“Group Rights” and Racial Affirmative Action
This article argues against the view that affirmative action is wrong because it involves assigning group rights. First, affirmative action does not have to proceed by assigning rights at all. Second, there are, in fact, legitimate “group rights” both legal and moral; there are collective rights—which are exercised by groups—and membership rights—which are rights people have in virtue of group membership. Third, there are continuing harms that people suffer as blacks and claims to remediation for these harms can fairly treat the (social) property of being black as tracking the victims of those harms. Affirmative action motivated in this way aims to respond to individual wrongs; wrongs that individuals suffer, as it happens, in virtue of their membership in groups. Finally, the main right we have when we are being considered for jobs and places at colleges is that we be treated according to procedures that are morally defensible. Morally acceptable procedures sometimes take account of the fact that a person is a member of a certain social group.
KeywordsAffirmative action Collective rights Diversity Group rights Harm Identitarian harms Identity Membership rights Probabilistic harms Race
Defining “Affirmative Action”
When in the Four Quartets, T. S. Eliot wrote that “human kind/Cannot bear very much reality,” he was not, of course, thinking about the attitudes of United States citizens to race. But the facts about how little racial reality most US citizens can bear are pretty dramatic.1 I rediscover a sense of awed astonishment each time I look at the data on the distance between US public opinion and US social reality when it comes to black and white. As Melissa Harris-Lacewell has observed, there is a “wide and persistent gap between the political attitudes of white and black Americans.”2 And that gap is clearly grounded in very different beliefs both about the state of race relations and about the extent of continuing differences between white and African-Americans life chances. On this latter issue, there is little doubt that the views of many white people are quite out of touch with the facts.
You might think that the situation was something like this. White people, asked about race in the US, tend to think (in most cases, correctly): “I and my friends and the other white people I know are just not as racist as white people used to be in this country. So things have clearly gotten better.” And black people tend to think (in most cases, correctly, also): “My friends and I and many of the black people I know about still face more obstacles, on average, than similarly situated white people. So, though things were once worse, they certainly aren’t where they should be. As for the idea that Americans no longer hold anti-black prejudices … excuse me, if I do not go along.”
In these circumstances, one of the most useful things that scholars can do is to remind black people that white people are indeed averagely less malevolently racist than previous generations, while insisting to white people that there is overwhelming statistical evidence of continuing racial inequalities in every major sector of social life. These facts about continuing racial inequality are clearly central to thinking about affirmative action, and the first of the many useful services that James Sterba’s Affirmative Action for the Future performs is to summarize in a straightforward way some of this statistical evidence.3
A second useful service is to give us, again in straightforward summary, a survey of the major forms of affirmative action currently practiced in this country and around the world.
(1) Affirmative action is a policy of favoring qualified candidates for positions of one group over qualified candidates of other groups. (2) The feature of candidates in virtue of which the policy favors them must be minority status, gender, or economic disadvantage.4 (3) The policy has to be aimed either at outreach or at remedying discrimination or at achieving diversity. And (4) it must have as its ultimate goal the achievement of a society that is racially, sexually and economically just. (5) Racial and gender justice will be achieved only when societies are color-blind and gender-free. (Sterba 2009, p. 32)
No definition of anything interesting is likely to satisfy everybody. I think I can imagine policies of affirmative action that fail (2) or (5): South African blacks are not a minority and there is no evidence that South African affirmative action aims, in the long run, for a society that is either color-blind or gender-free. And I do not think, in any case, that I want a gender-free society, even though I favor many forms of affirmative action. But Sterba’s definition has two very strong points in its favor. First, it mentions all of the major rationales that have been seriously considered for policies of group-based preference. And second, in building into the definition the idea that affirmative action does not entail favoring people for positions for which they are not relevantly qualified, it avoids one of the sillier arguments against affirmative action.
In avoiding that argument, however, it only defers the question of what qualifications are relevant; and it also obscures the possibility that membership in a group can be qualification, so that, in some cases, where race or gender is used as a basis for preference, the other candidates are in fact not relevantly qualified. I am thinking here, of the case for so-called “diversity role-models” made by Amy Gutmann in Color Conscious: The Political Morality of Race. (Appiah and Gutmann 1996, p. 132) In teaching white children, through social experience, the value of interacting with non-white people on terms of equality, being a non-white person is a qualification. It is not clear to me that policies of this sort would count as affirmative action in Sterba’s definition, because they do not advantage qualified members of one group over qualified others. Instead, they treat membership of the group as a qualification. But these are niggling worries about Sterba’s definition, not problems for his discussion. For in the discussion, he both addresses South African affirmative action and recognizes that group-membership can be a qualification. The definition does not get in the way of the analysis because Sterba is too attentive to the real issues to be distracted by his own official definition.
