Book Search:  


Google full text of our books:


Identity in Democracy
Amy Gutmann

Book Description | Reviews

COPYRIGHT NOTICE: Published by Princeton University Press and copyrighted, © 2003, by Princeton University Press. All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World Wide Web. Users are not permitted to mount this file on any network servers. Follow links for Class Use and other Permissions. For more information, send e-mail to

This file is also available in Adobe Acrobat PDF format



Identity groups occupy an uneasy place in democracy. Critics emphasize how much group identities constrain rather than liberate individuals. When people are identified as black or white, male or female, Irish or Arabic, Catholic or Jew, deaf or mute, they are stereotyped by race, gender, ethnicity, religion, and disability and denied a certain individuality that comes of their own distinctive character and freedom to affiliate as they see fit. When individuals themselves identify racially, ethnically, or religiously as a consequence of being identified with groups, they often develop hostilities toward other groups and a sense of superiority over them. Groups frequently vie against one another in uncompromising ways, sacrificing justice and even peace for vindicating their superiority as a group.

If critics told the whole story, we would have little reason to doubt that identity groups are up to no good from a democratic perspective. Defenders of identity politics point out some of the problems with the critics' image of the autonomous, self-made person who neither identifies nor is identified with groups. Without any group identities, defenders of group identity say individuals are atomistic, not autonomous. Group identities help individuals have a more secure sense of self and social belonging. Moreover, group identity propels women and disadvantaged minorities to counteract inherited negative stereotypes, defend more positive self-images, and develop respect for members of their groups.

What the defenders and critics of identity groups have to say is significant, but each captures only part of the relationship between identity groups and democratic politics. The relationship is far more complex yet no less important than that suggested by these and other common defenses and critiques of identity politics. People identify with others by ethnicity, race, nationality, culture, religion, gender, sexual orientation, class, disability, age, ideology, and other social markers. No single group identity or even all group identities taken together comprehend the whole of a person, yet a commonly shared identification around any of the above characteristics of a person often leads to a group identity. Group identities are as abundant in democracies as they are controversial. Politically significant associations that attract people because of their mutual identification are aptly called identity groups.

Were it not for the mutual identification of individuals with one another, there would be no identity groups. Although mutual identification is basic to human existence, it has been neglected in democratic theory, where the language of "interest" and "interest groups" (soon to be discussed), rather than identity and identity groups, is far more common. Yet no one should doubt that identification with others makes a difference in how individuals perceive their own interests. Psychological experiments demonstrate that something as basic as self-image changes when individuals identify with others. And just as remarkably, a difference in self-image can be based on a seemingly, irrelevant identification with others. Experimental subjects who view a beautiful stranger report an increase in their own self-image of attractiveness when all that they learn about the stranger is that they share her birthday. The experimental subjects apparently identify with the total stranger by virtue of sharing the same birthday, and that identification alone is enough to enable her beauty to enhance their own self-image.1 Conversely, a negative difference can arise from group identification when women students are reminded of their gender or African American students of their racial identity before taking tests in subjects where it is widely thought that women and African Americans perform poorly.2 Democratic theory and politics clearly cannot afford to neglect the differences, both positive and negative, that group identifications make in people's lives.

What difference does the existence of organized identity groups make for democratic theory and practice? When is nationality, race, religion, gender, sexual orientation, or some other group identity a good or bad reason for democratic action? What identity groups should be encouraged or discouraged, and what actions based on identity can aid or impede democratic justice?

The analysis that follows suggests that organizing politically on the basis of group identity is not a good or bad thing in itself. When identity groups put the group above opposition to injustice or the pursuit of justice, they are morally suspect. Identity groups do better when they offer mutual support and help combat injustice for disadvantaged people. Even when combating injustice is justified, it can be ugly. A completely justifiable struggle against the rights violations of an identity group, such as the Ku Klux Klan, is often ugly, bringing with it unavoidable pain and suffering, or avoidable only at the price of appeasement. Resistance to injustice often itself encounters resistance, and people may undeservedly suffer as a consequence. By subjecting identity groups to fair-minded scrutiny, we come to recognize the good, the bad, and the ugly of identity politics.

Basic questions about the political ethics of identity groups in democracy have been conspicuous by their absence in both academic and popular discourse, for reasons that are worth noting. Because political scientists have tended to treat all organized nongovernmental political actors as interest groups, they have benignly neglected the role that group identity plays in defining and guiding many politically relevant groups in democracies.3 At the other end of the spectrum, far from neglecting identity groups, popular political commentators often subject them to hypercriticism. Some claim, for example, that although interest groups are "an inherent part of the governing process of a democracy," identity group politics--by contrast--"is antithetical to the basic principle of one indivisible nation."4 If one thinks only of identity groups that teach hatred of others, sometimes martyring their members who are willing to kill innocent people, then it is easy to condemn identity politics. But this line of thought misleadingly narrows the notion of identity groups.

Nationality itself is a group identity in the name of which injustices have been both inflicted and resisted. Both slavery in the United States and apartheid in South Africa, for example, have been institutionalized and then opposed in nationalism's name. The nationalisms have differed dramatically in content. Democracies have fought both aggressive and defensive wars by encouraging nationalist impulses among their citizens. People have rallied around a wide variety of nationalist identities in support of tyrannical regimes, yet tyrannies also have been resisted by many nationalist movements.5 Nationalism is part of identity politics, and nations no less than other identity groups should be scrutinized according to considerations of democratic justice.

Identity groups are an inevitable byproduct of according individuals freedom of association. As long as individuals are free to associate, identity groups of many kinds will exist. This is because free people mutually identify in many politically relevant ways, and a society that prevents identity groups from forming is a tyranny. Associational freedom therefore legitimizes identity groups of many kinds.

Many political parties are identity groups, calling upon and cultivating shared identities around ideology, class, religion, and ethnicity, among other mutual recognitions. The myth that superior citizens are independent voters--citizens who do not identify in a stable way over time with a party as a partisan reference group--was an early casualty of survey research in political science.6 "Far from being more attentive, interested, and informed," The American Voter discovered that "Independents tend as a group to have somewhat poorer knowledge of the issues, their image of the candidates is fainter, their interest in the campaign is less, their concern over the outcome is relatively slight."7 The rise of independent American voters in the 1960s led many commentators to declare mutual identification around a political party to be an anachronism, but the revival of partisan political loyalty since the mid-1970s (matching its high 1950s level) underscores the importance of parties as identity groups.8

There now can be little reasonable doubt that mutual identification around a partisan group identity plays a central role in the official institutions of democratic politics. As the literature on party identification amply demonstrates, the relative successes and failures of political parties cannot be adequately understood without attending to the ways in which parties succeed or fail in calling upon and cultivating mutual identification among potential members.9 This book extends the finding that mutual identification is a central part of party politics by examining and evaluating the role of identity groups outside of political parties and the formal political processes of democratic government. Not only within but also outside of the formal democratic mechanisms, identity groups act in ways that both support and threaten basic principles of democratic justice.

Three basic principles are equal standing as a citizen--or "civic equality"--along with liberty and opportunity. The interpretation of these principles varies across democratic views, but the variation does not detract from the fact that civic equality, liberty, and opportunity are core principles of any morally defensible democracy. The broad range of views compatible with these principles all can be called democratic. Identity groups act in ways that both aid and impede democracies in expressing and enacting these principles. The benign neglect of identity groups by political scientists and the hypercriticism of popular commentators are not terribly helpful in understanding or assessing their role in democratic societies.

To assess some of the major issues that identity groups pose for democracy, I consider various illustrative examples, most taken from the United States, a context that provides examples of the major kinds of identity groups. The issues that identity groups pose for democracy are best analyzed in their specific political context, but other inquiries, I hope, will concentrate on other democratic countries. To indicate how ubiquitous and varied identity groups are in contemporary democracies, I also occasionally draw on examples from other democratic societies. These analyses must be even more suggestive, since the context can be less taken for granted.

