The Journal of Political Philosophy: Volume 9, Number 1, 2001, pp.

118

Equality of Whom?
Social Groups and Judgments of Injustice*
IRIS MARION YOUNG
Political Science, University of Chicago

ECENT philosophical debates about normative ideals of equality have

focused on questions of what we should be aiming at when we wish to make
people more equal. Should norms of equality be conceived in terms of welfare,
for example, which entails reference to subjective valuations, or should goals of
equality be directed at objective resources? Are concepts of the distribution of
resources adequate for comparing objective well-being or do we need to
conceptualize and compare what persons are capable of doing with these
resources? Recent debates have shown that the apparently simple question of
what we should compare when we make judgments of equality and inequality
harbors a nest of conceptual and practical tensions. Contributors to the debate
have greatly claried these issues, if not solved them.1
Little attention has been paid to another side of this question, however, namely
whom we are discussing when we compare people's situation with regard to any
or all of these targets of equality. Theorists usually assume that the units we
should be comparing when we make judgments of inequality are individuals.
Many assessments of inequalities and claims for redress on grounds that an
inequality is unjust, however, compare groups of individuals according to one or
more measures of equality. People commonly claim that women lack equality
with men, Blacks with whites, old people with younger people, children of
working-class parents with children of middle-class parents, and so on.
Governments, research institutes, and other organizations apparently legitimate
such group-conscious judgments about equality by disaggregating general
welfare measures according to gender, ethnicity, race, religion, caste, age,
occupation or region. This article asks whether and how such group-conscious
practices of assessing inequality are justied.
First I review both political and philosophical challenges to theories and
practices which assess inequality in terms of social groups. If equality is
*I am grateful to David Alexander, Richard Arneson, David Gauthier, Robert Goodin, Charles
Larmore, Thomas Ross and Larry Temkin for comments on earlier versions of this article. Thanks to
Michael Reinhart for research assistance.
1
For reviews of the debates, see: Richard Arneson, ``Equality,'' Robert Goodin and Philip Pettit,
eds, A Companion to Political Philosophy (Oxford: Blackwell, 1993); John Roemer, Theories of
Distributive Justice (Cambridge, Mass.: Harvard University Press, 1996); see also Andrew Levine's
useful discussion in Rethinking Liberal Equality (Ithaca, N.Y.: Cornell University Press, 1998), ch. 2.
# Blackwell Publishers, 2001, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street,
Malden, MA 02148, USA.

IRIS MARION YOUNG

important, from this point of view, it must be assessed strictly in terms of the
comparison of individuals. I argue against this position that assessment of
inequality solely by comparing the situation of individuals provides little or no
basis for making claims about social justice. The main reason moral and political
philosophers care about equality, however, is to contribute to assessments of
social justice and proposals for promoting greater justice. A large class of issues
of social justice, and those that concern claims that inequalities are unjust in
particular, concern evaluation of institutional relations and processes of the
society. Evaluating inequality in terms of social groups enables us to claim that
some inequalities are unjust, I argue, because such group-based comparison helps
reveal important aspects of institutional relations and processes.
Specically, identifying inequalities according to group categories helps
identify structural inequalities. Though many philosophers, social theorists and
political actors profess an interest in structural inequality, as distinct from
patterns or relations of inequality that are transient, accidental or more socially
supercial, many use the term without explaining its meaning. This article
theorizes structural inequality as a set of reproduced social processes that
reinforce one another to enable or constrain individual actions in many ways.
What we refer to by group differentiations of gender, race, class, age, and so on,
in the context of evaluating inequalities as unjust, are structural social relations
that tend to privilege some more than others. Identifying patterned inequalities
on measures of well-being among these groups is thus only the beginning, but an
important beginning, of identication of these forms of basic and persisting
injustice. We need also to be able to give an account of how social processes
produce and reproduce these patterns.
The article concludes by asking what this argument implies for social policy. It
certainly implies that public and private bodies charged with collecting and
analyzing data about the status and well-being of society's members, and their
relative well-being, should continue to organize such data in terms of social
groups and compare the standing of groups to one another. The particular
argument of this article, however, in itself implies nothing about steps which
policy ought to take to rectify unjust inequalities discovered by means of such
analyses. Some remedies practiced and proposed for structural injustices involve
afrmative targeting of the distribution of resources or affording specic
opportunities to groups or members of groups. While such a group-conscious
approach to remedy for structural disadvantage may indeed be justied or even
required, such justication requires arguments additional to those this article
offers.
I. CHALLENGES TO GROUP-CONSCIOUS MEASURES OF EQUALITY
The intuition is deep and widespread in our society that assessments of equality
and inequality should refer to social groups dened by such characteristics as

EQUALITY OF WHOM?

gender, race, age, ethnicity, occupation, class, education, and so on. For example,
there has been considerable debate on both sides of the issue about whether
inequalities between groups have disappeared now that laws make status
differences or group-based discrimination illegal. Social scientic research usually
documents and explains inequality on some dimension, such as political equality
or economic equality, by disaggregating society into social group categories.2 In
response to these interests, many government agencies and private organizations
routinely collect and publish statistics designed to show how various groups of
persons compare with other groups on measures of resource distribution,
opportunity, privilege or inuence. Everyday practical discourse about equality,
then, seems to express little doubt that judgments about inequality and equality
ought to refer to groups. Practices of evaluating inequality by comparing groups,
however, have recently come under both political and philosophical challenge.

