Introduction 61
accept the hierarchal system as legitimate and endorse its conception
of justice. Under these conditions, incentives to become more liberal
are likely to cause resentment within the society of peoples and also
compromise the self-determination of nonliberal societies. See Law of
Peoples, p. 85.
85 Some object to Rawls’s duty of noninterference since it seems to imply a
duty not to come to the assistance of democratic liberation movements.
But the duty of noninterference only prohibits assisting democratic liberation
movements in decent hierarchical regimes, not in tyrannical and
other indecent regimes. A decent nonliberal society should be deemed
capable of the eventual self-imposition of democracy; otherwise, Rawls
implies, its members are not likely to sustain democratic rule that is
imposed upon them.
86 See Brian Barry, The Liberal Theory of Justice, pp. 128–33; Charles Beitz,
Political Theory and International Relations (Princeton, NJ: Princeton
University Press, 1979), pp. 127–69, and Thomas Pogge, “An Egalitarian
Law of Peoples,” Philosophy and Public Affairs 23, n. 3 (Summer 1994).
87 This is the “target” of the duty of assistance; unlike a global distribution
principle, it has a cut-off point and no longer applies once a society
achieves political autonomy and economic self-sufficiency. In view
of the independence and self-determination of a people who take responsibility
for their political culture, and for their rate of savings and
investment, Rawls sees little justification for a global distribution principle
(like the difference principle) that applies continuously without
end ( LP 117). Citing Japan, Rawls says once unjust political causes are
removed and a people achieves independence, its wealth is largely determined
by its political culture and industriousness, not its level of natural
resources.
88 In earlier drafts of LP, Rawls said that the contention that the justification
of boundaries must rest in the difference principle begs the issue
for its global application.
89 Because The Law of Peoples was not near completion when this volume
was planned in 1996, an essay on the law of peoples was not commis-
sioned.
90 On political liberalism, see also Sections II and III of Scanlon’s chapter
in this volume (Chapter 3) containing a discussion of overlapping consensus
and public reason, Gutmann’s chapter (Chapter 4) on liberalism
and democracy (especially Section 6 on public reason), and Section V of
Freeman’s chapter on overlapping consensus (Chapter 7).
thomas nagel
1 Rawls and Liberalism
I
“Liberalism” means different things to different people. The term is
currently used in Europe by the left to castigate the right for blind
faith in the value of an unfettered market economy and insufficient
attention to the importance of state action in realizing the values of
equality and social justice. (Sometimes this usage is marked by the
variants “neoliberalism” or “ultraliberalism.”) In the United States,
on the other hand, the term is used by the right to castigate the
left for unrealistic attachment to the values of social and economic
equality and the too ready use of government power to pursue those
ends at the cost of individual freedom and initiative. Thus, American
Republicans who condemn the Democrats as bleeding-heart liberals
are precisely the sort of people who are condemned as heartless liberals
by French Socialists.
Both of these radically opposed pejorative uses have some basis in
the broad tradition of liberalism as a group of political movements
and political ideas, sharing certain convictions and disagreeing about
others. It is a significant fact about our age that most political argument
in the Western world now goes on between different branches
of that tradition. Its great historical figures are Locke, Rousseau,
Constant, Kant, and Mill, and, in our century, its intellectual representatives
have included Dewey, Orwell, Hayek, Aron, Hart, Berlin,
and many others. With the recent spread of democracy, liberalism has
become politically important in countries throughout the world.
Rawls occupies a special place in this tradition. He has explored
and developed its philosophical foundations to an unprecedented
depth – and thereby transformed the subject of political theory in
62
Rawls and Liberalism 63
our time – and he has defended a distinctive, strongly egalitarian
view that is at odds with many others in the liberal camp, although
he sees it as following the basic ideas of liberalism to their logical
conclusion.
One indication of the importance of a political theory is the
vehemence with which it is attacked and the need its opponents
feel to explain their disagreements and situate themselves in relation
to it. Rawls has been attacked relentlessly, and from many directions,
because his theory of justice has the kind of real substance
that arouses strong disagreement. Though the style of presentation
is always accommodating rather than challenging, the views themselves
are highly controversial. They do not, for example, represent
the main stream of liberal opinion in the United States today.
In brief, what Rawls has done is to combine the very strong principles
of social and economic equality associated with European socialism
with the equally strong principles of pluralistic toleration
and personal freedom associated with American liberalism, and he
has done so in a theory that traces them to a common foundation.
The result is closer in spirit to European social democracy than to
any mainstream American political movement.
Rawls’s theory is the latest stage in a long evolution in the content
of liberalism that starts from a narrower notion, exemplified
by Locke, which focused on personal freedom and political equality.
The evolution has been due above all to recognition of the importance
of social and economic structures, equally with political and
legal institutions, in shaping people’s lives and a gradual acceptance
of social responsibility for their effects. When the same moral attention
was turned on these as had earlier been focused on strictly
political institutions and uses of political power, the result was an
expansion of the liberal social ideal and a broadened conception of
justice. Indeed, the use of the terms “just” and “unjust” to characterize
not only individual actions and laws but entire societies and
social or economic systems is a relatively recent manifestation of
this change of outlook. Rawls’s liberalism is the fullest realization
we have so far of this conception of the justice of a society taken as a
whole whereby all institutions that form part of the basic structure
of society have to be assessed by a common standard.
The original impulse of the liberal tradition, found in Locke and
Kant, is the idea of the moral sovereignty of each individual. It
64 thomas nagel
implies limitations on the ways in which the state can legitimately
restrict the liberty of individuals even though it must be granted
a monopoly of force in order to serve their collective interests and
preserve the peace among them. Freedom of religion, of speech, of
association, and of the conduct of private life and the use of private
property form the core of the protected liberties. Mill gave a different,
rule-utilitarian justification to these limits on the authority of the
state over the individual. They have remained central to liberalism
through continuing arguments both about their moral foundation
and about their proper scope and interpretation.