Consider three political controversies that feature identity groups from three different democracies.

  • In Canada, often considered the home of multiculturalism, the Royal Canadian Mounted Police, for the first time in its history, decided in 1990 to exempt a group from a long-standing rule governing their uniforms. Sikhs were exempted from wearing the wide-brimmed hat that is otherwise a required part of the Mounties' official uniform. This exemption, while not in itself earth-shaking, had far-ranging implications for the accommodation of diverse group identities by public authorities in Canada. The accommodation of the Sikhs in Canada met with six years of protest and was appealed to the Canadian Supreme Court, which refused to hear the challenge, leaving intact the exemption based on identity group membership.
  • In Israel, also in 1990, a group of conservative and orthodox Jewish women petitioned the High Court of Justice to be given the same rights as Jewish men to pray in public. A year earlier, these women had peacefully marched to the Western Wall in Jerusalem, holding a Torah, determined to pray there without male approval. They were attacked by a group of mainstream orthodox Jews who were defending their religious prohibition of women from praying as orthodox men do in public.10 Ten years later, in 2000, the court recognized the women's right to pray at the Wall without abuse by other worshippers. The court also held that the fact that their prayer offends other Orthodox Jews must not annul their ability to exercise their equal rights in public. In response to this ruling, the Israeli Knesset (the unicameral parliament of Israel) introduced a bill that would impose a penalty on any woman who violates traditional Orthodoxy by praying at the Wall.
  • In the United States in 1990, James Dale, an assistant scoutmaster of New Jersey Troop 73, received a letter revoking his membership in the Boy Scouts of America. Dale rose up through the ranks from cub to eagle scout to assistant scoutmaster. When executives in the Boy Scouts learned from a newspaper article that Dale was copresident of the Lesbian/Gay Alliance at Rutgers University, they revoked his membership. Dale brought suit against the Scouts on grounds of discrimination. The New Jersey Supreme Court ruled in Dale's favor on the basis of the state's anti-discrimination law. In a 5 to 4 decision, the United States Supreme Court reversed the ruling and ruled that the Boy Scouts may discriminate based on their expressive freedom as a group to oppose homosexuality.11

These controversies pit politically-engaged identity groups against each other: Canadian Sikhs versus other Canadians, Orthodox Israeli Jews versus "Women of the Wall" (some of whom are also Orthodox), Boy Scouts of America versus gay men, along with other individuals and groups, many of them also part of identity groups, who supported one side or the other in these political battles. The political interests of the major groups in these controversies are intimately linked to their group identities. These identity groups represent only a small fraction of the groups that organize around a mutually recognized identity of their members and pursue a political agenda at least partly based on that group identity. Although non-mainstream groups like Sikhs and gay men are more commonly recognized as identity groups than are mainstream Canadians or the Boy Scouts of America, all are identity groups according to any impartial understanding of the term.

Once we recognize all these groups as identity groups, we are in a far better position to engage in nonpolemical analyses of the problems they raise and the contributions they make in a democracy. Here, in a nutshell, is the dilemma that identity groups present to democracy:

  • Identity groups are not the ultimate source of value in any democracy committed to equal regard for individuals;
  • Identity groups can both aid and impede equal regard for individuals, and democratic justice, more generally;
  • Some identity groups promote negative stereotypes, incite injustice, and frustrate the pursuit of justice;
  • Others help overcome negative stereotypes and combat injustice in contexts of civic inequality and unequal liberty and opportunity;
  • Identity groups can also provide mutual support and express shared identities among individuals whose lives would be poorer without this mutual support and identification.

Why are identity groups not the ultimate source of democratic value? Equal regard for individuals--not identity groups--is fundamental to democratic justice. A just democracy treats individuals as civic equals and accords them equal freedom as persons. If identity groups were the ultimate source of value, then they could subordinate the civic equality and equal freedom of persons (inside or outside the group) to their cause. Accepting an identity group as morally ultimate is inconsistent with treating persons as civic equals who are free to live their lives as they see fit. Living your life as you see fit therefore presupposes that self-appointed groups not impose their identity on you against your will.

Why, then, are identity groups not only legitimate but often also important, indeed even valuable, in democratic politics? First, because identity groups can significantly influence individual identities consistently with individual freedom. Freedom of association is a basic individual freedom. People freely associate (and express themselves) in, among other politically relevant ways, identity groups, which do not comprehensively define the identity of individuals but nonetheless influence their identities in important ways. People value identity groups in no small part because they value relationships of mutual identification and support. Second, numbers count in democratic politics as a legitimate means of exerting political influence, and individuals are most influential in groups. Third, the influence of identity groups includes their ability to play a critical role in combating civic inequalities and unequal freedoms and opportunities of individuals who identify and are identified in groups. Fourth, even when identity groups do not combat injustice, as long as they do not inflict it, they can be valued and valuable for the mutually supportive relationships that they provide their members, which is one reason why associational freedom is an important freedom for democracy to secure.


Identity groups may be organized or unorganized and may be inside or outside the official organs of government. This book concentrates on organized identity groups outside the official organs of government because they are of such great political significance yet neglected by political scientists and treated in a highly polemical way by popular commentators on politics.

What kind of group is an organized identity group?12 Organized identity groups are sometimes thought to be a kind of interest group, but that does not tell us very much since all politically relevant groups that are not part of the official apparatus of government are considered interest groups. Simply to subsume identity under interest groups also tells us too much when it suggests that all groups coalesce around the shared instrumental interests of their members. By separately identifying identity groups, we recognize another way in which individuals relate to politically relevant groups in democratic societies: on the basis of mutual identification.

People often join a group because they share an identity and therefore identify with the people represented by the group and want to support its cause. They usually don't join because they want some instrumental goods from the group that they could not otherwise obtain. Many members of organized identity groups could obtain the same instrumental goods even if they did not join the group. Why, then, do they join? The answer to this question should not come as a surprise (except to those who presume that all rational individuals act--and therefore join groups--out of self-interest). Shared identity is connected to identification with a group and, as a large body of psychological literature demonstrates, is independent of the pursuit of self-interest.13

Identity groups are politically significant associations of people who are identified by or identify with one or more shared social markers. Gender, race, class, ethnicity, nationality, religion, disability, and sexual orientation are among the most obvious examples of shared social markers, around which informal and formal identity groups form. Social markers of group identity also include age and ideology and other mutually recognized features around which groups of people identify or are identified with one another in politically significant ways. Identity groups need not be based on largely unchosen characteristics of persons, such as race or gender.

What distinguishes social markers of group identity is that they carry social expectations about how a person of the particular group is expected to think, act, and even appear. Social markers therefore contribute to the creation of collective identities of both individuals and groups. Collective identities can change over time, and they are also open to varying individual interpretations. Yet because these identities are collective, they may be very difficult for individuals who are so identified to change, even if they do not welcome the identification.

In the paradigmatic case, when a sizable group of people identifies as and therefore with each other, they constitute an identity group. When they act in an organized fashion in politics on the basis of their group identities--whether for the sake of gaining recognition for the group or furthering its interests--they are part of identity group politics. Because people identify as and with other people, they can join a group out of identification rather than (only, primarily, or even necessarily) to pursue their self-interest. (Alternatively, one might say that identification with others becomes part of a person's self-interest, but in saying this we should realize we are stretching the meaning of self-interest so that it no longer signifies an interest in the self but rather any interest of the self, which can even be altruistic. Such a broad definition of self-interest may be more misleading than illuminating.)