In political and legal debates about social justice in the United States, many
question giving any attention to the relative situation of groups in formulating
public policy. Certain Supreme Court decisions in the mid-1970s found that the
obligation to redress harm could properly be assessed in terms of the ``disparate
impact'' which action or policies had on groups of persons, especially those
historically discriminated against on grounds of sex, race or ethnicity. More
recently, the Court seems to be rejecting any argument that harm can be
evaluated in terms of results affecting groups unequally, and insisting instead on
claimants demonstrating that policies or actions contain invidious classications
intended to disadvantage individuals in them as compared with others.3 Recent
attacks on the diverse policies associated with the label ``afrmative action,'' to
take another example, not only seem to reject preferential treatment of
individuals associated with groups claimed to be unequal, but also to be
attacks on the very idea that groups can be unequal, or that we should care
whether they are, as long as we afrm principles of equal treatment and formally
equal opportunity.
At least in the United States, political controversy about data collection that
uses group classication and the use of such data for making group comparisons
on measures of well-being has focused almost exclusively on race and ethnicity
classications. Issues of data collection by gender, age, ability, occupation, sexual
2
For a recent comprehensive study of aspects and degrees of political inequality, see Sidney Verba,
Kay Lehman Schlozman and Henry E. Brady, Voice and Equality: Civic Voluntarism in American
Politics (Cambridge, Mass.: Harvard University Press, 1995). For one discussion of economic
inequality, see Sheldon Danziger and Peter Gottschalk, America Unequal (Cambridge, Mass.:
Harvard University Press, 1995). Both studies compare inequalities by gender, race and class.
3
One example: In his concurring opinion in City of Richmond v. J.A. Croson Co.488 US 469, 527
(1989), Justice Anthony Scalia specically says that the wrong of discrimination involves not groups,
but individuals.

IRIS MARION YOUNG

orientation, income, occupation of parents, or other classications reasonably

argued to have social meaning appear to be less contested. Prompted in part by
the claims of people whose lineage and self-identication cross two or more
classications, such as white, Black, or Asian, some people have argued that the
United States census should cease asking people to identify themselves by race or
ethnicity at all.4 Many of those arguing against collecting data by racial or ethnic
categorization do so because they opposed any group-conscious measurement of
inequality. They have a number of reasons.
The use of group classications for any purpose amounts to invidious
discrimination, they suggest. The long struggle against practices and policies that
privileged white male Christians was fought on the principles that each person
should be considered only as an individual; we should be blind to differences in
ascribed characteristics or social membership. The goal of a free society should be
color-blindness, gender-blindness, blindness to all those ascribed characteristics
that historically served as markers of inferiority and exclusion. This goal is only
thwarted by assessments of well-being that compare categories of people. Race
and ethnicity classications presume xed immutable identities, and they thereby
perpetuate the very social divisions their benign uses are intended to overcome.
Assessing inequality in terms of social groups and deriving claims of injustice
from such comparisons, nally, encourages damaging political divisiveness. It
encourages some to claim a victim status for their group, and public policy debate
becomes a zero sum contest among these xed identity groups. Political
cooperation will be more easily obtained if evaluations of well-being consider
individuals only.5

Philosophical theories of equality typically assume without argument that the

entities we ought to compare when making evaluations of equality and inequality
are individuals. A few political philosophers, however, do explicitly defend the
claims that only individuals, and not groups, are proper subjects of judgments
about equality.
Ronald Dworkin criticizes the group-conscious evaluations of equality he nds
at the base of Rawls's theory of justice. Rawls argues that reasoning about
principles of justice behind the veil of ignorance assumes the point of view of the
4
These challenges have failed. Some proposed a category of ``mixed race'' for the US census, a
proposal supported by very few. The solution that the Census has adopted is to allow respondents to
identify themselves by as many categories as they wish. Steven A. Holmes, ``The politics of race and
the census,'' The New York Times, Sunday, March 19, 2000, wk p. 3.
5
For some examples of such positions, see: Clint Bolick, The Afrmative Action Fraud: Can We
Restore the Civil Rights Vision? (Washington, D.C.: Cato Institute, 1996); Michael Lind, The Next
American Nation: The New Nationalism and the Fourth American Revolution (New York: Free Press,
1995); Nick Gillespie, ``Blurred vision: seeing beyond government racial categories,'' Reason
Magazine, July 1997.