The history of philosophy suggests that adequate definitions, conceived of as providing conditions individually necessary and jointly sufficient, are, in any case, rather hard to come by. Once concepts are available their contours are reshaped in use in order to do the jobs they help us to do. That has happened with the idea of affirmative action, which has come to do a range of different jobs for us, the most important of which Sterba discusses. So, if I were asked to say what affirmative action is, I think I would list the sentences of Sterba’s book, “JS,” remove every direct use of the expression “affirmative action,” replacing each instance with the variable, “x”, to produce “JS(x),” and say that affirmative action is the x such that most of the sentences of “JS(x)” are true of it. Ramsey-sentences continue to have their uses.5
By and large then, with the occasional demurral, I found myself in agreement with Sterba’s main thrust. But there is one place where I found myself out of sympathy with his approach. And that is in his treatment of the moral status of group rights. For on this topic he shares assumptions with many of affirmative action’s friends and almost all of its enemies that I think collapse upon examination.
So in this paper I want to argue something about various ways in which the notion of group rights might be understood and what moral rationales their can be for them. I will also insist on a distinction between moral and legal rights. In particular, I will claim that there is simply no general moral principle that warrants an across-the-board hostility to the legal assignment of rights to people on the basis of their membership in social groups. My main object, then, is to strengthen one of Sterba’s responses to some of affirmative action’s critics.
Two Kinds of Group Rights
It will help, though, to say where these issues fit in Sterba’s general picture. His book supports affirmative action by defending seriatim the three aims of outreach, remedying discrimination, and achieving diversity. Sterba considers affirmative action aimed at outreach so uncontroversial that he devotes his time simply to saying what it entails. The definition he offers is this: “All reasonable steps must be taken to ensure that qualified minority, women, and economically disadvantaged candidates are made as aware of existing jobs and positions that are available to them as are nonminority, male, or economically disadvantaged candidates.” (Sterba 2009, p. 37) It is already interesting, for my purposes, that this is an uncontroversial idea, because it would be natural to describe this proposal as giving, say, African-Americans as a group the right to equal access to information about available positions. Suppose a town had two newspapers, one with a predominantly white readership and one with a readership that was predominantly black. And suppose an employer in that town failed to advertise jobs in the black newspaper. If the law requires outreach—as Sterba thinks it uncontroversially might—do not blacks, as a group, have standing to object or bring action? But then, if group rights are morally troublesome, why is this policy not troublesome, too?
To explain why there is so little objection to “group rights” of this kind, let me make two distinctions. The first is between what I have elsewhere dubbed “collective” rights, on the one hand, and “membership” rights, on the other. (Appiah 2005a, b, pp. 72f.) Collectiverights are rights that are exercised collectively. To make sense of a collective right we have to be able both to identify a group and to define the mechanisms by which it can assert the right.
So, for example, the people of every nation have, as a matter of international law, the right to self-government. They exercise this right through democratic procedures, which, if they are to be morally acceptable, must treat them as equals. As a result there may be individual rights implicated by the collective right to self-government. Nevertheless, self-government is collective action: individuals cannot do it, even though they can participate in it. Someone might argue that this is just a loose mode of speaking and that all the rights really involved here are individual. But that someone needs to say precisely how this claim should be cashed out. Even then, it is open to us to respond that the individual rights here exist because there is a collective right, and not the other way round.6
Membershiprights, on the other hand, are individual rights that people have in virtue of their membership in groups. Voting rights are individual rights that we have as citizens, as members of a political people. The right to a trial by the House of Lords was once a membership right of its members.
It would be natural to say that only collective rights are genuine group rights; for only they involve assigning rights to groups. But I think that those who say they are skeptical of “group rights” often mean to be challenging membership rights. As I wrote in The Ethics of Identity: “It was an objection to the membership rights of whites (and the membership burdens of blacks) that underlay much of the opposition to American Jim Crow and to apartheid.” (Appiah 2005a, b pp. 72f.) So it is worth insisting that there are many membership rights that are quite uncontroversial. Citizens, as I said, have voting rights qua citizens. These are individual rights you get in virtue of being an adult member of, say, the US citizenry, which is surely a group in the sense in which people use the word “group” in speaking of group rights.