It should be obvious that the mutual identification of an identity group--such as gay, lesbian, feminist, Jewish American, Irish American, or African American--does not exhaust the individual identities of its members. Group identification is socially significant but not comprehensive of individual identity.14 Each of these group identifications is also subject to varying interpretations by the individuals who make it their own. A person may make a group identification more or less comprehensive of his or her identity. Individuals have multiple group identifications, and their individual agency modifies their group identifications just as group identifications shape individual agency. Individuals who mutually identify around a social marker often join together in a politically relevant and socially identifiable group.15 These are the organized identity groups that are the focus of this book. (There are also unorganized, or what might be called nominal, group identities that are attributed to individuals in popular culture--the geek, jock, bimbo, and hottie, for example. Nominal group identities, like organized ones, can degrade individuals or elevate them above others. I focus on organized identity groups for the sake of assessing their political importance in democratic societies, and therefore I use the term identity group to refer to organized groups. However, the more philosophical parts of my analysis may apply to nominal identity groups as well.)

An illustrative example of an identity group can help illuminate our understanding of such a group. A social marker of the Women of the Wall--their Jewish egalitarian feminism manifested by their wearing of prayer shawls ordinarily reserved only for Jewish men--is publicly identifiable and it carries a set of social expectations about how individuals within this group identity, as religious, feminist, and activist, are likely to think and act as well as appear. The expectations attached to this group identity do not comprehensively define the mutually associated individuals: a feminist woman is more than a feminist, and a Jewish woman is more than Jewish, and a social activist is more than just that. The social expectations can also change over time and social context. In the United States, Jewish egalitarian feminists have manifested this identity in different ways from the Women of the Wall. More generally speaking, the expectations of feminism have changed over the past century and even today they vary, for example, from religious to secular feminists as well as among religious feminists. Citizens of a single democracy therefore have widely varying views of what it means to be a feminist today. Notwithstanding this variation, members of identity groups associate and are associated with social markers that partly but nonetheless importantly define them.16

When individuals organize together around a recognizable social marker on the basis of their own mutual identification, they are a paradigmatic identity group.17 In other cases, which we can call negative identity politics, people are identified against their will by others by being given attributes of a particular kind of person, such as a "dirty Jew" or a "nigger," even though they disdain being so identified. (They might even prefer not to be identified at all as Jewish or black.) Historically, negative identity groups have rarely remained only or even primarily negative. When individuals are stigmatized because they are identified with a group, if they have the freedom, they often also publicly organize as a positive identity group to protest and transform their social markers from negative to positive ones. Since there are identity groups in the business of both negative and positive identity politics, to call a group an identity group is neither to praise nor to criticize it. The National Association for the Advancement of Colored People (NAACP) and the Ku Klux Klan (KKK) are both identity groups, and praise for one is often associated with criticism of the other. The KKK is responsible for a pernicious kind of negative identity politics, which the NAACP exists in part to overcome. The NAACP's identity includes joining blacks and whites together in mutual recognition of their common humanity and equal rights. A person need not be a member of the NAACP to benefit from the instrumental aim of the organization--the achievement of equal rights. Mutual identification around its egalitarian ideology helps explain why membership in the NAACP is robust and not irrational from an individual perspective.

Individuals identify in groups around their gender, race, ethnicity, nationality, class, sexual orientation, age, physical ability or disability, and ideology. These social markers are open to widely varying interpretations and also vary over time. The social markers of paradigmatic identity groups--religion and culture in the case of the Canadian Sikhs, gender and religion in the case of Women of the Wall, sexual orientation in the case of gay men--are sufficiently differentiated to enable both insiders and outsiders to distinguish the group from others on that basis.

Mutual identification draws individuals together to identify with politically relevant groups, sometimes in pursuit of instrumental ends. Mutual identification therefore does not preclude a group's pursuit of instrumental ends for its members and nonmembers. Identity and interest are often closely intertwined.18 Identity is an effective means of a group's organizing with the aim of pursuing instrumental goals, but identity groups are not reducible merely to instruments for such pursuits. Because interests have so often been the focus of political analysis, it is important to shift the focus in order to understand and evaluate the role that identity plays in democratic politics. Identity groups are sites of mutual identification and the pursuit of instrumental interests that are informed by mutual identification. Moreover, many identity groups pursue interests (both morally good and bad) for nonmembers as well as members, and therefore we cannot define them as serving only--or even primarily--the self-interests of their membership.


The time is long overdue for understanding the distinguishing features of interest and identity groups and how they interact. An interest group organizes around a shared instrumental interest of the individuals who constitute the group without any necessary mutual identification among its members. The members are not drawn to the group because of their mutual identification; they are drawn to it because they share an instrumental interest in joining the group.19 The political action of the group reflects the social identification of its members. Even when individual members dissent from some of the group's actions, the group's actions reflect back on the identity of its members because the members identify with the group (even when they dissent from some particular actions of the group). Whereas the defining feature of an identity group is the mutual identification of individuals with one another around shared social markers, the defining feature of an interest group is the coalescing of individuals around a shared instrumental goal that preceded the group's formation.

The pursuit of instrumental interests by identity groups must not be taken to mean that these interests preceded the mutual identification and that therefore the group is really an interest group, not an identity group. The contrary is often the case. For example, without their mutual identification as Israeli Jewish feminists, the Women of the Wall would not have an instrumental interest in breaking down the barriers that exist for Israeli Jewish women who wish to worship in public. This causal connection between group identity and individual interests is one reason why we should not be selectively skeptical of using the idea of group identity to understand the way people act in democratic politics. Speaking of the term "identity," one scholar rhetorically asks: "Do we really need this heavily burdened, deeply ambiguous term?"20 If we want to understand an important set of political phenomena in democratic politics, the answer turns out to be yes; the term "identity" is useful and illuminating. To be sure, "identity" has various meanings, but so does the term "interest," and "interest" is far more ambiguously invoked--to the point of meaninglessness--to describe each and every political action taken by apparently reasonable people in democratic politics.

What can group identity explain better than a ubiquitous invocation of interest? The example of African Americans who could pass as whites but chose not to, even in the midst of rampant racial prejudice, confirms the importance of group identity in informing individual interests and actions. For African Americans who could pass, it was the mutual identification with other African Americans that explains their interests in the group, not vice versa. Just as some African Americans could satisfy their self-interest by distancing their interests entirely from the group and living as white, so too can members of many other disadvantaged ethnic groups who are able but unwilling to pass. Most people, of course, cannot pass for a different gender or white rather than black, hearing rather than deaf, or vice versa. The existence of individuals who can pass but who choose to publicly identify with a less advantaged group is important to note because it demonstrates that group identification is not derivative of the self-interests of the members of the group.

The phenomenon of identity groups is therefore both real and distinctive. It expresses the robust idea that group identity provides a basis for individuals to develop a sense of their own interests in democratic politics. This idea has been neglected because "self-interest" is so often taken to be primordial rather than informed by, among other things, group identity. Identity groups encourage people to join by orienting themselves around some mutual identification that is broader than the specific interests that they are pursuing at any given time. The greater the role that identity or solidarity plays in attracting, retaining, or mobilizing members in political efforts, the more a group is distinctively an identity group. The greater the role played by the pursuit of shared instrumental interests of individuals--regardless of their group identity--in attracting, retaining, or mobilizing members in political efforts, the more a group is distinctively an interest group. One of the neglected issues that identity group politics therefore poses for democratic society is the way in which recognition of interests often follows from group identification rather than being given simply by the pre-existing interests of individuals apart from their group identifications.

As the distinction between a paradigmatic identity group and an interest group suggests, the mutual identification of individuals and the collective pursuit of shared instrumental goals are not mutually exclusive. Identity groups are so-called because members join, support the group, and act out of mutual identification. Identity groups may of course act collectively in pursuit of an instrumental interest of the group. The mutual identification and the collective pursuit of an instrumental interest are mutually reinforcing. When identity groups pursue instrumental interests of their members, they encourage more people to see their identities as bound up with a group. Conversely, interest groups that originally formed to pursue a discrete issue may solicit more members by orienting themselves around some mutual identification that is broader than the original issue that they organized to pursue.

