A Theory of Justice,
supra note 33, at 395-587 (especially the summary at 577-87).
Rawls suggests in A Theory of Justice, supra note 33, at 120, 136.
n47. On these highest-order interests, see
Political Liberalism, supra note 1, at 29-35, 72-77, 299-315.
Political Liberalism, supra note 1, at 6-7
& n.6; see also
Joshua Cohen, Moral Pluralism and Political Consensus, in The Idea of Democracy 276, 277-78
(David Copp et al. eds, 1992).
n49. On the idea of a political conception of justice, see
Political Liberalism, supra note 1, at 11-15.
n50. The protection of these liberties is sufficient, not necessary. The
suppression of liberties does not eliminate (or perhaps even limit) moral
diversity; it drives such diversity underground.
n51. On loose-jointedness, see Stephen Hart, What Does the Lord Require?: How
American Christians Think About Economic Justice (1992).
Political Liberalism, supra note 1, at 13-22.
n53. Id. at 13. This familiarity is required for a political conception of justice.
n54. Hampshire, The New Twist, supra note 1, at 46.
n55. Innocence and Experience, supra note 5, at 90.
n56. Id. at 73. More precisely, Hampshire's thought is that the evil of the
conditions and practices noted in the text is not so much acknowledged as
presupposed by conventional moral reasoning:
"These are some of the constancies of human experience and feeling presupposed
as the background to moral judgments and arguments," id. at 90, in the way that the absence of spontaneous disappearances of
middle-sized objects is a constancy presupposed as background to perceptual
judgments. This is the core of Hampshire's response to Humean subjectivism
about value. Id. at 81-110.
n57. The fanaticism consists in requiring
"other persons to observe duties and obligations, and to develop specific
virtues, without providing them with any reasons for abandoning their previous
conceptions of the good" and for accepting an alternative. Innocence and Experience, supra note 5, at
n58. Hampshire, The New Twist, supra note 1, at 46.
n59. See five elements of substantive justice supra p. 597.
n60. On the first two categories, see Dahl, Democracy, supra note 19, at 163-75.
n61. I thank Oliver Gerstenberg for stressing this point in correspondence.
n62. See my discussion of the right of freedom of expression in Cohen, supra note
n63. On contestation and inclusion as two dimensions of democracy, see Robert A.
Dahl, Polyarchy 1-16 (1971).
n64. Of course, the procedures themselves may have a substantive rationale. When I
say here that a right, for example, has a procedural rationale, I mean only
that it is required for a procedure. I am not addressing the rationale for the
procedures, though see discussion on rationale infra pp. 612-15.
n65. Ely, supra note 16, at 105.
n66. Thus Ely says that the
"central function" of the First Amendment is to
"assure an open political dialogue and process." Id. at 112.
n67. See five elements of substantive justice supra p. 597.
n68. Dahl, Democracy, supra note 19, at 167.
"might just as well be disenfranchised." Ely, supra note 16, at 84.
n70. Hampshire mentions a requirement of
"reasonable access" as one element in the idea of procedural justice. Innocence and Experience,
supra note 5, at 141.
n71. Id. The passage I quote from in the text suggests one element in a set of
conditions on fair process.
San Antonio Ind. Sch. Dist. v. Rodriguez, 411 U.S. 1 (1973), rejects the claim that local financing of school districts violates a
Constitutional guarantee of the worth of citizenship rights. But the Court did
not reject the claim that the Constitution sets a threshold below which the
worth of those rights cannot fall. They argued instead that there was no
showing of a failure to come up to that threshold:
"no charge fairly could be made that the system fails to provide each child with
an opportunity to acquire the basic minimal skills necessary for the enjoyment
of the rights of speech and of full participation in the political process."
Id. at 37.
Rawls on the fair value of political liberties in
Political Liberalism, supra note 1, at 324-31, 356-63; Dworkin on the political leverage that would
be guaranteed by ensuring equality of resources, in Ronald Dworkin, What is
Equality? Part 4: Political Equality,
22 U.S.F. L. Rev. 1, 21-22 (1987).
n74. Innocence and Experience, supra note 5, at 141.
n75. Meiklejohn, supra note 40, at 27 (emphasis omitted).
McDuffy v. Secretary of the Exec. Office of Educ., 615 N.E.2d 516, 548 (1993).
n77. See, for example, the discussion of
"immunity rights" in Roberto M. Unger, False Necessity: Anti-Necessitarian Social Theory in the
Service of Radical Democracy 524-30 (1987).
n78. See Adam Przeworski, Democracy and the Market 36 (1991). Przeworski uses the
point in the text as part of an argument for a purely procedural constitution.
Constitutions that reduce the stakes of political battles
"define the scope of government and establish rules of competition, leaving
substantive outcomes open to the political interplay." Id.
It is not clear what Przeworski means by
"substantive," but it is hard to find an interpretation that fits the U.S. case, with
constitutional protections of religion and of different sorts of substantive
due process rights, or the entrenchment provisions in the postwar German
Rawls's discussion of the value of taking certain matters off the political agenda.
Political Liberalism, supra note 1, at 151, 157, 161.