The other great moral impulse of liberalism, a hostility to the imposition
by the state of inequalities of status, overlaps at its point of
origin with the protection of liberty since both of them mean that
slavery, serfdom, and caste are ruled out. But opposition to inequality
extends gradually to more positive requirements such as equal
citizenship for all groups, universal suffrage, the right to hold office,
the abolition of hereditary political authority – in short, political and
legal equality as a general feature of public institutions.
What has led to the development of modern forms of egalitarian
liberalism of the kind that Rawls defends is the recognition that a
society may impose inequalities of status on its members in many
ways other than by making them legally explicit. The entire system
of social and economic institutions – partly made possible by laws,
such as the laws of contract and property, but really shaped by conventions
and patterns that are the sum of countless transactions and
choices by individuals acting in this framework over time – offers
very unequal life chances and opportunities to different persons, depending
on where they are situated in it by fate.
Consciousness of the hereditary inequalities of class led, of course,
to other political movements besides liberalism, but it expanded the
concerns of liberalism, through a natural extension of the opposition
to inequality, from inequality that was deliberately imposed to inequality
that was foreseeable and preventable but tolerated. This has
led to a great expansion of what liberalism can demand of the state
because it is not just a prohibition but a positive requirement – the
requirement that the state use its power to prevent certain severe
social inequalities from arising or from having their worst effects.
But the egalitarian impulse in liberalism, as opposed to movements
farther to the left, has always been strictly tied to the limits on
Rawls and Liberalism 65
state power imposed by each individual’s sovereignty over himself.
However much is required of the state in a positive direction to curb
the development of deep institutional and structural inequalities, it
may not violate the basic rights to liberty of individual citizens when
carrying out this charge. Putting these impulses together in a coherent
theory is not always easy, and the task has resulted in familiar
disagreements within the liberal camp.
Rawls’s theory is remarkable for the distance to which he has followed
both of these moral impulses and for the way he connects
them. Rawls interprets both the protection of pluralism and individual
rights and the promotion of socioeconomic equality as expressions
of a single value – that of equality in the relations between
people through their common political and social institutions. When
the basic structure of society deviates from this ideal of equality, we
have societally imposed unfairness, hence the name “justice as fairness.”
A society fails to treat some of its members as equals whether
it restricts their freedom of expression or permits them to grow up
in poverty.
It is the very strong interpretation he gives to the requirements of
justice for all of the basic institutions of society that makes Rawls’s
liberalism so controversial. It is very different from the liberalism of
Mill with its dominant insistence on limits to government action.
Mill was aware of the egalitarian appeal of socialism and responded
to it in his posthumously published “Chapters on Socialism.”1
His
doubts about the economic and psychological viability of a system of
that type were of the kind that have persisted and proven valid. But
the egalitarian impulse also persisted and eventually had its effect
on the development of the liberalism of the welfare state. How extensive
that effect will be remains uncertain; the question is very
much under current political debate in all broadly liberal regimes.
The other big difference from Mill is that Rawls’s account of the
individual rights central to liberalism is not instrumental. He does
not think they are good because of the results they will bring about;
he thinks they are good in themselves. Or rather, he holds that they
are principles of right and that the right is prior to the good. The
protection of certain mutual relations among free and equal persons,
giving each of them a kind of inviolability, is a condition of a just
society that cannot, in Rawls’s view, be explained by its tendency
to promote the general welfare. It is a basic, underived requirement.
66 thomas nagel
This noninstrumental conception of individual rights is also supported
by Rawls’s rejection of the utilitarian method of aggregating
advantages and disadvantages across persons and choosing the system
that maximizes the total. The importance for morality of the
distinctness of persons also accounts for the special form he gives to
the social contract as a foundation for political theory. But the details
would take us too far from the topic of this essay.
II
The relation of Rawls’s theory to other views will show up clearly
if we examine his two principles of justice in detail. We will then
see how his choices among alternatives express a specific moral position
and what other positions would have been expressed by other
choices. The two principles, in their latest formulation in Political
Liberalism, are as follows:
a. Each person has an equal claim to a fully adequate scheme
of equal basic rights and liberties, which scheme is compatible
with the same scheme for all; and in this scheme the
equal political liberties, and only those liberties, are to be
guaranteed their fair value.
b. Social and economic inequalities are to satisfy two conditions:
first, they are to be attached to positions and offices
open to all under conditions of fair equality of opportunity;
and second, they are to be to the greatest benefit of the least
advantaged members of society.2
The first principle (equal rights and liberties) has priority over the
second, and the first part of the second principle (fair equality of opportunity)
has priority over the second part (the difference principle).
Note that the first principle is a principle of strict equality, and
the second a principle of permissible inequality. The first applies
roughly to the constitutional structures and guarantees of the political
and legal systems, and the second to the operation of the social
and economic systems, particularly insofar as they can be affected
by tax policies and various approaches to social security, employment,
disability compensation, child support, education, medical
care, and so forth. The strict priority of individul rights and liberties
over the reduction of social and economic inequalities is the
Rawls and Liberalism 67
true core of liberalism, and it has attracted the scorn of the radical
left over a long period. This ideological battle is not over, as we see
from the denigration of “Western values” by the latest generation of
non-Western despots.
However, the issue of what to include in the required scheme
of rights and liberties marks an important division among liberals.
There are those who believe that the core rights are connected with
the protection of the democratic process and the prevention of political
oppression – such rights as freedom of speech, freedom of association,
due process of law, the right to vote and hold office, and
freedom of religion. On this view, purely personal and cultural liberties,
such as those involved in disputes over the legal enforcement
of sexual morality or the legality of abortion, do not have the same
status. On these issues Rawls’s interpretation of the scope of basic
rights tends to be broader for reasons having to do with the foundations
of those rights and with the ways in which a just society must
accept pluralism, reasons that will be discussed below.