In paradigmatic form, identity group politics is bound up with a sense of who people are, while interest groups politics is bound up with a sense of what people want. This distinction is as clear as it can be in theory, and its theoretical clarity can also help make more apparent the close connection between group identity and people's sense of their interests in political practice. How people identify themselves--the distinctive organizing feature of identity groups--importantly affects what they want, the distinctive organizing feature of interest groups. Democratic politics is bound up with both how people identify themselves and what they therefore want. For this to be true, identity groups must be politically relevant and worthy of our careful attention.

People's identification with one another influences their sense of what they want. Moreover, individuals who identify with others are better able to organize politically, and organized groups can be far more politically effective than an equal number of unallied citizens.21 Taken together, these two observations answer the question of why group identity matters so much in democratic politics. Group identification--whether it be focused on gender, race, religion, sexual orientation, ethnicity, nationality, age, disability, or ideology--provides people with motivating reasons of mutual identification to organize politically. Since mutual identification informs people's sense of their own political interests, group identity and collective interests are often mutually reinforcing in democratic politics.


In order to evaluate the broad phenomenon of identity groups in democracy, we must avoid defining identity groups so narrowly as to include only those groups that aid democratic justice or only those that impede it, but include both. By beginning with a nonpolemical definition, we are better able to understand what inclines identity groups, whether intentionally or unintentionally, toward or against democratic justice.

Mutual identification is central to the raison d'etre of identity groups. When mutual identification entails putting considerations of group identity above considerations of justice--for example, by preferring people of one's own "kind" above others even when matters of justice push in the opposite direction--identity group politics is morally suspect. Putting a shared identity above considerations of justice means elevating what is not morally fundamental above what is. Group identity is not morally fundamental. If it were, as I argued above, groups could subordinate just treatment of individuals (member and nonmembers alike) to the identity and interests of their group. This is a description of tyranny over individuals, whether by a minority or a majority.

Identity groups should be suspect whenever they encourage their members to ignore considerations of justice for the sake of supporting the group, thereby disregarding any injustice of their cause. Putting a group above justice is a common phenomenon. Call it "making morally too much of group identity." The phenomenon is so common as to tar the entire landscape of identity group politics. Not all identity groups act this way, and few do all the time, but when any do, their moral myopia needs to be exposed. The source of the myopia is thinking, feeling, and/or acting as if group identity is an overriding public good, which takes precedence over avoiding injustice or pursuing justice for individuals regardless of their group identity. This myopia is the source of ongoing injustice in democratic societies and world politics more generally.

Even identity groups like the NAACP that act overwhelmingly for the good sometimes succumb to the moral mistake of making too much of group identity. When Clarence Thomas was first nominated to the Supreme Court, the NAACP hesitated to oppose his nomination because he was black, despite the fact that his judicial philosophy was inimical to that of the NAACP. Although, after a short period, the NAACP strongly opposed Thomas, it lost important momentum in the early days after the nomination, and its hesitation in opposing Thomas made his approval by the Senate more likely than it otherwise would have been. Making too much of Thomas's race gave conservative Republicans an easier success than they otherwise would (or should) have had. At the time, President Bush publicly claimed that race had nothing to do with his choice of Thomas, but that claim is not credible in light of available alternatives to the Thomas nomination. What is credible is that race was related to the nomination not because Bush himself thought it should be but because he thought others thought it should be, and those others would not normally support a nominee with Thomas' conservative qualifications. Group identity can be used--whether sincerely or cynically--in ways that are ethically suspect but politically effective.

Identity groups are ethically suspect when they elevate group identity over considerations of justice. Critics suggest three reasons why identity groups will systematically subordinate justice to their cause. First, they say that identity is harder to compromise than interest, and democratic politics depends on compromise. Second, they say that identity politics is inherently sectarian and therefore inimical to egalitarian reform. And third, they criticize the involuntary basis of identity groups.

Each of these criticisms merits closer scrutiny. Even if identity groups on average are less compromising than interest groups, it is not clear what follows, since there is no more reason to commend or criticize identity groups as a whole than interest groups as a whole. The differences in political behavior among and within identity groups are so striking as to call into question the usefulness of this generalization about the entire phenomenon of identity groups in democratic politics. The critical premise is that by virtue of being an identity group, the group must be less compromising than an interest group because people do not compromise their identities. By contrast, they do compromise their interests. This reasoning, however, is misleading because the aim of many organized identity groups is to positively express the identity of their members while pursuing various instrumental interests of the group. In the pursuit of instrumental interests--such as equal pay for equal work for women--identity groups can be just as compromising--or not--as traditional interest groups that organize for minimizing taxes, for example.

One problem with criticizing identity groups for being uncompromising is that it does not follow that identity groups are more harmful to democracy than interest groups, even if they are less compromising. Everything depends on the nature of the issue and--in the case of identity groups that are committed to just causes such as equal rights for minorities and women--how effectively groups that are not organized around identity would otherwise fight for these causes. Democratic societies can benefit from the presence of noncompromising groups on some issues such as equal rights for women and minorities. At the same time, there are plenty of reasons to criticize the uncompromising positions of those identity groups that push for benefits for their group regardless of the merits of the case and even at the cost of denying civic equality, equal freedom, or opportunity to members of other groups. The same criticisms, however, can be directed at uncompromising interest groups--such as some corporate organizations--when their causes are similarly unfriendly to democratic justice.

When we pick prominent examples of identity and interest groups and ask whether one is more compromising than the other, the generalization seems even more doubtful. Is the National Rifle Association (NRA), a large interest group, more compromising than the NAACP, a large identity group, on the issues that concern its members? One cannot credibly claim that the racial identity of the NAACP makes it uncompromising while the antigun control interest of the NRA makes it compromising. Identities and interests are interpreted by groups over time, and neither identity nor interest as the basis of an organized group lends itself more or less to political compromise. Everything depends on the content of the identity or interest along with the context within which it is pursued.

Uncompromising critics of identity groups also say that they block progress toward greater economic equality in democracies.22 Is this true? The politics of multiculturalism often seems preoccupied with supporting particularistic identities and interests and therefore is either oblivious or hostile to egalitarian principles such as Rawls' difference principle (which requires income and other primary goods to be distributed so as to maximize benefits to the least advantaged).23 The critics are rightly critical of relatively privileged identity groups lobbying for greater recognition of their own interests in disregard for those of far more disadvantaged groups. Even equal pay for equal work for professional women or minorities, a just and worthy cause, is surely not as high a moral priority as a living wage for all workers, regardless of their gender, race, or other group identity.

Has identity politics as a whole actually distracted democracies from these more urgent causes? Women who mobilize for equal pay for equal work also are likely to be among the same people who avidly support a living wage for all workers. The implication of egalitarianism is not a critique of identity groups per se but a critique of those that present their claims solely in terms of their particularistic identities and not in egalitarian terms. This criticism, however, must also be levied against interest groups that are as particularistic as identity groups in this respect.24 Both identity and interest groups are suspect in this regard, and so is every other particularistic group that conceives of its collective interests as self-justifying rather than in need of taking other people's interests into account.

In the absence of identity politics, would democratic politics, as some critics claim, be more egalitarian in redistributing income and wealth and supporting equal opportunity? Alas, there is no evidence for this counterfactual claim. To the contrary, one credible explanation for the rise of identity politics in the United States in the late twentieth century is the failure of conventional interest group politics and government to concern themselves with the civic equality, equal freedom, and opportunity of disadvantaged women, people of color, and the disabled. Identity groups arose in the United States representing all of these groups--and more--and they have succeeded in bringing far greater attention to these egalitarian causes and in effectively lobbying for more progress along egalitarian lines than would have occurred without them. Many of these groups, moreover, defend the application of universal and egalitarian principles--nondiscrimination, equal pay for equal work, equal opportunity, civic equality--to correct long existing injustices that interest group politics have passed by.