Rights, Legal and Moral
The second distinction we need is between legal and moral rights. Even if there were, as many people seem to suppose, a deep moral objection to membership rights, it would not follow immediately that the objection extends to assigning legal rights to groups, if that is the best available way of securing the individual rights that are at stake. Suppose members of an American Indian tribe face employment discrimination. The discrimination is objectionable, no doubt, because it denies the rights of individual Indians. But giving group standing to the tribe might be the best legal way of guaranteeing those individuals their rights. There could be many objections to proceeding this way in a particular case, but none of them involves the premise that the legal assignment of a right to a group is morally wrong. Is it eo ipso wrong to allow the golfers' club to own land? Or the soccer club to have standing to sue the national league? We can make, or course, a distinction between the legal person—the tribe, the club—and the members as a group. And you might insist that it is that legal person—the collective agent—which has the right. But then it will be open for us to make the case for assigning legal personhood (and mechanisms for its exercise by real persons) to any group. These examples show that sometimes that case can be made.
I do not mean to concede that there are no moral membership rights, however. It is at least prima facie plausible that the individual right to democratic participation is a moral membership right, since it is precisely the legal denial of this right that was the moral objection to denying women or blacks the suffrage. And if that is so then it is simply wrong to claim that there are no moral membership rights. It would follow, as I shall argue below, that one premise of a familiar “group rights” argument against certain forms of remedial affirmative action is just false. Whether we take group rights to be collective or membership rights, there are moral and legal group rights aplenty.
With these distinctions in hand, we can now ask why outreach to groups might be so uncontroversial. Note first that the rights in question are legal. Note next that they are collective rights. Granting the right of qualified black people in the Chicago area to be made aware of jobs at the University of Chicago does not impose on the University a duty to make each qualified black person aware of it; nor even the lesser duty of ensuring that each qualified black person has the same probability of finding out about it as every qualified non-black person. Rather, it requires the university to ensure that qualified blacks are about as likely to learn of the jobs as whites; and the “about as likely …” here is, for the reason I just gave, a predicate of the group not of individuals.
Since legal collective rights are extremely common, as I pointed out, it is hard to mount a principled objection against a legal collective right to outreach in virtue of its being a legal collective right. But there is a further reason why a collective right to outreach for blacks is uncontroversial: It does not seem to entail denying the individual rights of non-blacks. That is because it is explicitly stated as granting the same collective right to other groups as it offers to the previously excluded groups. Even if an individual white person could show that a policy of outreach reduced his probability of employment—which, of course, it often would—it is hard to see how, in these circumstances, he could claim that this reduction in probability amounted to denying him a right. As we shall see below, it is often not the group right for the formerly excluded that opponents of affirmative action really object to, it is the alleged denial of some individual right to members of historically privileged groups. In this case there is no plausible argument of this form.
Once you spell out the duty of outreach here, though, I am not so sure that it should be uncontroversial. It would usually be wrong to deny access to (and, more particularly, information about) opportunities on the basis of race. That is because it is wrong to deny people substantial social goods in the public world on the basis of morally irrelevant facts about them. But from the fact that it would be wrong to deny an individual a good just because she was black it does not, I think, follow that the state is morally required to take affirmative measures to assure that black people as a group have the same share of that good as other groups, other relevant things being equal. I shall argue below that there is an additional assumption needed to connect premise and conclusion here.
Still, as I noted, it is hard to see how outreach of this form could be objected to on the ground that it infringes on the individual rights of anyone in the labor pool. The way labor markets work, no one thinks that it is an employer’s obligation to assure that every individual in the pool has an equal probability of hearing of each job opportunity.7 It is wrong deliberately to deny this information to members of some group, especially if your intention is to make sure you do not have to hire qualified members of it. But the right not to be excluded from information in this way is a negative right: and there is, so far as I can see, no affirmative moral duty (and there is certainly no legal duty) to assure each individual an equal probability of inclusion.
Remediation and Group Rights
Sterba’s own response to the “group rights” objection to affirmative action comes in the course of his discussion of remedial affirmative action for past discrimination, and I shall turn to that issue in a moment. But he also insists, rightly, that there is evidence of continuing racial discrimination in housing and employment in the present. And in thinking about what he notes about this issue we shall come across some difficulties in understanding his position.