On the other hand, there is one significant kind of right that Rawls
excludes from the full protection of the first principle, namely, property
rights. Those who give significant moral weight to property
rights – not just to the right to possess some personal property,
which Rawls includes, but significant rights of accumulation and
disposition of private property – belong to the libertarian branch of
liberalism. Even if strict libertarians are rare, the high valuation of
economic freedom is a significant element in the outlook of those
who retain a Lockean sympathy for the natural right of individuals
to enjoy the fruits of their labor and their gains from other uncoerced
economic transactions.
Rawls will have none of this. Entitlement to what one has earned
or otherwise legally acquired has a completely different status in his
theory from free speech, freedom of worship, or freedom to choose
one’s employment. Economically significant property rights are valued
not as an essential part of individual liberty but as indispensable
features of the economic system without which the reliable
expectations and security that are essential for long-term planning,
investment, production, and capital accumulation would not be possible.
Reliance on contract, salary agreements, the payment of dividends,
and so forth is economically essential, and it is only the justification
of the whole system that provides the moral support for
68 thomas nagel
an individual’s entitlement to what he earns or otherwise acquires
through the actions he and others take in accordance with its rules.
What he is entitled to is determined by the rules, and what the
rules should be, including the rules of taxation and redistribution,
is determined by which overall system would be most just in its
results, taken as a whole. In Rawls’s theory individual property rights
are the consequence, and not the foundation, of the justice of economic
institutions. In theories of a libertarian tendency the reverse
is the case.
This rejection of economic freedom as a value in itself is one
feature of Rawls’s view that has attracted opposition along with the
closely related rejection of individual desert as a fundamental political
value. For the purposes of political theory, at least, Rawls holds
that people deserve the product of their efforts only in the sense that
if they are entitled to it under the rules of a just system, then they
have a legitimate expectation that they will get it. This view is, I
think, more uncompromising than would be accepted even by most
of those who would describe themselves as liberals. There are certainly
those who would maintain that, even preinstitutionally, people
deserve what they gain by their own efforts and that this should
be allowed to have some effect on the form of a just economic system.
That might be expressed by some modification in the interpretation
of Rawls’s first principle, to admit a measure of economic freedom
as a protected right.
If we move now to the second principle, the first thing to observe
is that the inclusion of any such principle at all, limiting the inequalities
that can be permitted by a just state to arise through the
free choices of individuals acting under a regime of adequate and
fully protected individual rights and liberties, marks the difference
between laissez-faire liberalism and welfare state liberalism. The
second principle expresses the recognition that class stratification
and the resulting inequality of chances in life are social evils bearing
on the justice of a society.
To begin with the first part of the second principle: Equal opportunity
has come to be a central tenet of most liberal positions, but it
is open to two very different interpretations, negative and positive.
Negative equality of opportunity means the absence of barriers to
competition for places in the social and economic hierarchy, so that
anyone can rise to a position for which he is qualified. This is what
Rawls and Liberalism 69
Rawls calls the principle of “careers open to talents.” Positive equality
of opportunity, or what Rawls calls “fair equality of opportunity,”
requires more: It requires that everyone, whatever his starting place
in life, have the same opportunity to develop his natural talents to
the level of which he is capable so that he can compete for a position,
when the time comes, without handicaps that are due to a deprived
background. The second interpretation, enabling everyone to realize
his potentialities, demands much more state action than the first,
making sure the doors are open to anyone who qualifies.
Attachment to negative equality of opportunity – condemning the
deliberate exclusion of anyone on grounds of race, class, sex, or religion
from an equal chance to compete – is now nearly uncontro-
versial.3
And to some degree the value of fair or positive equality of
opportunity, or equality of chances, is more and more widely recognized.
The obligation of an affluent society to ensure access to education
through the university level to all who are willing and able
to benefit from it and some obligation to see that children receive
adequate nourishment and medical care, however poor their parents
may be, are accepted by most segments of the political spectrum in
broadly liberal societies. The disagreements are over the degree to
which inequalities of opportunity ought to be evened out.
Such inequalities cannot be eliminated entirely because differences
between families have a big effect on children that state action
cannot completely override. But there is room for disagreement
over how much has to be done. Some of that disagreement may be
due to differences of opinion about how powerful the effect of class
is on people’s options. Some parties claim that anyone can succeed
by hard work, whereas others point out how much more difficult it
is if you start at the bottom rather than at the top. But most of the
disagreement, I suspect, is due to a difference of moral focus. Those
who are inclined to regard the competitive advantages children get
from the luck of having been born to prosperous parents as unobjectionable
probably focus on the fact that they result from normal and
irreproachable family affection. Others, who think those advantages
and the corresponding disadvantages of those born poor are unfair,
probably focus on the fact that their recipients have done nothing to
deserve them.
Still, the debate over the proper form of equal opportunity is much
less divisive than that over whether a just society should go beyond
70 thomas nagel
this to strive for equality of results. That brings us finally to the
second part of the second principle – the difference principle – which
is Rawls’s most strikingly egalitarian requirement and one of his
most contested claims. It says, to repeat, that social and economic
inequalities “are to be to the greatest benefit of the least advantaged
members of society.” One can conceive of an even more egalitarian
principle, one which favored greater equality even if it would lower
everyone’s level of welfare, including that of the worst off. But this
does not hold much appeal outside the tradition of utopian socialism
and is in any case probably the reflection of something else – the idea
that strict equality of possessions would promote a universal level of
self-esteem and mutual respect that is impossible in a socially and
economically stratified society. That is the appeal of the perennial
fantasy of the abolition of all hierarchy. But Rawls’s difference principle
is still very egalitarian, and it can be contrasted with several alternatives
that command support within the spectrum of liberal views.