EQUALITY OF WHOM?

representative person in that class of persons located in the least advantaged

position. On this construction, evaluations of justice are essentially tied to a
comparison of the situation of groups or classes. Dworkin traces the causes of
such group consciousness to Rawls's overly ``at'' conception of equality, which
compares classes of people only in economic terms. All the many differences in
the tastes, choices and goals of individuals with similar income drop out.
Dworkin's own theory of equality of resources is precisely designed to take
account of such differences, with the particular aim of distinguishing differences
that derive from tastes and choices of individuals from those deriving from
natural misfortunes and other sources for which individuals are not responsible.
Dworkin offers his theory as methodologically and morally individualistic.
It aims to provide a description (or rather a set of devices for aiming at) equality of
resources person by person, and the considerations of each person's history that
affect what he should have, in the name of equality, do not include his membership
in any economic or social class....The theory proposes that equality is in principle a
matter of individual right rather than group position.6

Larry Temkin argues against measuring inequality in terms of groups because

it seems to locate moral concern with equality in a concern for groups rather than
individuals. Moral evaluation and public policy, however, should be concerned
with how individuals fare.
The core of the egalitarian's position is that it is bad (unjust or unfair) for one person
to be worse off than another through no fault of her own. Ultimately, then, the
egalitarian is concerned with how individuals in situations fare in relation to one
another. Correspondingly, concern about inequality between society's groups must
ultimately be understood as concern about inequality between groups' members.7

Douglas Rae worries that group-conscious assessments of inequality wrongly

collapse individual circumstance to a group average. Within any putative group
classication there are always many individual differences. On any measure of
resource distribution, welfare, capabilities or opportunity, there is likely to be a
wide range of individual variation within groups. Assertions that groups are
unequal on some measure of well-being thus must rely on comparing statistical
averages or medians between groups. Such comparisons create the false
impression of a unity or similarity within groups and polarize dissimilarity
between them. Discussions of inequalities among groups often fail to mention
that some members of the supposedly disadvantaged group are better off than
some members of the supposedly advantaged group. Those concerned with
gender equality, for example, might cite the fact that the average income of
women with a high school education is signicantly lower than the average
income of male high school graduates. Some women graduates, however, earn
6
Dworkin, ``What is equality? Part 2: equality of resources,'' Philosophy and Public Affairs, 10
(1981), p. 340.
7
Larry Temkin, Inequality (Oxford: Oxford University Press, 1993), p. 101.

IRIS MARION YOUNG

signicantly higher incomes than some men. When we criticize inequalities of

that sort between groups, moreover, we seem to be calling for change that would
equalize the group statuses. Justice is achieved, on this view, by employment or
taxation policy that aims to make the average incomes of women the same as the
average incomes of men. Such group-based action, however, is liable to deny
some individuals in the groups what they deserve for the sake of a patterned
outcome that itself may hide signicant individual variation and inequality
within each group.8
I agree with Dworkin and Temkin that the ultimate purpose for making
assessments of inequality is to promote the well-being of individuals considered
as irreducible moral equals. That is to say, if there are good reasons for groupconscious measurements of inequality, they are not that we should care about
groups as such, considered as independent sources of moral value. I will argue
below, however, that in making some of the most important judgments of justice
and injustice we must compare social groups such as women, African Americans,
migrants, or people with disabilities. Categories such as these name groups are
positioned by social structures that constrain and enable individual lives in ways
largely beyond their individual control. I will later explain this concept of
structural group and argue that comparing groups on measures of inequality is
necessary in order to discover such social structures and make judgments that
inequalities grounded in those structures are unjust.

II. INEQUALITY AND MORAL JUDGMENT

I assume that philosophers and political actors are interested in theorizing and
making evaluations of inequality primarily in order to support judgments of
justice or injustice. To be sure, not all claims of justice concern issues of equality,
and shortly I will refer to some that do not. Judgments of equality or inequality in
themselves, however, are simply factual comparisons of amounts or degrees of
some variables between or among entities. Such comparisons by themselves do
not yield judgments with the moral force that claims about social, economic or
political equality usually carry.
Some question whether considerations of substantive equality are relevant at
all when making judgments about justice. On this view, justice is primarily about
liberty. If justice includes reference to equality, it is only as formal procedural
equality. As long as procedural equality is observed in law and the enforcement
of contracts, on this account, and as long as people are free to try to realize their
goals through voluntary exchange and associations, then considerations of
inequality on some measures of power, inuence or material well-being are
irrelevant to judgments of justice.
8

81.

Douglas Rae, Equalities (Cambridge, Mass.: Harvard University Press, 1981), esp. pp. 2938, 76

EQUALITY OF WHOM?

I do not intend here to rebut this position, because it would take me away from
my main question. Instead I shall assume that if we are concerned about some
measure of equality and inequality, it is because we have a conception of justice
for which such assessments of equality are relevant. My thesis is, then, that
claims about social justice that invoke equality usually require comparison of
groups on measures of well-being or advantage. As I will explain below, claims
that some inequalities are unjust implicitly or explicitly compare groups in order
to identify social structures that involuntarily position people, constraining some
more than others and privileging some people more than others. Such a
conception says that these structures produce injustice insofar as they afford
people different degrees of opportunity to achieve well-being. My account here
does not claim that all injustice is reducible to these effects of structural
positioning. Many claims of justice that invoke liberty, need, or desert do not
require identication of structural social groups to become meaningful and gain
normative plausibility. The claim here is only that claims of justice that invoke
equality generally do.
Assessment of inequality in terms of the comparison of individuals yields little
basis for judging injustice. Suppose we discover that some individuals are
starving and others have access to more food than they need. We have discovered
a serious inequality, and one that indeed is likely to provoke moral judgment that
something is wrong. It is wrong, unjust, some would say, that some people
should starve when others have more than they need. Those who are well-off
have a prima facie moral obligation to give to the starving people. If there is such
an obligation, however, and I believe that there is, it derives not from the fact of
inequality as such, but from the fact of need. It is wrong for some people to lack
what they need to live a minimally decent life when others are able to contribute
to meeting those needs at relatively little cost to themselves. Many criticisms of
inequalities probably have this form. They are judgments about obligations to
help people reach a certain level of what Harry Frankfurt calls ``sufciency,''
however, rather than judgments that an inequality is wrong as such.9
So, let us assume a situation in which there is an accepted denition of
sufciency and everyone being compared on measures of well-being starts at or
above this baseline. Some individuals, let us suppose, have incomes 100 times
those with the lowest incomes, while some have incomes only 20 times those with
the lowest incomes. Or suppose that some individuals who are not
democratically legitimate public ofcials have signicantly more inuence than
others over important policy decisions. Or suppose that some people require
more time, effort and planning to get to a polling place or a public hearing than
others. Even assuming that all the individuals being compared have a sufcient
base line, each of these inequalities, I suggest, probably signals injustice. Simply
by focusing on comparing the situation of individuals, however, without any
9