The proper response to continuing discrimination, Sterba argues, begins with (a) trying to stop it and continues with (b) providing legal remedies for its individual victims. But none of this need involve affirmative action in his sense and, as he points out, most opponents of affirmative action agree that these responses are morally appropriate. (Sterba 2009, p. 40) Controversy begins when we move towards remedies that are not so individualized.
need not be by an institution that has itself discriminated;
may advantage members of the disadvantaged group who have not suffered direct discrimination, and
may “pass over” (that is deny preference to) only members of the advantaged group. (Sterba 2009, p. 53)
The answer, presumably, is that if a policy is remedial and the institution engaging in it has not discriminated itself, then it must be remediating discrimination by some other person or institution. Sterba gives the example of a company operating in an industry where there has been racial discrimination. It may engage in affirmative action, he argues, even if has not itself engaged in racial discrimination. But what other sorts of connection between a company and other institutions justify affirmative action? May a company justifiably engage in remedial affirmative action because it is located in a city, state or country that has a history of racial discrimination in other industries; or in government employment; or in housing and education? I do not see how to generalize Sterba’s idea that a company may remedy past discrimination in its own industry beginning with the individualist theory.
But I think these difficulties suggest we are being misled by the individualist account of remediation. Condition (i) derives more naturally from a different picture of the problem to which remedial affirmative action is the solution. What justifies remedial affirmative action is not the fact that its agent has done some wrong but rather the fact that some group has been wrongly disadvantaged. Their interests have been wrongfully set back. (Feinberg 1987) Social institutions may be required to restore them to where they should have been. As a result, the question to be asked is not how the institution is connected with the past wrong, but how it can contribute to restoring the situation of the disadvantaged. That would explain (i).
So far as (ii) goes, not all the wrongs involved in discrimination are direct wrongs. Arguably, any black person in a city with a history of employment discrimination has been subjected to the following wrong: discriminatory policies made it true of her that, had she applied, she would not have been given fair consideration, even if she had been qualified.8 The remedy here can thus be conceived of as a response to this individual indirect wrong. The membership right is justified by the fact that the group whose members have it consists of just those people who have experienced this wrong.
But our new theory suggests a problem for condition (iii). If someone has been disadvantaged and she is owed remediation, and if a policy of preference is a suitable remedy, then why should it matter whether those it passes over were themselves advantaged?9 Because it is a policy of preference the members of the advantaged group will get, on average, less under the new policy than they would have gotten without it. But if the policy is morally justified, that is no basis for complaint. If justice requires restitution to Japanese Americans for the wrongs they suffered in internment in World War II, I cannot complain, when my taxes are raised to pay this restitution, that I did not do the interning.
What people are owed in the employment market is fair consideration according to rules that are morally permissible. They are not owed the same treatment as others if there is a morally relevant difference between them and others. They are not owed consideration without regard to their race, in particular, if race is a morally relevant feature. Nor can they complain that the policy of affirmative action sends a message of inferiority (as the old discriminatory policies often did). As Justice Stevens said, in a passage Sterba refers to, we can surely tell the difference between a welcome mat and a “No Tresspassing” sign. (Sterba 2009, p. 45)
For any particular job, there will be one successful candidate. An employer does not owe the other candidates that that person is the best qualified, in the sense that she is the person who is likely to make the largest contribution to the institution’s productivity. Companies are free to decide that they will contribute to things other than their own profits, provided there is no moral or legal bar to their doing so. In a society with a history of racial inequality, whose consequences are evident in continuing unequal distributions of social goods, contributing to eradicating racial inequality is a perfectly reasonable aim.10
There is, however, a general problem here. Because Sterba shares with most people the assumption that the only real rights and interests at stake are individual ones, he is required to propose forms of affirmative action that connect their beneficiaries individually with some wrongful disadvantage and connect those who are passed over individually with some advantage. But it looks as though no amount of statistical data about racial differences in income, wealth or opportunity is going to be enough to establish either of these things. And, indeed, the US Supreme Court’s tendency to require a showing of individual intent to discriminate has mostly been used to resist policies of affirmative action.
Part of the difficulty here, as I have argued before, comes from confusions about various senses in which our moral theory might be individualist. One basic idea, endorsed by, among others, Thomas Pogge, is that “the ultimate units of concern are human beings, or persons—rather than, say, family lines, tribes, ethnic, cultural, or religious communities, nations, or states.” (Pogge 1992, p. 48) Call this basic individualism (and ignore the fact that there are creatures, like fish, who are not persons that may nevertheless be objects of concern).
I am not sure that basic individualism is true, though many modern people appear to assume that it is. It strikes me as quite possible (and here I should admit to retracting a claim I have made elsewhere) that among the things that matter is the survival of cultures, and that this may matter beyond their worth to individuals. But let me accept this doctrine here. Because the crucial point, for Sterba’s discussion, is that basic individualism needs distinguishing from a different doctrine, which is that only individual moral agents have rights. Call this individualismaboutrights.