First there is the view that the only equality required for justice
is equality of opportunity and that since the inequalities that arise
under a regime of equal opportunity are the result of what people
make of their opportunities, they are not unjust. Somewhat more egalitarian
is the view that certain forms of misfortune, including disability,
serious illness, and particularly low earning capacity due to
lack of skills or overwhelming parental responsibilities, should not
be allowed to render their victims helpless and destitute. The provision
of some kind of social safety net is widely favored to deal with
such cases, although there is disagreement over how high the net
should be – what level of social minimum it should guarantee. This
view is perhaps best interpreted not as a fundamentally egalitarian
one but rather as the consequence of something else: the judgment
that certain absolute forms of deprivation are particularly bad and
that no decent society should tolerate them if it has the resources to
prevent them.
Another view that has egalitarian consequences, although it is
not fundamentally egalitarian, is utilitarianism, the position that
the maximization of total welfare should be a social goal. Destitution
seriously brings down the total, and the diminishing marginal
utility of resources means that transferring some of them from the
rich to the poor, if it can be done without too much loss, will increase
total welfare. It seems likely that most support for moderate policies
Rawls and Liberalism 71
of assistance to the disadvantaged is due to moral positions like
these rather than to the much more deeply rooted egalitarianism
that Rawls defends.
Rawls’s difference principle is based on the intuitively appealing
moral judgment that all inequalities in life prospects dealt out to
people by the basic structure of society and for which they are not
responsible are prima facie unfair; these inequalities can only be justified
if the institutions that make up that structure are the most
effective available in achieving an egalitarian purpose – that of making
the worst-off group in the society as well off as possible. This is
an egalitarian aim because it blocks the pursuit of further equality
only if that would make everyone worse off.
This may be a radical position, but it should be kept in mind that
it applies only to deep structural inequalities that affect statistically
large numbers of people in the different social categories. It does not
apply to the countless inequalities among individuals that will inevitably
arise as people make choices and interact, and succeed or fail
in their efforts, in the context of any socioeconomic structure, however
just. If the broad structure of society satisfies the principles of
justice in its large-scale statistical effects on the life prospects of different
groups, then, according to Rawls, any individual inequalities
that emerge from its operation will be ipso facto just. That is what
he means by calling it a system of pure procedural justice: The broad
design of the system confers legitimacy on the specific outcomes,
whatever they are.
Nevertheless, the difference principle means that the broad design
of the system is supposed to be evaluated by its success in eliminating
those inequalities that are not needed to provide maximum benefit
to the worst off. And this imperative depends on the moral claim
that it is unfair if people suffer or benefit differentially because of
differences between them that are not their fault. A society that
does not try to reduce such differentials is not just, and that applies
whether the differences in question are racial, sexual, or religious or
disparities in the fortunes of birth, such as being born rich or poor,
or being born with or without unusual natural abilities.
It is this last point, the unfairness of society’s systematically rewarding
or penalizing people on the basis of their draw in the natural
or genetic lottery, that underpins the difference principle. Even under
ideal conditions of fair equality of opportunity, such inequalities will
72 thomas nagel
arise from the normal operation of a competitive market economy in
which there is bidding for scarce productive skills. According Rawls
those inequalities are unjust unless supplemental policies ensure
that the system works to the maximum benefit of the worst off.
People do not deserve their place in the natural lottery any more
than they deserve their birthplace in the class structure, and they
therefore do not automatically deserve what “naturally” flows from
either of those differences.
One other point about the second principle deserves attention: the
priority of the first of its conditions over the second. Rawls holds that
fair equality of opportunity may not be sacrificed even if this would
benefit the worst-off group in a society. It may be difficult to imagine
how that might be so, but I mentioned earlier in this section the deviation
from equality of opportunity represented by affirmative action,
and it is perhaps possible that, even in the absence of the historical
legacy of slavery or a caste system, someone might favor an ongoing
program of preference in assignment of desirable positions to the
less talented, or perhaps some randomization of assignment, in order
to prevent the development of a hereditary meritocracy. That kind
of reversal of priority between equality of opportunity and equality
of results would represent a more radically egalitarian position than
Rawls’s, and also one that was in a sense more anti-individualistic.
This brief survey of the alternatives shows that in putting forward
his two principles of justice, Rawls has not only expressed a distinctive
position but provided a framework for identifying the morally
crucial differences among a whole range of views on the main questions
of social justice. I now want to go more deeply into the justifications
for the most controversial features of his view – its pluralism
and its egalitarianism.
III
An important element in Rawls’s conception of liberty is the requirement
that a just state refrain, so far as possible, from trying to impose
on its members a single conception of the ends and meaning of life.
This is most straightforward in the requirement of freedom of religion,
and Rawls assigns great importance to the historical descent
of ideas of toleration from the seventeenth-century wars of religion
and their aftermath. But he applies the principle much more widely
Rawls and Liberalism 73
to cover all deep differences in fundamental conceptions of the good.
Toward these, he believes a just society should adopt an attitude of
toleration and the expectation of pluralism and should leave people
free to pursue their ultimate aims provided they do not interfere with
the other requirements of justice.