The record of success of egalitarian identity groups in the United States falls far short of the aims of democratic justice, but this failure of American politics long preceded the identity politics targeted by egalitarian critics. The critique of identity politics is at best partial and polemical when directed at organized feminists, people of color, ethnic minorities, and gays, lesbians, and bisexuals rather than at the institutions, ideologies, and interest groups that actively oppose any egalitarian reforms. "The politics of identity," Todd Gitlin writes, "struggles to change the color of inequality," meaning that the affluent are now black as well as white, but inequalities of income and wealth are no less (and in fact even more) than when only whites were affluent.25 This claim might appear less forceful to many as a critique of identity politics if it instead read "the politics of identity struggles to eliminate the gender of inequality." Eliminating the gender and color of inequality is no small moral accomplishment, even if it is not enough. It is also untrue that identity politics struggled only to change the gender and color of inequality. Feminists and many African American, Latino American, and other identity groups continue to lobby to equalize work, income, and educational opportunities for disadvantaged members of their groups who still encounter discrimination in the marketplace, workplace, and educational institutions. Gay, lesbian, and bisexual identity groups also have struggled--although not always successfully--for equal sexual freedom as well as equal opportunity in workplaces and institutions, including the military.

Affirmative action is another target of critics of identity politics who blame such politics for dividing disadvantaged groups, particularly whites and nonwhites, into "internecine warfare" to gain entry into selective colleges and high-level social offices.26 If affirmative action changed the gender and color of educational inequality by admitting more African Americans and women to selective colleges and universities, the change should not be disparaged as ethically insignificant. Revealingly, far fewer critics disparage the use of affirmative action for women, which changed the gender of inequality, than disparage its use for African Americans. Changing the gender and color of inequality in higher education, while far from sufficient to satisfy democratic standards of civic equality, is not morally insignificant. Affirmative action in higher education, we learn from the careful empirical investigations of William Bowen and Derek Bok, has done considerably more than change the color of inequality.27 It has improved the quality of education as reported by both whites and blacks who attend selective colleges and universities and brought more black professionals into community service positions in their communities.

Many of the same individuals and identity groups, moreover, defend not only affirmative action but also egalitarian causes such as raising the minimum wage, increasing job opportunities, improving health care and public health, and making taxation more progressive.28 Following the lead of William Julius Wilson, many African American identity groups also give higher priority to improving economic opportunities for the least advantaged, regardless of their color, and ally themselves with labor and other economically disadvantaged identity groups. But success on these egalitarian fronts in the United States in recent decades has not been easy, to say the least. And Wilson does not make the mistake of opposing affirmative action as if it had been responsible for blocking these other, more urgent egalitarian reforms.

Other things being equal, egalitarian politics has been more successful in those democracies where working class identity has been more politically significant. The reason not to criticize identity politics per se is not because group identity is good in itself--it is not--but because group identity can make organization around almost any cause easier. Egalitarian critiques of identity groups often fail to recognize that some of the most successful egalitarian movements in modern democracies have been organized around class identity, mobilizing large numbers of citizens to ally with one another to support welfare rights, minimum wages, and other economic redistributions. Although the most successful political alliances are not purely made on class grounds, neither do they succeed without strong support from class-based organizations (including class-based parties like Labor in many democracies that build on class identity). Only by dismissing identity groups based on class can one say that identity politics is an enemy of egalitarian politics. Perhaps this criticism can be recast and redirected against identity politics that lacks a significant class basis. If identity politics neglects class and concentrates on group identities such as race and ethnicity, as is the case in the United States, then it is less likely to pursue an egalitarian agenda.

To blame identity groups for the failure of egalitarianism in the United States, however, distorts the larger picture of resistance to egalitarianism among a citizenry that overwhelmingly identifies itself--and with one another--as middle class rather than poor or working class. In the last three decades of the twentieth century, while the top four quintiles (80 percent) of Americans have grown better off, the bottom quintile (20 percent) has grown worse off. Even if identity group politics has not helped improve this problem, the critics of identity politics have not made a credible case that in its absence, life would now be better in the United States for the bottom quintile, and the country would be further along the road to civic equality, equal freedom, and opportunity for all. Have interest groups done any better? At one time, unions did better, but their decline, and their relative lack of success in the United States compared to in many European countries, is attributable in no small part to the weakness rather than the strength of identity politics: the considerably weaker identification of American workers with a working class group identity than their European counterparts. A working class group identity can help motivate individuals to join labor movements and identify their interests with this group rather than with the ubiquitous middle class, the group with which so many relatively poorly paid workers in the United States identify themselves.

When ridding democracy of identity groups is viewed as the path to egalitarian justice, we might paraphrase Madison's comment about factions in Federalist 10: It would be as much folly to try to abolish what causes identity groups to form--particularistic group identities of individuals and freedom of association--as "it would be to wish the annihilation of air, which is essential to human life, because it imparts to fire its destructive agency."29 Far from being antithetical to representative democracy, identity politics is an important manifestation of individual freedom within it. Far better to address the bad effects of identity group politics in a way that is consistent with free association than to try to abolish identity groups. In significant instances, identity groups have been effective in addressing inequality. More often than not, feminist politics and African American identity politics have been prominent sources of movement in the direction of civic equality, equal opportunity, and other egalitarian dimensions of democratic justice. In some instances, however, the efforts of these identity groups have been misdirected, in a direction away from democratic justice.

The ideology of identity groups is so diverse as to defy the generalization that it is either antiuniversalist and antiegalitarian, or the reverse. There are certainly many identity groups whose philosophies and practices work counter to democratic justice, and they are worthy of political criticism. Brian Barry directs his relentless critique of identity politics against illiberal groups that oppose or undermine the protections of equal rights that liberal democracies should provide.30 So directed, the critique is correct. But Barry chooses to ignore the fact that many identity groups, including many of those that support affirmative action in the United States, struggle politically for precisely the reverse: equal effective rights by means of better economic and educational policies.31 To be convincing to open-minded people, criticism of identity group politics needs to be discerning. Identity group politics as a whole cannot fairly be said to undermine a politics of redistribution.32 Many identity groups--most feminist groups, for example--do precisely the reverse: they strongly support a politics of redistribution, and they ally with many other identity groups that do the same. On the other hand, many other identity groups are highly sectarian and inegalitarian.

Better than both sectarian identity politics and sectarian interest group politics, as egalitarians have long suggested, would be a universalistic and united egalitarian political movement to raise the standards of all disadvantaged individuals, regardless of their cultural, ethnic, gender, racial, or national identities. The idea that such a universalistically based movement will be forthcoming in the absence of identity politics is doubtful in light of American history and the relative weakness of working class identification. The history of egalitarian reform in many European and Latin American democracies has been based in no small part on a politics of working class identification that is far weaker in the United States, where income and wealth inequalities are correspondingly greater. politics per se: an alternative that consistently follows from the egalitarian critique of identity politics is a democracy in which more individuals ally together politically as free and equal persons. Such a justice-friendly alliance based on mutual identification as egalitarian democrats would be yet another form of identity group politics. Democracies need more people to ally together politically and defend just causes, whether out of mutual identification as free and equal persons, or out of their commitment to justice, or both. (The identification and commitment are fully compatible.) The need for a politics that is more conducive to justice does not come any closer to being met, however, when critics condemn identity groups per se.