Basic individualism strikes me as much less demanding than it is normally taken to be. From the fact that it is individuals that matter fundamentally, it does not follow immediately that we have no good reason to assign rights to things other than individuals.11 For doing so might be the best available way of advancing the interests of individuals—which is our fundamental concern according to basic individualism. Given the complex facts of social and political life, it may well be, for example, that there are morally sufficient reasons to assign to nations a right to self-determination. And even if there are not, there are certainly sufficient reasons to assign to democratic societies the right to manage their own affairs as a matter of international law. So an argument would be required to show that basic individualism entails individualism about rights; and it would have a very stiff wind blowing against it, precisely because we do normally suppose that many groups have both moral and legal collective rights, as I have shown.
These problems show up when it comes to Sterba’s explicit response to the argument against remedial affirmative action as immorally entailing group rights.
The “Group Rights” Objection to Remedial Affirmative Action
Sterba cites the basic group rights objection to remedial affirmative action in a passage from Carl Cohen, which runs, in part: “Moral entitlements are not held by groups. … Rights are possessed by persons. As when persons are entitled to be made whole for some injury earlier done to them, the duty owed is … to them as individuals.” (Sterba 2009, pp. 57–58) I have already said that I think that the first two claims here are false. Sterba points this out, too. US citizens have a right to self-government. This right is distinct from the individual rights they have. More than this, it is a moral right, not a legal one; and everything I have argued is nevertheless consistent with basic individualism. But it is worth asking why Cohen’s claims figure in a discussion of remedial affirmative action at all.
After all, Sterba’s definition of “affirmative action” does not mention rights. It discusses preferences. And a policy could grant preferences to a group without granting a right to those preferences and without saying that the beneficiaries of the preference were independently morally entitled to them. When a company or a private university sets up a policy of affirmative action, it can do so by stating it will give certain preferences while denying that the preferences it grants are entitlements. And to do so it need claim only that it is morally permissible to grant these preferences. It need not argue that they are morally required.
Nor, since the policies in question are largely legal policies, is it relevant to deny that moral entitlements are held by groups, since it is legal entitlements that are in question. Even if it were true that there were no moral collective rights, there would certainly be legal ones. The relevant moral claim here would have to be that it is morally impermissible to grant moral entitlements to groups. And as I suggested earlier, if that were so, large areas of the law would have to be abandoned.
The fact that Cohen goes on immediately to discuss remediation for individual harms suggests that the real issue is somewhat different. What he really objects to appears to be granting to individual members of the disadvantaged group a remedy (in the form of a preference) that is not a response to an individual wrong against them. And this is—as I said earlier—an objection not to a collective right but to a membership right.
Now it is hard to get a focus on what the problem is supposed to be here. Suppose we announce a policy of preferring qualified blacks to equally qualified whites for positions in our company. Is this wrong because it may advantage a black person who has not been the individual victim of a racial wrong? The depressing truth is that almost every black person in the US has been the individual victim of racial wrongs. But even if an exceptionally lucky black person should show up, what moral principle declares it wrong to adopt a policy that occasionally rewards the undeserving, if it normally awards the deserving? And, in any case, if you are worried about the tailoring of remedies to wrongs, why not draw attention to the black people who may have suffered unrecompensed wrongs? If the problem is the proper distribution of remedies, we should be as worried about under-recompensing as about over-recompensing. Policies are hard to tailor tightly to the contours of moral desert. After all, civil law, like all law, gives remedies to people who do not deserve them and imposes burdens on those who have done nothing wrong, because juries make mistakes about the facts. It would be odd to object to tort law as a whole on the grounds that you cannot guarantee that it will never order undeserved compensation.
Perhaps Cohen thinks such a policy wrong not because some black people have not been subjected to racial wrongs, but because many have not been subjected to wrongs that are serious enough to warrant such an advantage. If so, what is needed is not a general argument against collective rights, but a specific moral argument against assigning particular rights (Or a proposal as to how to tie remedies to wrongs more closely). It is hard to see how a sound argument drawing on a general premise about the impermissibility of assigning membership rights could show that no such policies were morally acceptable, since, as I argued, some groups consist of people who have been subjected to the same wrong.
The obvious objection to granting preference to a black person who has not been seriously wronged by racial discrimination is that it is a substantial undeserved benefit and that, had it not been granted, somebody else at least as qualified could have gotten the job. But what exactly is the complaint on behalf of that other person? The policy did not discriminate against them. It gave preference for a morally sufficient reason to people in the group to which the successful candidate belonged. And, as I said before, what people are owed is fair consideration according to rules that are morally permissible. That they got.