What this position opposes, in particular, is one or another form of
perfectionism based on commitment to a particular contested idea
of the ends of life and insistence that it is the proper role of a political
community to guide its members in that direction by coercion,
education, the exclusion of other options, and control of the cultural
environment. Rawls opposes perfectionism not merely because the
contest for religious or cultural hegemony has divisive results and is
potentially dangerous for all parties. That would be to accept pluralism
and toleration as a mere modus vivendi, necessary for practical
reasons though falling short of the ideal. Rawls believes, on the contrary,
that pluralism and toleration with regard to ultimate ends are
conditions of mutual respect between citizens that our sense of justice
should lead us to value intrinsically and not instrumentally. In
the original position, this ideal receives formal expression through
the fact that parties to the hypothetical contract are supposed not to
know their own full conception of the good – so they have to choose
principles of justice based on a thin, purely formal conception that
they know would be consistent with any of the thicker conceptions
that might be their actual one. This feature of the veil of ignorance,
like not knowing one’s race or class background, is required because
Rawls holds that equal treatment by the social and political systems
of those with different comprehensive values is an important form
of fairness.
The distinction between comprehensive values and more narrowly
political values is discussed extensively in Political Liberalism,
and Rawls suggests that in A Theory of Justice he failed
adequately to attend to this difference.4
This is a rather subtle matter.
I myself think that the aim of making the theory of justice independent
of any particular comprehensive view was already implicitly
present in the earlier book, though the later discussion is very important
in working out how Rawls believes this can coherently be
accomplished. In any case, the questions whether it is possible, and
if so whether it is desirable, have generated a great deal of attention.
Rawls himself points to others in the liberal tradition, such as Kant
74 thomas nagel
and Mill, who take it for granted that political liberalism should be
derived from a comprehensive moral conception. That outlook still
has many adherents. And since Rawls has raised the issue, a number
of skeptics have argued that it is impossible to ground a political
theory of justice on a much narrower base, as he wishes to do – that
the kind of neutrality or abstinence that he requires of us when thinking
about justice is unavailable and incapable of sustaining the moral
commitment to principles of tolerance and antiperfectionism.5
This corresponds to a heated dispute that arises again and again
in public debate: Do the typical liberal demands for tolerance and
individual liberty with respect to religion, sexual conduct, pornography,
abortion, assisted suicide, and so forth really depend on the
requirement of state impartiality toward deep and contested personal
convictions, or are those demands in reality based on the quite specific
contested convictions of those very liberals, convictions which
they think it politically inadvisable to invoke directly – religious
skepticism, sexual libertinism, and moral endorsement of abortion
and assisted suicide? Alternatively, the charge may be that the true
basis of all liberal positions is a comprehensive belief that the best
thing for each person is to live his life in accordance with his own
autonomous choices, whatever they are, and that that is what a just
society should make possible so far as it can be managed for people
with widely varying preferences and commitments. This is an important
issue both theoretically and substantively; the appropriate
form of liberal toleration turns on it.
It is true that with respect to any issue of individual rights, such as
homosexuality, two very different arguments can be offered on the
side of liberty. The first is that there is nothing wrong with homosexuality,
so it should not be prohibited. The second is that, whether
or not homosexuality is morally wrong, sex is one of those highly
personal matters that should not be controlled by a society on the
basis of the convictions of a majority of its members. It is also true
that many of the people who would be willing to offer the second
argument would also endorse the first, and perhaps not many who
would reject the first would be persuaded by the second. Still, there
is an important point to the appeal by some liberals, in the style of
Rawls, to the second, higher-order argument, which belongs specifically
to political rather than overall moral theory. Whether or not
it actually commands wide acceptance, the second-order argument
Rawls and Liberalism 75
tries to appeal to a value which all members of a pluralistic liberal
society could reasonably accept even if they disagreed fundamentally
in their beliefs about sexual morality. It is not the overriding
value of individual personal autonomy, which may be rejected by
many religious and other comprehensive views. It is the value of
mutual respect, which limits the grounds on which we may call on
the collective power of the state to force those who do not share our
convictions to submit to the will of the majority.
All government, all society, requires that the state must have such
power; the issue concerns only its extent and the admissible grounds
of its exercise. The way it defines those limits is one of the most
important features of any liberal position – what makes it a liberal
theory of democracy rather than mere majoritarianism. As we know
from the case of Mill, strict limits on both the extent and the admissible
direct grounds for the exercise of state power can be defended
directly by appeal to the comprehensive value of happiness and individual
human flourishing without relying on any principle of secondorder
impartiality among comprehensive views. On Liberty is a powerful
rule-utilitarian defense of liberal principles.
But Rawls wants something else – something that is in a way
more difficult and perhaps less likely to persuade in real political
argument. He wants a justification for liberty and pluralism that
does not rely on the individualistic system of values so many liberals
share. Political liberalism should be compatible with religious
orthodoxy. Rawls wants this because, when it comes to constitutional
essentials, it is insufficiently respectful toward those many
members of a liberally governed society who do not share those comprehensively
individualistic values to justify the institutions under
which we all must live, and the rights which those institutions guarantee,
by reference to grounds those individuals cannot reasonably
be expected to accept. The reach of a justification for constitutional
guarantees of individual freedom must be wider than that even if this
means its grip will be more precarious.
Rawls identifies the type of argument he has in mind in his extensive
discussions of what he calls “public reason” and its relation to
the fact of reasonable pluralism. These concepts are very important
in Political Liberalism and receive their most developed treatment
in a still later essay of 1997.6
The greatest difficulty in defining such
a view is to distinguish between those conflicts of value that belong
76 thomas nagel
within the domain of public reason and those that do not. Disagreements
outside of the public domain, religious disagreements being
the clearest example, should so far as possible be avoided when justifying
the design of basic social and political institutions. But disagreements
within the domain of public reason can be just as fundamental,
yet Rawls believes that those who hold the balance of
political power need not hesitate to exercise it on the basis of their
views on such questions or to impose the result on those with opposite
views. This happens all the time in political debate over issues of
war and peace, economic policy, taxation, welfare, or environmental
protection, for example. So what is the difference?