Harry Frankfurt, ``Equality as a moral ideal,'' Ethics, 98 (1987), 2143.

IRIS MARION YOUNG

reference to attributes or afnities they share with others or generalized social

relations in which they stand, we have no reason to call these inequalities
injustices.
Why not? Because if we simply identify some inequality of condition or
situation between individuals at a particular time we have no account of the
causes of this unequal condition. It is the causes and consequences of some
pattern of inequality, rather than the pattern itself, that raise issues of justice. If
the causes of an inequality lie in the uncoerced and considered decisions and
preferences of the less well-off persons, for example, then the inequality is
probably not unjust. Dworkin and many other equality theorists agree. Their
theories are devoted to distinguishing those causes for an individual's resource
holdings or level of welfare for which he or she can be said to be responsible from
those for which he or she is not responsible. Most of these theories place the latter
set of causes under a vague category of bad luck.
A large set of the causes of an unequal distribution of reources or unequal
opportunities between individuals, however, is attributable neither to individual
preferences and choices nor to luck or accident. Instead, the causes of many
inequalities of resources or opportunities among individuals lie in social
institutions, their rules and relations, and the decisions others make within
them that affect the lives of the individuals compared.10 When a government
body holds a hearing in the city center on issues that affect a poor rural
population, neither dispersing ofcials to other locations nor providing
transportation to the affected group, then this group's opportunity to inuence
the decision is not equal to those who live close to the center. They are not
victims of bad luck, however, but of institutional routines that convenience
public ofcials and business leaders, a history of transportation planning or the
lack of it, regional development inequalities, and so on. People similarly
positioned in social structures frequently experience multiple forms of exclusion,
unequal burdens or costs deriving from institutional organization, rules, or
decisions, and the cumulative consequences of each.
Those interested in equality in the United States often cite the following fact in
order to elicit intuitions of unfairness: roughly 80 percent of the wealth in the US
is owned by roughly 20 percent of the population. It is disingenuous, however, to
hold that this manner of aggregating individuals and comparing their conditions
in itself gives grounds for a judgment of injustice. If everyone has enough to meet
their needs, if there is nothing preventing the less wealthy from becoming
wealthy, if the wealthy receive no social privileges unavailable to the less wealthy,
and so on, then such an unequal distribution may be one of those things, a
combination of luck and choice. To reach a judgment that there is something
wrong with such an unequal distribution of wealth, we must ask more about the
lives of those in the wealthy group as compared with the group with little wealth.
10

Elizabeth Anderson, ``What Is the Point of Equality?'' Ethics:, 109 (1999), 287337.

EQUALITY OF WHOM?

When we learn that more of the wealthy had wealthy parents, were educated at
the most elite and resourced universities, and so on, and we compare their life
opportunities with those in the less wealthy group, then we can begin to make
judgments of justice. We have moved from assessment of inequality in terms of
aggregations of individuals to comparisons of social groups, in this case social
classes. The importance of measuring inequality in terms of social groups such as
class, gender, race, I argue below, lies in that it reveals the structural inequalities
which are particularly relevant for making judgments of justice and injustice.

III. STRUCTURAL INEQUALITY

Social critics often describe the inequalities they nd particularly important or
broadly conditioning of people's lives by referring to structural or systemic
relations. Though political philosophers less often distinguish a structural from a
more individual level of assessing inequality, there are notable exceptions. Thus
in The Community of Rights, Alan Gewirth claims that his human rights
approach to justice shares with libertarianism the valuing of individual
responsibility for using one's freedom to support oneself by one's own efforts.
The human rights approach fundamentally differs from libertarianism, however,
in that the libertarian ``overlooks that many persons do not have and cannot use
their freedom in the ways he exalts, and that these negative facts derive from, or
are vitally affected by, the structural inequalities of existing institutional
arrangements.''11
Jean Hampton also recommends attending to social structural or systemic
factors for theorizing justice. Political theorizing that is too individualist, she
argues, ``can fail to acknowledge that all sorts of social structures in our society,
including our legal institutions, family structures, and systems of educating the
young, have been worked out by generations of people responding to a variety of
problems in ways that are complicated, nuanced, and often highly successful.''12
Hampton recommends a post-liberal political theory that not only attends to
issues of individual freedom and the dangers of state tyranny, but also theorizes
``the harm of oppression coming from systemic effects of certain kinds of social
institutions in which individuals nd themselves and operate. These social forms
are such that individuals, despite any good intentions they might have, are forced
to act and react in ways that result in considerable damage to some people.''13
This post-liberal theory, according to Hampton, realizes that ``it is a mistake to
try to impose what one takes to be a `fair' pattern of distribution on a society