Group Wrongs: Probabilistic Harms
The possibility that Cohen believes that many contemporary blacks are not victims of serious race-based harms makes it urgent, I think, to underline the sorts of facts about continuing racial disadvantage that, as I said at the start, Sterba has done us all the service of assembling. But I want now to point to a couple of mechanisms by which race-based harm is done and to argue that both these mechanisms are likely to cause harm to almost all black people in the US, giving them a continuing basis for individual claims to remedy.
Black people in the US—like whites and Asian-Americans and Latinos—have a shared racial identity. That racial identity leads them to be classified in many public contexts as black and responded to as such. One such form of response is simple race prejudice. And the presence of significant numbers of US citizens with anti-black prejudices, combined with fact that black people are in the minority, means that, on average, a black person enters most public contexts with a serious risk of paying higher psychic and material costs than otherwise identical white people. You are likely to get worse service, to have to wait longer for it, and to have to pay more for any goods whose prices are negotiable. You are more likely to be insulted or treated with contempt. Police officers are more likely to stop you and more likely to arrest you after stopping you. Indeed, you are more likely to be racially profiled in criminal justice contexts. Prosecutors are likely to give you worse plea deals and ask for longer sentences. Juries are more likely to convict you and judges are likely to give you longer sentences than similarly accused whites. As a result of facts like these—the sort of facts Sterba rightly begins by drawing attention to—you are more likely to be subjected to stress and to experience the ill health that flows from higher levels of cortisol and other physiological expressions of stress. Facts like these mean that the probability of race-based harms is generally higher for black people than for a similarly placed whites.
This is one reason the bare fact of significant statistical differences in wealth and education by race would be important, even if it were not the result of racial animus. Because people cognitively store information (and misinformation) about race, the effect of these inequalities is to lead each black person, whatever her own place in the distribution, to be treated as presumptively less endowed with certain desirable attributes than the average white person. Such beliefs help to perpetuate, in obvious ways, the very inequalities they presuppose. And that adds to the race-based burdens facing black people. This is the basis for thinking that we might be able to establish the additional premise needed to get from the fact that it would be wrong to deny an individual a good just because she was black to the conclusion that the state is morally required to take affirmative measures to assure that black people as a group have roughly the same share of that good as other groups, other relevant things being equal.
You can be harmed directly by being denied a job or being physically assaulted. But you can also be harmed when the probability of your getting a job is lowered or the probability of your being assaulted is raised. Being black is associated in the US both with direct harms and with a multitude of probabilistic ones, like these. While some lucky or well-off or well-situated blacks can escape the direct harms, we cannot usually avoid the probabilistic ones. If I take special care, I may be able to avoid arrest, but I can do nothing to protect myself from the fact that, if I were arrested, I would be more likely to be badly treated than similarly placed white people. Without attention to group differences like these you cannot assure that there will not be socially-caused racial difference in life chances that wrong individuals.
Group Wrongs: Identitarian Harms
Our society associates racial identities with stereotypes of two kinds.12 One sort of descriptive stereotype associates particular racial groups with behaviors they take to be typical. These associations may or may not be empirically grounded. It may or may not be true, that is, that white people are generally racist. But true or not, such stereotypes shape the responses of other people to you. If your racial group is associated with negative qualities, you will enter many contexts with the burden of that negative expectation, whether or not you yourself fit that negative stereotype.
There are also normative stereotypes: ways that people think those of your racial group should behave. Normative stereotypes are one reason racial identities are important to some people: because of them, a black person can comport herself in the way she takes to be apt for a black person. Some descriptive stereotypes flow from the fact that certain normative stereotypes are part of our common knowledge. People expect black people to talk a certain way, in part because they know that black people think they should talk that way. The racial history of this country and the facts about the racial distribution of resources and burdens—differences in everything from wealth and income to incarceration rates—mean that many of the negative descriptive stereotypes of black people place a burden upon the majority of US blacks who do not fit them. The harms of an identity damaged in this way—the burdens of racial stigma—are borne by all US blacks.
In short there are harms that black people suffer asblacks and you can only remedy them if you grant black people entitlements asblacks. If you do not know these things you are among the many in the US who cannot bear the reality of race in the US.
Diversity and Group Rights
Sterba’s final rationale for affirmative action is diversity. And its most obvious application is in college admissions. Why is diversity a good? Because diversity of social identities makes education and research better. Exposure to people of the major social kinds in your society on terms of equality and in the college setting is part of an education for life in a society of diverse identities. Universities are also research institutions and in large areas of the social sciences and the humanities interest is differentially distributed by social identity, in both senses of the term: social identity affects what you find interesting and it affects what stakes you have in the answers to questions. And that means that a community with qualified researchers a variety of social identities will explore more options, including resisting more effectively answers that reflect the interests of one group—whose amourpropre is fed, for example, by the belief that “we” are cleverer than “they” are. We do not need to assume that social identity determines what you are interested in, or that people will necessarily work harder to undermine mistakes about their own group. But because identity is associated with these things, we can assume that having a diversity of identities makes it more likely that we will have both a richer diversity of questions asked and a real pressure to challenge the old errors that pleased the privileged.