Rawls emphasizes that public reason is not to be thought of as an
effective decision procedure, guaranteed to produce agreement, but
rather as a special kind of disagreement, argument, and counterargument,
which tries to use mutually recognized methods of evaluation
and evidence, whether these produce consensus or not. Even if we
are not convinced by an opponent’s arguments about distributive justice,
for example, we can recognize them as offering grounds that he
thinks would be reasonable for us to accept, simply in virtue of the
reasoning capacity that we all share. The same cannot be said for
appeals to faith or revelation.
Whether an argument constitutes an appeal to public reason is
itself likely to be a contested issue (think of the question of the permissibility
of abortion). But the concept of public reason is not put
forward by Rawls as a mechanical test for the admissibility of arguments
but rather as a characterization of what we should be looking
for in an admissible ground for the design of basic institutions. In
applying the concept there will be higher-order disagreements, just
as there are conflicting arguments within the domain of public reason.
But the sense of justice should lead us to try, in good faith,
to offer to our fellow citizens grounds for the exercise of collective
power that we believe they, from their point of view as fellow reasoners,
have reason to accept – even if they do not actually do so. To
invoke only our private convictions is, according to Rawls, a violation
of the requirement of reciprocity that applies to members of a
just society.
In addition to these problems of definition there is the big problem
of justification. How can we put aside some of our deepest convictions
– convictions about the ultimate ends of life – in deciding
Rawls and Liberalism 77
how our society should be arranged? It can seem like a betrayal of our
values to deliberately refuse, if we have the power, to put everyone
on what we believe to be the true religious path to salvation, or
on the contrary the true secular path of individual autonomy and
self-realization, through the design of the political, social, and educational
systems. To base political values on something less than
our most comprehensive transcendent values can seem both morally
wrong and psychologically incoherent. For how can these narrower
political values have the leverage to hold in check transcendent religious
values, for example – particularly when the latter are concerned
not just with my own interests but with what I take to be the most
important interests of everyone, and therefore of my fellow citizens,
whatever their own convictions may be? The same question arises
about individualistic secular values, which would seem to justify
political opposition to orthodox religion.
This is a difficult question of moral theory, lying at the foundation
of the idea of individual rights and therefore at the foundation of a liberalism
based on rights. The central issue is whether a requirement
of mutual respect, operating in the context of the exercise of collective
power over the individual members of a society, is strong enough
to hold in check not only the unlimited pursuit of the self-interest
of the majority at the expense of the minority but also the unlimited
pursuit of the ostensibly transcendent values of the majority against
the will of the minority who do not share them. Skeptics answer that
to base our principles of political right and wrong on something less
than our full system of values is to accord those values only superficial
importance by comparison with an abstract, almost contentless
universality.
Rawls’s attempt to answer the question by grounding liberal toleration
and freedom on principles of right that are prior to conceptions
of the good is one of his most significant contributions. The difficulty
of the task is considerable, and the suspicion remains on the
part of many critics that such views are a kind of liberal camouflage
for much more partisan arguments – that the proposed ecumenical
appeal of liberalism is hollow. Some of these critics are themselves
liberals who believe it is better to defend liberal ideals by appealing
to an explicitly liberal conception of the human good.
But I believe Rawls’s alternative is a moral idea of the first importance
and that it represents a political ideal worth striving for.
78 thomas nagel
Even if it is much harder to explain and defend than a liberalism
based straightforwardly on individualistic and utilitarian values, a
Rawlsian political liberalism that could be justified even to those
of orthodox religious belief who do not share those values would be
preferable as a ground for determining the legitimacy of the exercise
of power by a state over all its citizens. Rawls has tried to describe a
form of liberalism that can claim the allegiance not only of secular
individualists and not only as a modus vivendi or second best. I believe
he has identified a source of moral conviction and motivation
that does not depend on religious skepticism or an ethic of individual
autonomy, and that has an important role to play in the justification
of liberal democratic institutions.
IV
The other great source of controversy in Rawls’s moral outlook is
his strong egalitarianism, exemplified by the difference principle.
Not only the principle itself but various of the claims offered in its
support have aroused substantial opposition. He qualifies its status
somewhat in Political Liberalism, saying that it is part of basic
justice but not a constitutional essential and that it is much more
difficult to ascertain whether it has been realized than is true of the
basic liberties; but it remains a very important part of his overall
view.
Rawls defends the difference principle most fully in Chapter 2 of
A Theory of Justice, arguing that it follows intuitively by a kind of
analogy from other principles of equality that are less controversial.
His main point is that we cannot be content with equality of opportunity.
Even the principle of negative equality of opportunity, which
excludes deliberate discrimination, depends on the belief that the
social system should not assign benefits or disadvantages solely on
the basis of differences between people for which they are not responsible
and which they have done nothing to deserve. To exclude
qualified candidates from a profession because of their race or sex is
to penalize them on grounds that are arbitrary in the worst sense,
and a society that permits such a thing is unjust.
This is only a first step, however, because people are no more
responsible for the socioeconomic status of the family into which
they are born than they are for their race or sex. Yet a system which
Rawls and Liberalism 79
guarantees only negative equality of opportunity permits class inequalities
to develop and accumulate without doing anything to
counteract the enormous differences they generate in the opportunities
for individuals to acquire the training and background needed
to develop their abilities and so to compete for formally open positions.
Negative equality of opportunity is therefore not full equality
of opportunity. It must be supplemented by positive provision of the
resources that will permit each potential competitor to develop his
natural abilities and therefore to be in a position to take advantage
of his opportunities. That is what Rawls means by fair equality of
opportunity.
The same reasoning leads him further. Even under a regime of
fair equality of opportunity, undeserved inequalities would continue
to arise. Fair equality of opportunity, to the extent that it can be
realized, guarantees only that persons of equal natural ability will
have roughly equal chances to prosper. But people are not equal in
natural ability, and their natural or genetic differences will continue
to affect the benefits they gain from interaction with the social and
economic order. Yet this too is morally arbitrary, for people are no
more responsible for their genetic endowment than for their race or
the economic status of their parents. Consequently a just society
will counter these undeserved differences in benefit to the extent
that it can do so without hurting the very people whose arbitrary
penalization it is most concerned to rectify, namely, those who come
in last in the socioeconomic race. Hence, the difference principle.