11

Gewirth, The Community of Rights (Chicago: University of Chicago Press, 1996), p. 34.
Jean Hampton, Political Philosophy (Boulder: Westview, 1997), pp. 18990.
13
Ibid, p. 191. It may be too strong to say that people are ``forced'' by structures to act in ways that
result in damage to some people. Rather people act under constraints in ways I will detail below.
12

10

IRIS MARION YOUNG

without realizing that these systemic forces are going to be propelling the society
toward a certain (unfair) distribution anyway.''14
Gewirth and Hampton refer to structural inequality in order to distinguish
individual attributes, actions and choices from more socially collective or
institutional conditions under which these occur, and which limit individual
options and action. They argue that such a distinction is necessary both in order
to identify what in their situation individuals are responsible for, and what kinds
of policies are likely to work to promote greater equality. Yet neither gives a clear
account of the meaning of this concept of structural inequality. I propose to take
some steps toward such clarication.
Marilyn Frye likens oppression to a birdcage. The cage makes the bird entirely
unfree to y. If one studies the causes of this imprisonment by looking at one wire
at a time, however, it appears puzzling. How does a wire only a couple of
centimeters wide prevent a bird's ight? One wire at a time, we can neither
describe nor explain the inhibition of the bird's ight. Only a large number of
wires arranged in a specic way and connected to one another to enclose the bird
and reinforce one another's rigidity can explain why the bird is unable to y
freely.15
At a rst level of intuition, this is what I mean by social structures that inhibit
the capacities of some people. An account of someone's life circumstances
contains many strands of difculty or difference from others that, taken one by
one, can appear to be either the result of decision, preferences or accidents. When
considered together, however, and when compared with the life story of others,
they reveal a net of restricting and reinforcing relationships. Let me illustrate.
Susan Okin gives an account of women's oppression as grounded in a gender
division of labor in the family. She argues that gender roles and expectations
structure men's and women's lives in systemic ways that result in disadvantage
and vulnerability for many women and their children. Institutionally, the entire
society continues to be organized around the expectation that children and other
dependent people ought to be cared for primarily by family members without
formal compensation. Good jobs, on the other hand, assume that workers are
available at least forty hours per week year round. Women are usually the
primary caretakers of children and other dependent persons, due to a
combination of factorstheir socialization disposes them to choose to do it,
and/or their job options pay worse than those available to their male partners, or
their male partners' jobs allow them little time for care work. As a consequence
the attachment of many women to the world of employment outside the home is
more episodic, less prestigious, and less well paid than men's. This fact in turn
often makes women dependent on male earnings for primary support of
themselves and their children. Women's economic dependence gives many men
14

Ibid, p. 206.
Marilyn Frye, ``Oppression,'' The Politics of Reality (Trumansburg, N.Y.: The Crossing Press,
1983).
15

EQUALITY OF WHOM?

11

unequal power in the family. If the couple separates, moreover, prior dependence
on male earnings coupled with the assumptions of the judicial system make
women and their children vulnerable to poverty. Schools, media and employers'
assumptions all mirror the expectation that domestic work is done primarily by
women, which assumptions in turn help reproduce those unequal structures.16 In
this account, interlocking social structures of family and economy, as well as
cultural norms, shape choices and thus explain the unequal conditions of women
and their children. Without such an account, it is difcult to describe women's
unequal position as a matter of injustice.
Processes that produce and reproduce residential racial segregation provide
another example of the structural conuence of many distinct actions,
expectations and effects to limit the options of many inner city dwellers in the
United States. Racially discriminatory behavior and policies limit the housing
options of people of color, conning many of them to neighborhoods from which
many of those whites who are able to leave do. Property owners fail to keep up
their buildings and new investment is hard to attract because the value of
property appears to decline. Because of more concentrated poverty and lay-off
policies that disadvantage Blacks or Latinos, the effects of an economic downturn
in minority neighborhoods are often felt more severely, and more businesses fail
or leave. Politicians often are more responsive to the neighborhoods where more
afuent and white people live; thus schools, re protection, policing, snow
removal, garbage pickup, are poor in the ghetto neighborhoods. Economic
restructuring independent of these racialized processes contributes to the closing
of major employers near the segregated neighborhoods and the opening of
employers in faraway suburbs. As a result of the conuence of all these actions
and processes, many Black and Latino children are poorly educated, live around
demoralized people in dilapidated and dangerous circumstances, and have few
prospects for employment.17
These accounts exhibit structural inequality of gender and race respectively.
While attributes of individuals also condition how they will be identied and
treated by others, the primary account of gender or racial inequality here is
structural. They describe a set of relationships among assumptions and
stereotypes, institutional policies, individual actions following rules or
choosing in self-interest, and collective consequences of these things, which
constrain the options of some at the same time as they expand the options of
others. One could tell analogous stories of how the economic class position of
one's parents, neighborhoods and friends condition much about a person's life
options because of the structural inequalities of class.
Now let us systematize the notion of structure. To do so I will build up an
account using elements derived from several theorists. Peter Blau offers the
16