As with employment, the only right everyone has in admissions is to be fairly assessed according to the stated rules by morally permissible criteria. My own view, pace most in the US, is that there is no notion of pre-institutional desert that entitles anyone to a place at any particular university or company.13 Qualified people may deserve access to a quality higher education. It is simply not obvious that people who are more qualified deserve access to a higher quality education. Nor is there a notion of intellectual merit such that those who possess it are entitled to places at the best schools in order of merit, not least because different fields require different aptitudes and trainings. Since race and gender are qualifications in this area—racial and gender diversity are part of what is legitimately being sought—there is, in any case, no legitimate complaint to be made of the form, “If I had been black, I would have gotten in,” any more than there is a legitimate one of the form, “If I had been a great football player, I would have gotten in,” or “If I had had a higher S.A.T., I would have gotten in.”
In sum: there are good reasons for doubting the claim that affirmative action is wrong because it involves assigning group rights.
First, affirmative action does not have to proceed by assigning rights at all. A preference can be granted that is not a right.
Second, whether by “group rights” you mean collective or membership rights, and whether you mean moral rights or legal ones, it is just not true that there are no legitimate groups rights: I have made a prima facie case that there are both moral and legal collective rights and legal and moral membership rights.
Third, there are continuing harms that blacks suffer in the US as blacks—some probabilistic and some identitarian: and claims to remediation for these harms can fairly treat the (social) property of being black as tracking the victims of those harms. Affirmative action motivated in this way aims to respond to individual wrongs; wrongs that individuals suffer, as it happens, in virtue of their membership in groups.
Finally, the main right we have when we are being considered for jobs and places at colleges is that we be treated according to procedures that are morally defensible. Morally acceptable admission procedures sometimes take account of the fact that a person is a member of a certain social group, as we see in the case for diversity affirmative action. To say that someone “deserves a place” (or a job) is not to point to an interesting normative property that connects them with the point of a college (or firm) independently of its admission procedures: it is to say, rather, that (supposing the current procedures are morally permissible) proper consideration under them would lead to her admission.14
I do not think Sterba needs to deny any of these conclusions … though, since he is a philosopher, he probably will! My only complaint is that he does not endorse all these truths explicitly in making his case for affirmative action in the years ahead.
Eliot (1968, p. 14).
Harris-Lacewell (2007, p. 107). This article provides a useful survey of the data on differences in political attitudes by race.
As Sterba points out, gendered differences of perceptions of gender inequality are not as widely studied as racial differences in perceptions of racial inequality. But he is surely right that many US citizens (of both sexes) underestimate the scale of gendered inequality as well. I am going to focus in this paper on questions about race.
Sterba actually uses “minority” not “race” here, but the rest of the definition proceeds in such a way that it is natural to assume he means “racial (or ethnic) minority.” If he meant to exclude other kinds of minority, such as, for example, gays and lesbians, it would have been helpful to say so. And we could still ask why he includes a historically disadvantaged majority race (such as blacks in South Africa) in his discussion.
My thought is this: someone might hold that the reason that citizens have the right to vote is that they belong to a nation that has a collective right to self-government. The mechanisms by which a particular constitution allows individuals to participate in the exercise of that collective right may be normatively questionable. (Why, for example, can a majority of those voting, who may constitute a minority of those eligible to vote, determine what counts as the national will?) But the problems with the ways in which this idea is actually implemented do not, in themselves, show that we cannot proceed by deriving individual rights to vote from the collective right to self-government.
Even if someone did think this, they would be obliged to show that it was possible, since ought implies can. It is enormously difficult to imagine a system that guaranteed that the probabilities of every person finding out about a job were identical. This is not to deny that it is a reasonable ideal that there should not be systematic attempts to deny information about job opportunities to some specified class of people; nor is it to say that jobs should not be widely advertised. It is just to say that the reason why these are reasonable ideals cannot be that we are morally required to guarantee that each potential candidate has the same chance of hearing of the opportunity.
Someone might argue: “No, she only suffered a wrong if she actually applied and was denied a job.” But that is like saying that I did not wrong you when I set a man-trap in your yard, on the grounds that you did not walk into it. You can wrong somebody by raising the probability that they would suffer some direct harm in certain counterfactual circumstances.