Despite the persuasiveness of these analogies, not everyone is convinced
that there is anything unfair about people’s benefitting differentially
from the employment of their own natural abilities even
though they have done nothing to deserve those abilities. Even if
they have done nothing to deserve it, their genetic makeup is part of
their identity, and it can seem like an assault on the independence
of persons to say that they have no right to the benefits which flow
from that identity, except insofar as this also benefits others. Such
reactions have seized on Rawls’s striking remark that “the difference
principle represents, in effect, an agreement to regard the distribution
of natural talents as a common asset and to share in the benefits
of this distribution whatever it turns out to be.”7
The issue identifies a fundamental cleavage in the liberal tradition
between those who identify justice with the fight against any kind of
80 thomas nagel
undeserved inequalities that the design of the social system can ameliorate
and those who believe the scope of justice is narrower – that
society is exempt from responsibility for certain forms of “natural”
difference, even if they are in a nonpolitical sense unfair. In this more
limited conception, a just society should provide a framework, with
fair equality of opportunity and a decent social minimum, in which
people can rise by their own efforts to the level to which their natural
abilities and efforts are able to take them.
The moral significance of the choice between this vision and
Rawls’s is quite difficult to characterize. Both are interpretations of
the vague idea of relations of mutual respect and cooperation among
the separate, autonomous individuals that make up a society. We do
not own one another and we want to interact on equal or reciprocal
terms in some sense. But in Rawls’s conception, we should not want
the collectively sustained system of which we are all equally members
to allow us to reap benefits on the basis of lucky accidents of
fate which we do not deserve, at the expense of others less fortunate
who also do not deserve their fate. The fact that one’s draw in the
natural lottery is undeserved communicates itself morally to what
flows from it through the operation of the economy. As Rawls says
in another memorable formulation, “In justice as fairness, men agree
to share one another’s fate.”8
The opposite view is that we retain more independence than this
of the claims of others when we enter a society and do not even
metaphorically hand ourselves over to it. Just as basic personal freedom
remains protected by liberal equality, so does the right to benefit
from one’s efforts and one’s talents. Our responsibility for one another,
as fellow members of a society, is substantial but nevertheless
definitely limited by our continued independence.
The moral key to Rawls’s more expansive position is in the idea
that, because of the essential role of the state, the law, and the conventions
of property in making possible the extraordinary productivity
and accumulations of a modern economy, we bear collective
responsibility for the general shape of what results from the sum
of individual choices within that framework. We are therefore responsible
for large-scale inequalities that would not have arisen in
an alternative framework, and if they are morally arbitrary, we have
reason to want to alter the system to reduce them. There is simply
something repellent about a joint enterprise in which rewards are
Rawls and Liberalism 81
apportioned in accordance with genetic endowment – unless there
is some further instrumental justification for this apportionment, as
there is when an inequality satisfies the difference principle.
Among those who would agree with Rawls in accepting society’s
responsibility for all outcomes that it permits, and not only for those
that it produces deliberately, there is still room for disagreement with
the strong egalitarianism of the difference principle. The strict priority
given to improvements in the situation of the worst off, in
preference even to greater individual and aggregate improvements to
the situation of those better off, seems unreasonable – particularly
to those drawn to utilitarianism. Utilitarians might agree that social
inequalities require justification but that they may be justified because
they contribute to the general welfare, not just to the benefit
of the worst off.
Even those who would admit some priority to the needs of the
worse off over the better off – after all, the better off already have
what the worse off need – may think the difference principle too
absolute. It seems to devaluate the interests of the middle class unreasonably
to say that a socioeconomic order will always be more just
if it sacrifices them to the interests of the lower class. Such doubts
are also voiced at the level of the hypothetical contract: it is often
questioned whether the parties in the original position would be rational
to adopt the maximin strategy of choice, which leads to the
choice of the difference principle, as a way of ensuring that the worst
possible outcome will be as good as possible. Rawls’s strong egalitarianism
displays an exceptionally strong aversion to the generation
by social institutions of what he regards as undeserved differences.
In addition to the familiar opposition from his right on the grounds
that the difference principle is too egalitarian, there is an interesting
criticism from the left to the effect that Rawls is too ready to countenance
economic inequalities under the difference principle even if
they are the result of acquisitive motives on the part of members of
the society – motives diametrically opposed to the ideal of equality.9
The point is that, in a market economy, it is assumed that inequalities
in income and wealth will arise as a result of the wage and
profit incentives that drive economic activity. The claim that these
inequalities are necessary for the benefit of the worst off depends
on the assumption that individuals will not be adequately motivated
in their roles as participants in the economy without personal
82 thomas nagel
incentives that appeal to the purely individualistic desire to accumulate
resources for the discretionary use of oneself and one’s family.
But the question then arises, Can a society be truly just if there is
such a gulf between the egalitarianism that determines the design
of its institutions and the individualism that motivates its members
when they act in the context of those institutions?
The fact that Rawls accepts this division is a mark of his unqualified
attachment to the liberal tradition despite his strong institutional
egalitarianism. Political theory is one thing; personal morality
is another. Justice is conceived as a specifically political virtue, leaving
individuals free to live their lives in pursuit of their own aims
and commitments, be these hedonistic or puritanical, libertine or
devoutly religious. The special demands of equal respect for the interests
of all that justice imposes apply to the sphere of collectively
sustained institutions, not to personal life. So liberalism involves
a division of the moral territory and leaves individuals free to instantiate
a great plurality of forms of life, some of them highly selfabsorbed,
so long as they are compatible with a just basic structure
of cooperation.