Critics of identity groups also focus their attention on the involuntary basis for some of these groups. The members of ascriptive groups of a particular race, gender, or nationality generally do not have a choice of being identified with the group. Involuntary ascription raises a set of distinctive problems concerning identity groups in democratic contexts that is worthy of extended analysis. But the involuntary nature of some group identities cannot be the basis for a wholesale critique of identity groups or identity politics for many reasons, the most basic of which is that involuntary identity groups are only a subgroup of identity groups and not necessarily the most prevalent or powerful subgroups at that. Memberships in many identity groups--most religious groups and all voluntary associations whose members are drawn to them out of mutual identification (in some cases because of their ascriptive identities)--are voluntary, as are memberships in many interest groups.

Nor would an involuntary basis of membership suffice for criticizing a group, as the defense of labor unions (some of whose membership is involuntary) suggests.33 The membership of most democratic societies is involuntary: Most citizens do not have the effective freedom to pick up and leave and settle in another country that would accord them civic equality and equal freedom. On its face, the left-handed invitation "love it or leave it" presents a false choice for most citizens, who cannot leave their country any more easily than they can change their gender (which, after all, is no longer impossible). The absence of an effective freedom to leave one's country is not a sufficient ground for criticizing an otherwise just democracy. But it is a good ground for criticizing democracies that are unjust, as all democracies today are, in not according civic equality and equal freedom and opportunity to all their members. Something similar can be said of identity groups based on race, gender, and ethnicity, for example, the fact that they are involuntary is not a sufficient reason to criticize them, but the absence of an effective freedom to exit some identity groups renders them suspect in the same way that societies are suspect vis-à-vis their citizens who are not accorded civic equality and equal freedom and opportunity within the group. When identity groups are voluntary, for their members, as fraternal and soraral clubs like the Masons are thought to be, they have at least a partial moral defense of the way they treat their members--based on informed consent--that is absent from involuntary groups of all sorts. The fact that democratic societies are not voluntary associations may also make it all the more important for other groups within those societies to be voluntary. This issue is critical to considering both cultural identity groups and ascriptive identity groups, which are the topics of chapters 1 and 3, respectively.

Religious groups are also identity groups, and their politics is an important part of identity group politics in modern democracies. Many critics of identity politics exempt religious groups from criticism, and treat them as deserving of special treatment. In chapter 4, I ask whether religious identity groups should be treated with special consideration in democracies. I argue that they should not, but the reason many people may think that they should is instructive. The ultimate ethical commitments of individuals--which may be religious or secular in their source--are an especially valued and valuable part of individual identity. The contemporary shorthand for those ultimate ethical commitments--conscience--instantiates the identity of individuals as ethical persons. As such, conscience should be treated with a degree of deference in democracies out of respect for persons as moral beings, which is a basic principle of democratic justice. A degree of deference does not mean that conscience trumps legitimate laws. It means that democratic governments may legitimately treat conscience as special in democratic politics when so doing does not violate anyone's basic liberty, opportunity, or civic equality. By recognizing that conscience can be either secular or religious, democratic governments avoid discriminating either in favor of or against religious citizens.


Identity groups are suspect when they elevate group identity above justice, but they often fight against precisely this problem, as it is often unconsciously ingrained in discriminatory political practices that are well established and therefore taken for granted. Identity groups need to be assessed by the same standards that one would apply to any groups that make political claims and exert political influence in democracies. I apply the democratic standard of civic equality, broadly understood to include equal freedom and opportunity for all individuals. There is no ethically neutral place to evaluate the contribution of identity groups to democratic societies, nor would a neutral place be desirable if it were available. When I use the term democracy, it signifies a political commitment to the civic equality of individuals. A democracy also can and ideally should be a deliberative democracy, offering opportunities for its citizens to deliberate about the content of democratic justice and to defend their best understanding of justice at any given time.

A just democracy therefore respects the ethical agency of individuals, and since individuals are the ultimate source of ethical value, respect for their ethical agency is a basic good. Ethical agency includes two capacities: the capacity to live one's own life as one sees fit consistent with respecting equal freedom for others, and the capacity to contribute to the justice of one's society and one's world. All democratic theories that take ethical agency seriously also honor three principles in some form. One is civic equality--the obligation of democracies to treat all individuals as equal agents in democratic politics and support the conditions that are necessary for their equal treatment as citizens. A second principle is equal freedom--the obligation of democratic government to respect the liberty of all individuals to live their own lives as they see fit consistent with the equal liberty of others. A third principle is basic opportunity--the capacity of individuals to live a decent life with a fair chance to choose among their preferred ways of life. Of course other principles may be considered basic by other theories, but for purposes of assessing the place of identity groups in democracies in this book, civic equality, equal freedom, and basic opportunity serve as critical standards. Since civic equality also requires many equal freedoms and basic opportunities, I refer to civic equality as a shorthand throughout the book of a principled basis for assessing the relationship between identity groups and democratic politics.

Why is it important to evaluate rather than just to describe the role of identity groups in democracies? Much more work needs to be done in describing the role of identity groups in democracies, far more than I can do in this book. But no matter how thoroughly we analyze identity groups, our discussion will be incomplete if it is merely descriptive. Description alone is not even sufficient to describe completely the role of identity groups in democracies, and it is not even the prior task to considering how identity groups at their best can contribute to democratic justice and how at their worst they impede its pursuit. This, after all, is part of a description of the role of identity groups in democratic politics.

To describe identity groups in a value neutral way would be to misdescribe and misunderstand not only identity groups but the nature of democracy. Democracies are not neutral political instruments; they are worth defending to the extent that they institutionalize in politics a more ethical treatment of individuals than the political alternatives to democracy, which range from benevolent to malevolent autocracies and oligarchies. Some identity groups aid democracies in institutionalizing more equal treatment of individuals and others impede it. A critical part of a description of the role of identity groups in democracies must therefore be to develop a language that helps us to understand their role in both aiding and impeding the pursuit of democratic justice.

Critics who do not labor under a false sense of value neutrality still may shy away from judging identity groups by democratic standards because they realize that democratic standards are themselves often contested. But so, too, are empirical descriptions of groups, and therefore avoiding controversy is not a good reason to seek value neutrality. Judging from the depth of controversies over historical accounts that are ostensibly empirical, it is not even clear that disagreements about ethical values arouse deeper passions than empirical disagreements (for example, about the causes and effects of ending slavery or instituting affirmative action).

Reasonable contestation or challenge is something to be encouraged in democratic politics, out of respect for individuals as ethical agents. Mutually binding laws and policies should be justified to the extent possible for ethical agents; efforts at justification, even if unsuccessful (as they often are) at achieving agreement, express mutual respect among persons as civic equals. Whatever specific principles we defend in democratic politics to assess identity groups, we therefore can also defend democratic deliberation about those principles and their application out of a commitment to mutual respect among persons, which itself is a way of treating people as civic equals.34

The principles invoked in this book to assess identity groups--civic equality, equal freedom, and basic opportunity--are defended by a wide range of democratic theories. The three general principles are still subject to disagreement in their application, especially in difficult cases. Does civic equality, equal freedom, or basic opportunity, for example, require the Canadian Mounties to exempt the Sikhs from its uniform policy? Assuming that both religious freedom and separation of church and state are violated by the inability of Israeli women to pray as Israeli men do at the Holy Wall in Jerusalem, what other features of Israel's nonseparation of church and state violate the equal religious freedom of Israeli women who identify as Jewish? James Dale's challenge to the Boy Scouts raises yet another kind of principled question concerning the competition between two principles. Which is more basic to a democratic justice: freedom from discrimination on the basis of one's sexual orientation or freedom of expression for voluntary associations? What difference does the nature of the voluntary association make in how we compare what is at stake on both sides?