There are other difficulties for the demand that remedial affirmative action pass over only those who have benefited from the form of discrimination it aims to remedy. Sterba argues that whites have been the beneficiaries of the racially unequal distribution of educational resources. But many whites—in rural Appalachia, for example—have been given inadequate educational resources by the US educational system too, and it is not clear that they have been benefitted even by its racially discriminatory features. The difficulty here is not primarily factual: it is conceptual. How are we to individuate the practice whose harms to black people we are remedying before going on to identify its beneficiaries? Consider local systems of racial disadvantage. Should we take the educational system as a whole? Or do we look just at its racially discriminatory practices? If so, should we examine these singly or as a whole? The educational system as a whole does not advantage the white poor. The racially discriminatory practices within it may actually disadvantage some inner-city whites and advantage some suburban blacks. Similar difficulties arise in assessing the beneficiaries of racial discrimination in the labor market. In most of the US legal cases in this area that Sterba discusses, the harms remediated derive from local discrimination in employment and contracting, and blacks who enter the labor market from elsewhere are unlikely to have been victims of these local practices, just as whites who enter from elsewhere are unlikely to have been beneficiaries. In these circumstances, it will be impossible to satisfy Sterba’s condition (iii) by any plausible policy of affirmative action, unless it denies preference to blacks and grants it to whites when they are each newly arrived in the area.
I realize that the claims of this paragraph will strike some as controversial. So far as I can see, however, the only moderately controversial presupposition of the argument here is that a private corporation is morally free to pursue goals other than maximal productivity within the law.
The discussion here is a (simplified) version of the account offered in Appiah (2001, pp. 55–71).
The number of places available at a university is limited. I take it to be obvious that, absent special circumstance, this is morally permissible. Either the number of people pre-institutionally entitled to these places is greater than the number of available places, or it is not. If it is greater, then the size of the number of places available is morally impermissible, which is inconsistent with what I just suggested was obvious. So it must be less. But then surely some explanation of the remarkably small number of those with pre-institutional entitlements to these places is due. The most likely explanation, so far as I can see, would be that everyone in the world is ranked with respect to pre-institutional entitlement to places at this university. This would require that there was some function from all the many dimensions of assessment that are relevant to assigning people to places into a final ordering. Then, if there were n places, the first n people in the ranking would be entitled to them (or, perhaps, the university would be morally obliged to offer places to those who applied according to their order in this ranking). The reason I doubt that this is so is that I cannot believe that there is such a function. Here is just one kind of reason: the abilities required to make the best use of a university place are very various, because so many different combinations of intellectual and affective attributes are required in different fields. We admit people to colleges, however, not to fields and we leave them free to discover what fields they enter. Now consider two students, with different combinations of intellectual gifts and temperaments: John would do very well, let us suppose, in English, if she chose it, Mary in Math. (Suppose, if you think this is relevant, that they come from similar homes and have worked equally hard so far.) How should we rank their pre-institutional entitlement to a place, not knowing what major they will choose?
So there is one way in which you might be said to deserve a place at an institution independent of its actual procedures, that is, if you would be admitted under any morally acceptable admissions process properly applied.
I am grateful to J. Angelo Corlett for comments and suggestions on an earlier draft.
- Appiah, Kwame Anthony, and Amy Gutmann. 1996. Color conscious: The political morality of race. Princeton: Princeton University Press.Google Scholar
- Appiah, Anthony. 2001. Stereotypes and the shaping of identity. In Prejudicial appearances: The logic of American anti-discrimination law, ed. Robert Post, 55–71. Durham: Duke University Press.Google Scholar
- Appiah, Anthony. 2005a. The ethics of identity. Princeton: Princeton University Press.Google Scholar
- Appiah, Anthony. 2005b. Thinking it through. Oxford: Oxford University Press.Google Scholar
- Corlett, J. Angelo. 1994. The problem of collective moral rights. Canadian Journal of Law and Jurisprudence, 7:242–256.Google Scholar
- Eliot, T.S. 1968. Four quartets. New York: Harcourt.Google Scholar
- Feinberg, Joel. 1987. Harm to others. Oxford: Oxford University Press.Google Scholar
- Harris-Lacewell, Melissa. 2007. Political science and the study of African American public opinion. In African American perspectives on political science, ed. Wilbur Rich. Philadelphia: Temple University Press.Google Scholar
- May, Larry. 1990. The morality of groups: Collective responsibility, group-based harm, and corporate rights. Notre Dame: University of Notre Dame Press.Google Scholar
- Sterba, James P. 2009. Affirmative action for the future. Ithaca: Cornell University Press.Google Scholar