Susan Okin, Justice, Gender and the Family (New York: Basic Books, 1989).
See Douglas Massey and Nancy Denton, American Apartheid (Cambridge, Mass.: Harvard
University Press, 1993).
17

12

IRIS MARION YOUNG

following denition: ``A social structure can be dened as a multidimensional

space of differentiated social positions among which a population is distributed.
The social associations of people provide both the criterion for distinguishing
social positions and the connections among them that make them elements of a
single social structure.''18 Blau exploits the spatial metaphor implied by the
concept of structure. Individual people occupy varying positions in the social
space, and their positions stand in determinate relation to other positions. The
structure consists in the connections among the positions and their relationships,
and the way the attributes of positions internally constitute one another through
those relationships. The position of supervisor in a workplace hierarchy is
constituted by relationship to determinate subordinates, and their positions are
constituted by relations to the supervisor position.
For purposes of inquiring about social justice, we are interested not in such
specic positions as foreman on the automobile assembly line, but in more
generalized positions that constitute what John Rawls calls the ``basic structure''
of society. Rawls says that the basic structure concerns ``the way in which the
major social institutions distribute fundamental rights and duties and determine
the division of advantages from social cooperation.''19 Major institutions
include, he says, the legal system's denition of basic rights and duties, market
relations, the system of property in the means of production, and family
organization. To these I would add the basic kinds of positions in the social
division of labor.
Taken together as one scheme, the major institutions dene men's rights and duties
and inuence their life prospects, what they can expect to be and how well they can
hope to do. The basic structure is the primary subject of justice because its effects are
so profound and present from the start. The intuitive notion here is that this
structure contains various social positions and men born into different positions
have different expectations of life determined, in part, by the political system as well
as by economic and social circumstances.20

Basic social structures consist in determinate social positions that people

occupy which condition their opportunities and life chances. These life chances
are constituted by the ways the positions are related to one another to create
systematic constraints or opportunities that reinforce one another like wires in a
cage. Structural social groups are constituted through the social organization of
labor and production, the organization of desire and sexuality, the
institutionalized rules of authority and subordination and the constitution of
prestige. Structural social groups are relationally constituted in the sense that one
position in structural relations does not exist apart from differentiated relation to
other positions. A position in the social division of labor, for example, is what it
18
19
20

Peter Blau, Inequality and Heterogeneity (New York: Free Press, 1977), p. 4.
John Rawls, A Theory of Justice (Cambridge, Mass.: Harvard University Press, 1971), p. 7.
Ibid.

EQUALITY OF WHOM?

13

is only in the context of the larger organization of productive activity to which it

is related.
It is certainly misleading, however, to reify the metaphor of structure, that is,
to think of social structures as entities independent of social actors, lying
passively around them easing or inhibiting their movement. On the contrary,
social structures exist only in the action and interaction of persons; they exist not
as states, but as processes. Thus Anthony Giddens denes social structures in
terms of ``rules and resources, recursively implicated in the reproduction of social
systems.''21 In the idea of the duality of structure, Giddens theorizes how people
act on the basis of their knowledge of pre-existing structures and in so acting
reproduce those structures. We do so because we act according to rules and
expectations and because our relationally constituted positions make or do not
make certain resources available to us.
For example, the parents of upper-class children are supposed to act, and
usually do act, to make sure that their children are able to maintain their class
position. They have income resources to nurture their children in comfortable
surroundings, and to provide high quality private or public schools. They have
acquired the education and cultural habits through which they supplement the
school's education with knowledge, skills, tastes and expectations that socialize
their children into what others of the class admire, or nd valuable. In their
interactions with teachers and local politicians they are treated with particular
respect because of the resources they can mobilize, or because of their knowledge,
demeanor and position. For all these reasons the road to high status, power and
afuence is much smoother for these children than those of a low-income single
mother without a high school education. The everyday interactions of people in
child raising both presuppose the rules and resource mobilization capacity that
attend a higher status and reproduce them.
Dening structures in terms of the rules and resources brought to actions and
interactions, however, makes the reproduction of structures sound too much like
the product of individual and intentional action. The concept of social structure
must also include conditions under with actors act, which are often a collective
outcome of action impressed onto the physical environment. Jean-Paul Sartre
calls this aspect of social structuration the practico-inert.22 Most of the
conditions under which people act are socio-historical: they are the products of
previous actions, usually products of many coordinated and uncoordinated but
mutually inuenced actions over them. Those collective actions have produced
21
Giddens, The Constitution of Society (Berkeley: University of California Press, 1984), p. 25. I rely
on Giddens' account mindful that it may be too vague and static for some purposes. William Sewell,
``A theory of structure: duality, agency, and transformation,'' American Journal of Sociology, 98
(1992), 129, makes a careful critique of aspects of Giddens' theory, and proposes revisions in the
meaning and relations of rules and resources better able to account for structural change as well as
stability.
22
Sartre, Critique of Dialectical Reason, trans. Alan Sheridan-Smith (London: New Left Books,
1976), bk. I, ch. 3.

