Curtis Crawford - March 2004

After a 25-year silence on the subject, the Supreme Court has pronounced on the constitutionality of race-based affirmative action in university admissions. Those who had hoped that the issues would be wisely clarified and cogently weighed must have been greatly disappointed. The two cases accepted for review, Grutter v. Bollinger and Gratz v. Bollinger, provided valuable information on how universities actually implement preferential admissions. But the Court's rationale in Grutter followed a blind alley from which its deliberations never recovered.

The litigation of these two cases revealed large racial inequalities in the treatment of applicants with similar academic credentials. For example, at the trial in federal district court, the Michigan Law School admission grid for 1995 (the year Ms. Grutter was rejected) was offered in evidence. For all applicants, identified by race but not by name, the grid included data on Undergraduate Grade Point Average (UGPA), Law School Aptitude Test score (LSAT), and admission or rejection. Each cell of the grid combined a small range of grades and scores. The racial or ethnic groups designated by the School for preference were Black, Mexican, Puerto Rican, and Native Americans. White, Asian, Other Hispanic, Foreign, and Racially Unidentified applicants received no such preference.

The size of the preference is indicated by the gap between the rates of admission for Favored Minorities and for Other Applicants. The cell containing the median grade and score for all applicants ranged from UGPA 3.25-3.49 and LSAT 161-163. In this cell, all Favored Minorities were admitted but only 5% of Other Applicants. Indeed, the 100% admission rate for Favored Minorities in the median cell was higher than the 95% rate for Other Applicants in the top cell (UGPA: 3.75+ and LSAT: 170+). Down at the 30th percentile (applicants with grades and scores below 70% of their rivals), 83% of Favored Minorities but just 1% of Other Applicants gained admission. Farther down, at the 10th percentile level (UGPA 2.75-2.99, LSAT 148-150), the 6% admission rate for Favored Minorities was higher than the 5% rate for Other Applicants in the median cell. In sum, Favored Minorities in the 10th percentile cell had a better chance of admission than Other Applicants in the median cell, while Favored Minorities in the median cell had a better chance than Other Applicants in the top cell.

The Supreme Court's Contribution

The general question before the Supreme Court was whether such a policy is constitutional. The framework for answering the question had been indicated in previous decisions: Since the policy requires the classification of people based on race or ethnicity by a state agency, it is presumably a violation of the Constitution's Equal Protection Clause. However, this presumption may be overridden, if the policy is "narrowly tailored" to serve a "compelling state interest."

This framework seemed to provide a judicial freedom to reach major issues of right and wrong, while carefully weighing the public good. In the Michigan lawsuits, the Court could have asked whether placing large numbers of Black, Latino and Native Americans in our elite colleges and professional schools is a compelling state interest, which justifies racial preference by the government. What are the benefits and harms to be expected? Do the former, despite the latter, create a public necessity?

These were not the questions the Court chose to address. The possible compelling interest it preferred to consider was "student body diversity." In making this choice, it failed to face an insuperable logical problem. According to the dictionary, "diversity" simply means difference. Student body diversity or difference is so indeterminate an interest that no reasonable judgment can be made as to whether it is compelling, or whether any given means is narrowly tailored to serve it. Such judgments are possible if, and only if, the type and extent of diversity are specified. Without such specification - without stating the kinds and amounts of difference at issue - the Court cannot know what it judges. Without such knowledge, the judgments have no basis.

These absurdities were self-inflicted wounds. Nothing in the nature of "diversity" required the Court to do what it could not. It was free, under reason and law, to frame the question in a way that could be answered. Rather than student body "diversity" in general, the subject of deliberation could easily have been student body "diversity," specified in kind and amount. For example, the Court might have weighed, as a possible compelling interest, Michigan Law School's stated goal of enrolling "underrepresented" minority students (kind) in sufficient numbers (amount) for extensive campus interracial interaction. But deliberation and judgment concerning student body "diversity," without such specificity, is nonsense. It suggests the proverbial warrior, mounting his steed and riding off in all directions.

Contemplating this essay before the Michigan decisions, I had planned to supplement the Court's constitutional inquiry with an examination of the underlying moral issues. There is not much to supplement, which makes it even more important for the public to revisit and clarify the moral considerations that bear on race-based unequal treatment. The Justices have said what the Constitution permits; the rest of us need to consider what the policy should be.