V
This division between the personal and the political, and the assignment
of justice firmly to the political category, has come to prominence
in Rawls’s writings after A Theory of Justice, culminating in
Political Liberalism. He has emphasized that justice as fairness is a
freestanding political conception partly in response to criticisms of
Theory alleging that it relied on a conception of the self as an autonomous,
unconstrained subject of choice whose good consisted in
forming its own preferences and pursuing their satisfaction, whatever
they were. While most of those criticisms depended on misinterpretation,
including the gross misinterpretation of attributing
to Rawls the view that actual persons were like the stripped-down
characters in the original position, the criticisms also threw into relief
the difficult question of the coherence of a position that makes
political values independent of comprehensive values and capable of
dominating them in the political sphere, even if they are concerned
with the most important things in life such as salvation and self-
realization.
Rawls and Liberalism 83
One of the important points Rawls has made is that the alternative,
of deriving the political order from a particular comprehensive
value system, is often supported by nostalgia for a communitarian
past that never existed, in which all the members of a society were
united in devotion to their common conception of the good: the
Christian world of the middle ages – in fantasy. Rawls points out
that the maintenance of orthodoxy of that kind has always required
oppression because harmonious agreement over fundamental values
does not maintain itself naturally. The Inquisition was no accident;
the persecution of heretics and apostates is an inevitable part
of the attempt to maintain comprehensive unity and to prevent the
outbreak of conspicuous dissent. Pluralism, on the contrary, is the
natural result of a regime of basic individual rights and freedoms.
It follows that support for the core of liberalism, the guarantee of
basic rights, must be compatible with pluralism. Now admittedly, it
would be possible to argue for such rights purely instrumentally on
the ground that each party in the plurality of comprehensive views
has more to lose from the danger of becoming an oppressed minority
than it has to gain from the chance of being the controlling majority.
Then liberalism would be adopted as a modus vivendi among parties
each of which would prefer, if only it were possible, to impose its
comprehensive conception on the others. But Rawls favors the more
demanding standard that the equal respect for others expressed by
recognition of their rights should be valued for itself and that this
should be the highest value in the sphere of political institutions,
although not in the conduct of personal life.
The importance of liberal rights depends precisely on the fact that
there are things people care about more than the political order but
with respect to which a plurality of beliefs and commitments is inevitable.
The only way to live together on terms of equality with others
with whom we disagree fundamentally about the ends of life, in
a framework that imposes its basic shape on all our lives, is to adopt
principles for the evaluation of the framework that can be accepted
by as many of us as possible. Their basis must therefore be compatible
with a wide range of reasonable but mutually incompatible
comprehensive views.
That means that some comprehensive views are not reasonable
because they do not permit their own subordination to the requirement
of reciprocity, that is, to the aim of seeking a collectively
84 thomas nagel
acceptable basis of cooperation. Fanatical movements which subordinate
the individual to the community depend on comprehensive
values that are unreasonable in this sense. But Rawls believes that
a wide range of views, forming the plurality typical of a free society,
are reasonable and can support the common institutional framework.
That is what he means by an “overlapping consensus.” Overlapping
consensus does not mean the derivability of common principles of
justice from all the comprehensive views in the pluralistic bouquet
but rather the compatibility of each of those comprehensive views
with a free-standing political conception that will permit them all
to coexist.
There are many forms of liberalism, and there will continue to
be. And while the liberal tradition is now in the ascendant politically
in economically advanced countries and making considerable
inroads elsewhere, it continues to be the object of attack not only
from apologists for tyranny and fanaticism but from many others
who cannot accept its severe restraints on the legitimate use of government
power – its insistence that the end, however worthy, does
not justify the means. Rawls’s advocacy of a specific liberal position
and his deep exploration of its foundations in ethical and political
theory constitute an enduring contribution to this tradition.
endnotes
1 In S. Collini, ed., On Liberty and Other Writings (Cambridge, UK: Cambridge
University Press, 1989).
2 Rawls, Political Liberalism (New York: Columbia University Press,
1993), pp. 5–6 (cited as “PL” with page numbers).
3 It has been breached by the policy of affirmative action, which is of
course highly controversial in liberal societies, but is probably best understood
in Rawlsian terms as an attempt at corrective justice – an
attempt to rectify the residual consequences of a particularly gross
violation in the past of the first principle of equal rights and liberties.
Affirmative action therefore does not form a part of what Rawls
would call “strict compliance theory” or ideal theory, which is what the
two principles of justice are supposed to describe. See A Theory of Justice
(Cambridge, MA: Harvard University Press, 1971/revised ed. 1999),
pp. 8–9/7–8 rev.
4 See the introduction to PL, pp. xvii–xx, and pp. xix–xxii in PL, 1996
paperback edition.
Rawls and Liberalism 85
5 See, for example Joseph Raz, “Facing Diversity: The Case of Epistemic
Abstinence,” Philosophy & Public Affairs 19 (1990).
6 “The Idea of Public Reason Revisited,” in John Rawls, Collected Papers,
ed. Samuel Freeman (Cambridge, MA: Harvard University Press, 1999),
573–615.
7 TJ, p. 101/87 rev., sentence rewritten.
8 TJ, p. 102/88 rev., sentence eliminated.
9 See G.A. Cohen, “Incentives, Inequality, and Community,” in The
Tanner Lectures on Human Values, vol. 13 (1992), Grethe B. Peterson,
ed.; “Where the Action Is: On the Site of Distributive Justice,” Philosophy
and Public Affairs 26 (Winter, 1997): 3–30; If You’re an Egalitarian,
How Come You’re So Rich? (Cambridge MA: Harvard University Press,
2000), Lectures 8–9.