14

IRIS MARION YOUNG

determinate effects on the physical and cultural environment which condition

future action in specic ways. As I understand the term, social structures include
this practico-inert physical organization of buildings, but also modes of transport
and communication, trees, rivers and rocks and their relation to human action.
American metropolitan areas, for example, have increasingly separated the
physical space of family life from other spaces over the last three decades. The
location of suburban residences are relatively farther away from schools,
shopping districts, factories and ofces than were urban residences fty years
ago. This fact, coupled with the assumption that care for children is primarily the
responsibility of family members as unpaid work in and around the home,
contributes to conditioning and constraining the lives of these home care workers
who, due both to the rules and expectations of gender, and the often greater
resources and opportunities available to men, are usually women. The spatial
relation of residences to the space of other activities means that women's care
work often consists in travelingto schools, after-school activities, health
providers, etc. If she wishes or needs also to have paid employment, the physical
organization of her unpaid work tends to restrict her choices to close to home or
one of these other activities. These facts make mothers of school-age children
something of a captive labor force for part-time suburban retail and service work,
and often make the workers more willing than others to work for lower wages.
Reference to such physical aspects of social structures leads us to a nal aspect
of the concept. The actions and interactions which take place among persons
differently situated in social structures using rules and resources do not take place
only on the basis of past actions whose collective effects mark the physical
conditions of action. They also often have future effects beyond the immediate
purposes and intentions of the actors. Structured social action and interaction
often have collective results that no one intends and which may even be counter
to the best intentions of the actors.23 Even though no one intends them, they
become given circumstances that help structure future actions. Presumably no
one intends that many children be vulnerable to poverty in the way that Okin
claims the gender division of labor makes them. Indeed, this result contradicts a
common justication for the gender division of labor, which says that it is for the
sake of the children that one parent should have primary responsibility for their
care at home.
In summary, structures refer to the relation of basic social positions that
fundamentally condition the opportunities and life prospects of the persons
located in those positions. This conditioning occurs because of the way that
actions and interactions in one situation conditioning that position reinforce the
rules and resources available for other actions and interactions involving people
in other structural positions. The unintended consequences of the conuence of
many actions often produce and reinforce such opportunities and constraints,
23

Sartre calls such effects counter-nalities; see ibid, especially pp. 27792.

EQUALITY OF WHOM?

15

and these often make their mark on the physical conditions of future actions, as
well as on the habits and expectations of actors. This mutually reinforcing
process means that the positional relations and the way they condition individual
lives are difcult to change.
Structural inequality, then, consists in the relative constraints some people
encounter in their freedom and material well-being as the cumulative effect of the
possibilities of their social positions, as compared with others who in their social
positions have more options or easier access to benets. These constraints or
possibilities by no means determine outcomes for individuals in their ability to
enact their plans or gain access to benets. Some of those in more constrained
situations are particularly lucky or unusually hardworking and clever, while
some of those with an open road have bad luck or squander their opportunities
by being lazy or stupid. Those who successfully overcome obstacles, however,
nevertheless cannot be judged as equal to those before whom few obstacles have
loomed, even if at a given time they have roughly equivalent incomes, authority
or prestige. Unlike the individualized attributes of native ability that often
concern equality theorists, moreover, structural inequalities are socially caused.
For this reason there is an even stronger argument than in the case of given
individual attributes for social institutions to remedy these inequalities.

IV. GROUPS AND JUDGMENTS OF INJUSTICE

We can now return to our initial question: What, if anything, justies comparing
social groups in various measures of well-being or status, in order to nd
inequalities between the groups? Assessing degrees of inequality between groups
in this way is necessary and justied because it helps to identify structural
inequalities. If observation discovers that a category of individuals is unequal to
others on certain important measures of well-being, and a plausible story can be
told about how the relations, rules, expectations, and cumulative consequences
of collective action specically condition the lives of members of that group, then
there are grounds for saying that members of the group suffer some injustice.
There is a recursiveness in this method. When we criticize social conditions for
being unjust, we do not choose random attributes of individuals, such as their
taste in music or their bone structure, as a basis for grouping them. We construct
the groups for comparison according to generally recognized social positions
which we already know have broad implications for how people relate to one
anotherclass, race, ethnicity, age, gender, occupation, ability, religion, caste,
citizenship status, and so on. The process of evaluating group inequality for the
sake of making judgments about injustice begins with the hypothesis that
comparing the average status of members of some of these groups will reveal
patterns of inequality. By a pattern, I mean the mapping of the distribution of
some good across all social positions at a particular time.