"Faced with the fact of reasonable pluralism, a liberal view removes from the
political agenda the most divisive issues, serious contention about which must
undermine the bases of social cooperation." Id. at 157. To remove an issue from the political agenda is, of course, not to
forbid people from discussing it, writing books about it, or forming
organizations to promote a particular outlook on it. For this reason, Seyla
Benhabib's concerns about removing issues from the agenda seem misplaced. See
Seyla Benhabib, Models of Public Space, in Situating the Self: Gender,
Community and Postmodernism in Contemporary Ethics 89, 106-07 (1992).
n80. The virtues of a court as a device for protecting rights are a matter of
controversy. See Dahl, Democracy, supra note 19, at 187-91; Jonathan D. Casper,
The Supreme Court and National Policy-Making, 70 Am. Pol. Sci. Rev. 50 (1976);
Robert A. Dahl, Decision-Making in a Democracy: The Supreme Court as a National
6 J. Pub. L. 279 (1957).
n81. On consociationalism and the general distinction between consensual and
majoritarian democracies, see Arend Lijphart, Democracies: Patterns of
Majoritarian and Consensus Government in Twenty-One Countries (1984); Arend
Lijphart, The Southern European Examples of Democratization: Six Lessons for
Latin America, 25 Gov't
& Opposition 68 (1990).
n82. Innocence and Experience, supra note 5, at 140.
n83. That these are great evils is not controversial; indeed it is presupposed by
conventional moral reasoning. Innocence and Experience, supra note 5, 81-110.
n84. Hampshire appears to advance three arguments for the absolute duty of justice.
One is the argument from the great evils sketched in the text. Id. A second is
an argument from individuality, id. at 113-36, while a third treats fairness
itself as a fundamental value. Id. at 153-54. See
Joshua Cohen, Hampshire on Morality and Justice (Jan. 7, 1994) (unpublished manuscript, on
file with the author).
n85. Hampshire acknowledges that there are moral views - associated, for example,
with extreme forms of religious fundamentalism - that lack the resources for
supporting the great value of justice, even in
""all normal circumstances.'" Innocence and Experience, supra note 5, at 140, 144-46. Hampshire calls such
"evil moralities," distinguishing them both from views that accept the absolute duty of justice
and from the Nazi rejection of morality. Id. at 66-72, 76-77.
n86. Innocence and Experience also emphasizes the positive value of such diversity,
which is rooted in human imagination - standing alongside reason as a
"fundamental power of mind," Innocence and Experience, supra note 5, at 48, - and an associated
"drive to radical diversity and individuality." Id. at 33. See generally id. at 32-38, 41-48.
n87. Amy Gutmann appears to attribute such relativism to Hampshire in Amy Gutmann,
The Challenge of Multiculturalism in Political Ethics, 22 Phil.
& Pub. Aff. 171 (1993).
n88. Innocence and Experience, supra note 5, at 55-59, 63.
n89. Id. at 58.
n90. Hampshire does not explain why this is so. One natural explanation is that the
practices put such great burdens on some people that they cannot be justified
except if those burdens are understood to be naturally necessary.
n91. Hampshire's view of historical possibility is, I think, too simple. For an
interesting and more subtle view of the terrain, and the complexities it raises
for ethical thought, see Bernard Williams, Ethics and the Limits of Philosophy
n92. A rough sketch of the
"minimum decencies of procedural justice" is provided (implicitly) at Innocence and Experience, supra note 5, at 140-41.
n93. On the chaotic character of majority rule, see Richard D. McKelvey,
Intransitivities in Multidimensional Voting Models and Some Implications for
Agenda Control, 12 J. Econ. Theory 472 (1976); on bargaining from a Hobbesian
state of nature, see James M. Buchanan, The Limits of Liberty: Between Anarchy
and Leviathan (1975).
n94. Hampshire counts
"reasonable access" for
"interested parties" as one of the
"minimum decencies of procedural justice." Innocence and Experience, supra note 5, at 141.
n95. On the points in this paragraph, see Dahl, Democracy, supra note 19.
n96. Id. at 187.
n97. Id. at 172.
n98. Id. at 179.
n99. Here I follow the discussion in Adam Przeworski, Social Democracy as a
Historical Phenomenon, in Capitalism and Social Democracy 7 (1986).
"This fact makes it rational for them to move out of the narrower circle of
their own views and to develop political conceptions in terms of which they can
explain and justify their preferred policies to a wider public so as to put
together a majority."
Political Liberalism, supra note 1, at 165.
n101. What makes this a dilemma is that widening the appeal and redefining the
nature of politics threatens socialist parties with a loss of working class
support. For a discussion of the structure and dynamics of the resulting
political calculus, see Adam Przeworski
& John Sprague, Paper Stones: A History of Electoral Socialism (1986).
n102. Przeworski, supra note 99, at 28.
n103. The discussion of the expressive interest draws on Cohen, supra note 42, at
n104. I say
"perhaps" because expression often has nothing to do with communication. See C. Edwin
Baker, Human Liberty and Freedom of Speech 51-54 (1989).
n105. Hampshire's discussion of the fundamental value of individuality suggests a
rationale, within a variety of comprehensive moralities, for the importance of
the expressive interest. See Innocence and Experience, supra note 5, at 113-34.
Rawls on self-respect and the political liberties in
Political Liberalism, supra note 1, at 289-371; Charles R. Beitz, Political Equality: An Essay in
Democratic Theory 109-10, 123-213 (1989); Dworkin, supra note 73, at 4-5.
n107. Jean-Jacques Rousseau, On the Social Contract, bk. 3, ch. 14 (Judith R.
& Roger D. Masters ed., 1978) (1762).
Rawls urges in his discussion of the social bases of self-respect. See
Political Liberalism, supra note 1, at 318-20
n109. Sunstein, supra note 40, at 130.
N110. See supra text accompanying note 90.
n111. See A Theory of Justice, supra note 33, at 75-83. For discussion of the
connections between the difference principle and the interest in recognition,
Joshua Cohen, Democratic Equality, 99 Ethics 727, 736-43 (1989). 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, 10 Phil.
& Pub. Aff. 283 (1981).
n112. The idea of a freestanding conception of justice - and of what I am calling
the autonomy of political argument - lies at the heart of
Political Liberalism, supra note 1, at xvii, 11-14.
Prepared: January 24, 2003 - 5:02:29 PM
Edited and Updated, January 25, 2003
Philosophy 111A Page