Some controversies concerning identity groups that I draw upon are not hard cases to assess whereas others are. It is important to consider both kinds of cases in order to understand and evaluate identity groups in democracies. When a case is hard, however, and there is reasonable disagreement among affected parties about how democratic principles should be interpreted and applied, there is also a special need to move beyond basic substantive principles and call for deliberation within as inclusive a group as possible of the people who are significantly affected by the decision. Substantive principles inform democratic deliberation, but they do not take its place. Conversely, deliberation does not take the place of substantive principles; it would be a politically hollow (and ethically senseless) exercise if it were not substantively informed.

Liberty, opportunity, and civic equality are defensible on grounds offered by almost all democratic theories, which converge in support of these principles from different starting points.35 Yet some people identify democracy far more simply with one single principle, majority rule. Why join democracy to a defense of any principles other than majority rule? First, because majority rule is not a principle by itself. It is a rule of procedure that cannot possibly define a defensible democratic politics, since majority rule can be used by oligarchic decision makers. Any democracy that is defensible on ethical grounds--grounds that respect the ethical agency of all persons--must do more than establish order through a decision-making procedure, since order alone does not respect the ethical agency of all persons.

I use the word "democratic" in this book as a concept of political ethics to signify a public commitment to treating individuals as ethical agents. (Democratic justice does not view individuals as atomistic individualists; it views them as ethical agents, which is quite a different matter.) A just democracy helps secure for all persons the conditions of civic equality, equal freedom, and basic opportunity, principles that are preconditions of a fair democratic process but are also valuable in their own right as expressions of the freedom and equality of individual persons as ethical agents.

A democratic state that respects individuals as free and equal persons does its best to secure civic equality for every person. There is room for deliberative disagreements about what counts as civic equality, but there is also a broad range of reasonable agreement possible among democrats. For example, democrats today count among the equal liberties freedom from slavery, serfdom, forced labor, and other forms of subordination of persons and the correlative freedoms of expression, conscience, assembly, and association. Basic opportunities are broadly agreed upon to include adequate schooling, subsistence, and nondiscrimination in the distribution of educational and career opportunities. Civic equality refers to both a set of political rights of equal suffrage and political participation in a fair competitive process of democratic decision making and a set of civil rights including due process and equal protection by the laws.

Defending these principles that are widely shared by democrats does not require anyone to be what John Rawls calls a "comprehensive liberal" because these principles do not encompass the entirety of a moral philosophy or a cultural identity. Nor does the democratic perspective that informs them encompass the entirety of any moral philosophy or cultural identity. Many democrats committed to civic equality for all persons are, to use Rawls's term, "political liberals." There are many versions of political liberalism other than Rawls's theory of justice, just as there are many versions of comprehensive liberal and nonliberal perspectives. Although the perspective taken in this book may be seen as a kind of political liberalism, it is compatible with many comprehensive philosophies as well.

Democracy, as I understand it, takes all persons, regardless of their ascriptive identities, as deserving of equal political regard or respect. Because a democratic view of politics is committed to equal respect for persons regardless of their particular group identities, it provides a potentially critical vantage point for assessing identity groups. While the political landscape is often divided between those who assume that all identity groups violate democratic principles and those who assume the reverse, neither assumption is warranted by this analysis. Distinguishing among identity groups is critical not only to understanding but also to assessing their role in democratic societies.


Four kinds of identity groups--cultural, voluntary, ascriptive, and religious--are worthy of separate consideration because each highlights a different set of ethical issues raised by the presence of identity groups in democracies and the ways in which they can either aid or impede democratic justice. Although the four kinds are not mutually exclusive, by focusing on culture, choice, ascription, and religion we can more readily examine the most important issues revolving around the relationship between group identity and democracy. I therefore have devoted a chapter to each kind of identity group.

Identity groups call attention to both the extent and the limits of expression of free identity, and the tension between free identity expression and its limits is a central issue in chapters 1 and 2. Chapter 1 considers cultural identity groups. Culture is the most common category around which controversies over identity groups gravitate. This is because of a tendency to consider every identity group as representative of a culture and culture as constituting its members' identities. Although culture can be considered the universal glue that unites or divides people into identity groups, so broad a definition verges on the vacuous. Political theorists of culture are more careful in specifying what they understand a cultural group to be. Cultural groups, they say, represent ways of life that are comprehensive or encompassing, terms that are used interchangeably to refer to the provision by culture of a context of choice that determines the range of what is feasible for its members.36 Cultural groups also can give their members a sense of security and belonging. In chapter 1, I question the comprehensiveness of culture and pursue some corresponding concerns about the claims of cultural identity groups vis-à-vis both members and non-members. The concerns follow from democratic principles such as equal freedom and civic equality to which many theorists of culture themselves subscribe.

At the other end of the associational spectrum from cultural identity groups are voluntary associations. They are so far at the other end that they are often not even considered to be identity groups. Voluntary associations offer the opposite of a comprehensive context of choice to their members. Each association offers one among many options, along with disassociation itself, within an associational diverse context. Chapter 2 focuses on the voluntary groups that inhabit a democratic landscape, which de Tocqueville described in his portrait of democracy in America as of "a thousand different types--religious, moral, serious, futile, very general and very limited, immensely large and very minute."37 Contemporary theorists of choice eloquently defend free association in democracy, reminding us that "to be free, to live as one likes, includes associating on one's own terms, which means engaging in voluntary relationships of all sorts."38

However essential they are to the fabric of a democracy, voluntary associations are far from problem-free. They pose a problem that flows from the very nature of the freedom that they represent in non-ideal societies that have been ridden by systematic social prejudices. Freedom of association entails the freedom to exclude people from one's association, yet prejudicial exclusion of people because of their gender, skin color, or other unchosen characteristic exacerbates civic inequality in many forms, including unequal freedom and opportunity. Prejudicially blocked entries into voluntary associations may therefore be considered unjust. Yet the freedom to form an exclusive group and the freedom to join one are both valued freedoms. Whichever way a democracy resolves this conflict between the freedom to join and the freedom to exclude, the freedom of some people to express their identities as they see fit will be limited by the freedom of others. How a democracy resolves the conflict of freedoms is critical to the value of voluntary groups vis-à-vis democratic justice, which is the subject of chapter 2.

In chapter 3, our attention turns to those organized groups that explicitly bring people together based on unchosen characteristics: mutual identification by ascription. Many identity groups organize around unchosen social markers such as gender, color, ethnicity, sexual orientation, and disability. Groups that organize around ascriptive identities are themselves enormously varied in their relationship to democratic justice. Some are justice-friendly, actively promoting democratic justice by advocating equal freedom and opportunity for individuals who are disadvantaged because of their ascriptive identities. Other ascriptive identity groups--like the National Association for the Advancement of White People, which arose in response to the NAACP--impede the very same cause. Even with regard to justice-friendly ascriptive groups, special issues arise. Some people claim that women, African Americans, and other people of color have special moral obligations to their ascriptive groups because if they do not contribute to the cause they will be free-riding on the efforts of justice-friendly groups. Although these special obligations are not considered legally binding, they are publicly defended as morally incumbent on individuals, and they therefore differentially burden some individuals who are identified with certain ascriptive groups and not others.

An alternative view--the identification view--suggests that all individuals, regardless of their ascriptive identities, can live morally better lives by identifying with disadvantaged people and contributing to just causes out of that identification. Justice-friendly groups make it possible for individuals to contribute to just causes without undue sacrifices in their lives. The identification view has the ethical advantage of applying to everyone, not only members of disadvantaged groups, and it also extends beyond any single democracy. Identifying with others and contributing to just causes can help make everyone's lives better, morally speaking. This kind of identification--for the sake of our living morally better lives and contributing to democratic justice--needs more prominence in democratic cultures if democracies are to have their best chance of becoming more just.