16

IRIS MARION YOUNG

Discovering one such pattern on one parameter, however, does not yet take us
to a judgment of injustice. We must discover that such inequality is systemic by
nding a pattern of average difference in level of status or well-being along
several parameters. When we nd that Native Americans as a group have the
lowest incomes, highest infant mortality rates, least education, and so on, of any
group in American society, then we are entitled to say that members of this group
probably suffer injustice. We are not warranted in the full evaluation, however,
unless we can tell a plausible structural story that accounts for the production of
the patterns. To complete the analysis and evaluation, we must explain how
institutional rules and policies, individual actions and interactions, and the
cumulative collective and often unintended material effects of these relations
reinforce one another in ways that restrict the opportunities of some to achieve
well-being in the respects measured, while it does not so restrict that of the others
to whom they are compared, or even enlarge their opportunities. This story will
be aided, moreover, by evidence that the basic conguration of the patterns
shows little change over decades. The measures of group inequality are like wires
in the cage; seen alone they reveal nothing about people's freedom or well-being.
Seen together as spaced and reinforcing, however, they explain a great deal.
Ultimately the judgments of injustice, then, are not about the distributive
patterns. Each distributive pattern only offers a piece of a puzzle, a clue to an
account of generalized social processes which restrict the opportunities of some
people to develop their capacities or access benets while they enhance those of
others.24 A large class of social inequalities can be judged as unjust because they
violate a broad principle of equal opportunity: that it is unfair to some
individuals to have an easy time ourishing and realizing their goals, while others
are hampered in doing so, due to circumstances beyond their control. Many
theories of equality interpret the main source of such circumstances as bad luck. I
agree with Elizabeth Anderson that the purpose of equality theory is less to
identify unlucky sources of inequality than to identify how institutions and social
relationships differentially conspire to restrict the opportunities of some people
to develop and exercise their capacities and enact their goals. Individuals alone
are not responsible for the way they are enabled or constrained by structural
relations. To the extent that injustices are socially caused, however, this
conception of justice claims that democratic political communities are
responsible collectively for remedying such inequalities, perhaps more than
they are obliged to remedy the effects of so-called ``brute luck.'' These judgments
of injustice take institutional organization and social practice as their object more
than distributive patterns per se.

24
I have argued in another place that social process is more important than distributive pattern for
making evaluations of justice and injustice. See I. M. Young, Justice and the Politics of Difference
(Princeton, N.J.: Princeton University Press, 1990), ch. 1.

EQUALITY OF WHOM?

17

V. RESPONSE TO OBJECTIONS AND POLICY IMPLICATIONS

Assessing structural inequality by referring to groups does not fall prey to most of
the criticisms of group-based assessments of inequality discussed earlier. Nothing
in the argument about assessing inequality in terms of structures contradicts the
claim that individual persons are and ought to be the nal targets of judgments
and policies aimed at producing or improving well-being. My argument does not
assume a reied concept of group distinct from individuals. Instead it notices that
in pursuing individual and institutional goals, people themselves treat others as
group members, and that the product of many such actions sometimes results in
structural inequalities. Comparison of groups is necessary to judge inequalities
unjust because doing so helps show that different individuals have different
opportunities and inhibitions related to their structural social positions.
This analysis of structural inequality, moreover, shows why the groups of most
concern to political actors and policy makers when they make group-conscious
judgments of equality are not arbitrary collections of persons. Identifying groups
for the purposes of such evaluations depends on knowing social positions,
relations and structures, and how they open or restrict opportunities. There are
two levels of knowing: a practical consciousness of action, where people know
what are the primary categories according to which persons in the society are
often treated differently or noticed as socially different. This practical knowledge
motivates the initial data-gathering that aggregates measures of well-being in
terms of groups. But more theoretical and systematically empirical investigation
must take place before we have warrant to claim that there are unjust structural
inequalities. I have said that this involves nding patterned inequality and being
able to tell a plausible story about how the position in structures accounts for
that inequality.
The worry that group-conscious measures of inequality suppress individual
variation within groups is not warranted, because the purpose of assessing
inequalities between groups is not to equalize the groups as blocs, without
consideration of the relative status of individuals within groups. The goal of
equality here considered is strong equality of opportunity among individuals,
rather than some static average state of affairs between groups. Measuring
inequalities between groups is a means of arguing that structures of privilege and
disadvantage give individuals unequal opportunities.
Using group-conscious measures of inequality may indeed exacerbate political
conict to the extent that it identies injustices that call for remedy, especially if
the design of remedies takes the social circumstances of different groups into
account. Even assuming that this is true, it cannot count as a reason against the
practice of assessing inequality in terms of groups. Making claims of injustice,
especially when they involve large numbers of people or social segments, often
produces conict, because people disagree either on principle or because they see
their interests threatened. Once judgments of injustice have been made, attempts

at remedy are often divisive, because some people believe that they benet from
the status quo more than from change that would address the injustice. It may be
true that identifying some groups as candidates for judgments of unjust inequality
gives some people incentive to propose themselves as members of structurally
unequal groups who deserve remedial policy. Some of these claims are likely to
be groundless, and they ought to be contested in public by means of principles
and criteria of justice and injustice. Far from inhibiting such efforts to challenge
spurious claims of group disadvantage, the concepts and methods of analyzing
structural inequality offered here provide basis for such challenges.
Judgments of injustice based on discovering inequalities between groups invite
action to redress the injustice. Such judgments of inequality between groups,
however, do not imply any specic principles or strategies for remedial social
policy. In particular, it does not follow from the fact that the judgments of
injustice use comparison of groups on measures of inequality that policies to
correct this injustice ought to target groups. The argument of this paper, then,
should not be construed as a defense of afrmative action policies, or other