Religious identification is among the most prominent grounds for identification in modern democracies, so much so that religious identity, the subject of chapter 4, often calls for special treatment by democratic governments. Religious identification seems to be one of the strongest and most persistent sources of mutual identification known to humanity. Perhaps this is because religions hold out the hope of immortality (although not all religions do so) or because religions are often also "thick" cultures that mark the human life cycle with holidays and sacraments with which families and communities can mutually identify. Recognizing how powerful religious identity can be, many people--religious and secular alike--have been concerned to separate the power of organized religion from the power of the state. Long before anyone coined the term identity politics, religious identity groups, or "churches," were both honored and feared as potent political forces.

In order to protect churches and states from one another, some democracies have tamed the political power of both by a settlement that I call "two-way protection." Two-way protection is committed to protecting the religious freedom of individuals in exchange for protecting the democratic state from the political power of churches. By contrast, one-way protection aims to protect religion from the state but not protect the state from religion. Defenders of two-way and one-way protection disagree as to whether and how religious identity is special, and with what political implications. Chapter 4 asks whether religious identity is special in democracies for the most commonly suggested reasons: its truth value, its contribution or threat to the public good, or its constitution of conscience, the ultimate ethical commitments of sincere believers. I conclude that religious identity is not special for its truth value, its contribution or threat to the public good, or its unique constitution of conscience.

The conclusion that religious identity is not special, however, has a significant twist. In considering the claims of religious conscience, I conclude that it is not special, although conscience, more generally understood as the ultimate ethical commitment of individuals, is special. Conscience has some special claim to be considered by democratic governments because democracies are supposed to be committed to respecting the ethical agency of individuals. Since religious identity is not the only source of binding ethical commitments, democratic governments cannot defer only to religious conscience without discriminating among citizens. Whether and when democratic deference to conscience is desirable turns out to be a separate issue from the question of whether religious identity is special, but its importance is inseparable from an appreciation of the centrality of ultimate ethical commitments, including those given by a person's religion, to individual identity.

Freedom of religion is a subclass of freedom of conscience, which is freedom to pursue one's ultimate ethical commitments within the limits of legitimate laws. Two-way protection suggests that the limits of legitimate laws should sometimes accommodate conscientious dissenters, when such accommodation does not threaten discrimination. A great advantage of accommodating conscientious dissent is that it respects the ethical agency of persons in democratically defensible terms. Accommodating dissent based on ethical identity recognizes a reciprocal relationship--of mutual respect--between conscientious citizens and the democratic governments that imperfectly represent them. Neither democratic laws nor conscience receive full rein: both are constrained by the other in principled ways that any liberal democrat can defend.


Are democrats also an identity group? A democrat is anyone who is committed to civic equality--public respect for persons as ethical agents. Democrats therefore need not be so committed out of identification with others. They may be so committed because they believe every person deserves respect as a civic equal. Some democrats, however, are so committed to these principles out of identification with other people. Their democratic impulse may emerge from a common human reaction to the suffering of others: "There but for good fortune (or the grace of God) go I." Logically speaking, democracy does not need to rest on mutual identification among people, but it may do so. Practically speaking, identification with others can be a path by which many people become committed democrats.

In light of the common philosophical view that justice should be based on impartiality, it is important to emphasize that there is nothing wrong (even from an impartial perspective) with supporting just causes out of mutual identification with those who are suffering from injustice or with others who are committed to organizing against injustice. Those whose ethical commitments operate independently of any mutual identification still need to join with others to be politically effective; therefore, practically speaking, there is not much to be gained by an extensive discussion about whether democrats are necessarily an identity group. They can be. What is important to recognize is that in identifying with other people, democrats may be better able to act effectively in democratic politics, since identification helps people join together in political causes. Democrats identity groups are therefore good from a practical as well as an ethical democratic perspective.

The idea of a democratic identity raises another question of political ethics. Is being a democrat a comprehensive identity, and if so, does it threaten creative human agency? As I understand it, a democratic identity is not comprehensive. I can identify as a democrat and also as a feminist, Jewish-American, friend, foodie, political philosopher, lover, mother, and lots more (or other). Democrats are not--or at least need not be--any single identity either exclusively or above all else. Multiple group identities coexist in individual persons. Democracy does not offer individuals a comprehensive template for making choices in life. Rather, it informs contributions to politics, and gives people great freedom to live their lives as they see fit, consistently with the equal freedom of others. Equal freedom includes the liberty to identify with many groups.

What, then, lies beyond group identity? Individuals have many identities that are not captured by any identity group. We have identities that are idiosyncratic to our own personalities, and we also have identities that are shared with others--who are wise or foolish, careful or careless, neat or sloppy, serious or light-hearted, and so on--but these shared characteristics, however important to our personal identities, are not the kind that constitute an identity group.39 Identity groups are characterized by mutually recognizable social markers that draw people together in politically significant groups. Many important features of an individual's identity are not captured by identity groups. Even if all our personality characteristics have a social dimension--otherwise they could not even be expressed as personality characteristics--not all are group identities of the kind that are featured in this book or that are central to democratic politics. Yet individual identities that do not lend themselves to identity groups may be at least as central to a person's sense of self. It is important to keep in mind that the group identities around which political identity groups form do not encompass the entirety of individual identities or even the most important parts of people's identities, even if they are often among the most politically significant.

Corresponding to the coexistence of multiple identities in a single person (some group identities, others not) is the coexistence of multiple groups and individuals that play a role in democratic politics. Some groups are identity groups; others are not. Many are a mixture of both. Whatever the mix, we should also recognize that some individuals, sometimes acting outside of interest and identity groups, can be politically effective. This is consistent with our opening observation that most people can make a significant political difference in democracies only by acting in groups. There are exceptional individuals who can make a difference without joining any identity or interest group by following a personal calling that happens to have political relevance. When Albert Einstein spoke out about politics, he was heard. When Nobel Prize winners like Toni Morrison speak out politically today, they are heard more than most other individuals, even though they are acting alone. The same can be said about the many rich and famous celebrities like Jane Fonda, Charlton Heston, Robert Redford, Barbra Streisand, and Steven Spielberg, although it is significant that those who seek the greatest political influence create or lead identity or interest groups (as Redford has done for environmentalists and Heston for the National Rifle Association). Individual political action is just as variable in value vis-à-vis democratic justice as identity group politics, but generally it is less effective because the numbers legitimately count in a democracy. Since wealth also counts, often unjustly, rich people are less dependent on group identities for effective political action. Yet they, too, form and join identity groups for some of the same reasons as anyone else.

In democratic politics, most people are most influential in groups, and identity groups are a manifestation of a basic freedom of association. Democrats therefore need to think about the ways in which a politics that depends in no small part on identity groups can work to better secure equal liberty, opportunity, and civic equality for all individuals, not only for the most privileged or the most powerful members of disadvantaged groups. The relationship between group identity and democratic politics is far more complex than blanket critiques and defenses of identity politics suggest. A democratic perspective attends to the interplay between group identities and democratic politics and assesses their relationship on the basis of broadly defensible principles of justice.

Identity groups as such are neither friends nor enemies of democratic justice. They pose distinctive challenges that have been neglected by political theorists who overlook the advantages of organizing on the basis of mutual identity in democratic politics and by political scientists who lump all politically relevant organizations together under the rubric of interest group politics. It is the aim of this book to overcome both kinds of neglect. A democratic view recognizes the legitimate but also problematic parts played by group identity in democratic politics and therefore the importance of distinguishing between the good, the bad, and the ugly of identity group politics.

Return to Book Description

File created: 8/7/2007

Questions and comments to:
Princeton University Press

New Book E-mails
New In Print
PUP Blog
Princeton APPS
Sample Chapters
Princeton Legacy Library
Exam/Desk Copy
Recent Awards
Princeton Shorts
Freshman Reading
PUP Europe
About Us
Contact Us
PUP Home

Bookmark and Share