Racial affirmative action began almost forty years ago with efforts to make sure that people were not being treated unequally because of their race. It soon developed into programs conferring special treatment based on race, especially in higher education and employment. Decisions typically affected have been admission to college and graduate school; and hiring, promotion and training for private and government jobs. The groups now regularly designated for favorable treatment based on race or ethnicity are Blacks, Latinos and Native Americans. Asians sometimes receive it; Whites, almost never. The advantage is usually conferred by applying a double standard, whereby the requirements for selection are less exacting for members of the favored group.

These programs have been upheld as a remedy for past injustice, yet condemned as an instrument of present injustice. They have been praised for increasing minority access to business and professional careers, and blamed for debasing standards in the process. They are supposed by some to have raised and by others to have undermined the self?esteem of their recipients and the value placed on them by others. The controversy is fierce, partly because people on both sides believe that their position is what justice requires. But contrary views cannot both be right. We must dig deeper than usually occurs in public discussion to uncover and disentangle the relevant standards for moral judgment.

Unequal Treatment in General

At the outset, we need to distinguish between unequal treatment in general, and unequal treatment based on race. The latter may or may not be a special case, with special rules. Unequal treatment is simply treatment that favors one person over another. People are treated unequally for so many reasons, in so many contexts, that the existence of a general moral rule may seem impossible. But I suggest that we have such a rule. Ask yourself if and when you think that treating people unequally is the right thing to do. Is it all right when there is no reason for it? That would be arbitrary. Is it morally permissible if there is a good reason? For example, is it permissible to favor one applicant over another if they differ in ability, character, training, experience, and the like? Of course. Concerning something as important as the opportunity for education or employment, should people ever be treated unequally without good reason? No. But if there is a good reason, is it morally permissible to treat them unequally? It is not only permissible, it may be required.

What if the individual difference on which special treatment is based has nothing to do with an applicant's ability or need? Suppose that a public university gives an admissions preference to in-state residents, or a scholarship preference for veterans. Does the rule still hold, that unequal treatment is morally permissible when it is reasonable? The reasons commonly offered are, in the first case, that a state university is financed by and owes a primary educational responsibility to the residents of the state; in the second case, that such scholarships are both reward and incentive for service in the armed forces. The reasons seem good to me, and my sense of right and wrong does not bar the unequal treatment in either example. Others may think the reasons poor and the treatment wrong. In either case, whether unequal treatment seems permissible depends on whether there seems to be a good reason for it. Preferential admission to a private university for the children of alumni is supposed to strengthen the school's relationship with its former students, thereby solidifying their continued interest and financial support, without which the quality and even the survival of the school might be jeopardized. Whether these are good reasons is disputed, but again the point is that, if one thinks the reasons good, one does not consider the preference immoral.

Supporters of racial preference think that the reasons for it are good: better, indeed, than for many kinds of preference that are generally accepted. Hence they conclude that there is nothing morally wrong with the unequal treatment they advocate. This conclusion is valid, if the rule for unequal treatment based on race is the same as the rule for unequal treatment in general. But are the rules the same?

Before tackling this question, a clarification is in order. Two words, preference and discrimination, are inseparable from a discussion of unequal treatment. Without realizing it, people define these terms differently, confusing and misunderstanding each other unnecessarily. "Preference" may refer to any unequal treatment, or only to unequal treatment that one disapproves of. "Discrimination" is variously defined as different treatment, as unequal treatment, or as unequal treatment that one disapproves of. In this inquiry, I define both "preference" and "discrimination" as unequal treatment (favorable in the former case, favorable or unfavorable in the latter). "Racial preference" and "racial discrimination" will always mean unequal treatment based on race or ethnicity.

Unequal Treatment Based on Race

Does the rule, that unequal treatment is morally permissible when there is a good reason, still hold when it is based on race? During the campaign to overthrow American discrimination against blacks and others, it was never suggested that if the discriminators had a good reason, their actions would be morally acceptable. The legislatures, schools, professions, businesses and unions that practiced racial discrimination were not asked about their reasons; they were simply told to quit. Any claims that their policies were "reasonable means to legitimate ends" were rejected as rationalizations for racial injustice. The overriding conviction was that racial discrimination was morally out of bounds, no matter what reasons the discriminators might offer.

Based on this moral principle, laws were enacted between 1940 and 1970 at the local, state and national levels, barring race-based unequal treatment in voting, housing, health care, public accommodations, public facilities, education and employment. These statutes established the right not to be discriminated against, and the corresponding duty not to discriminate, on account of "race, color or national origin." Rights are not absolute: they may be overridden by superior rights or by public necessity. But when unequal treatment on a particular basis is barred as a matter of right, people are not free to discriminate on that basis simply because they feel they have good reasons. The right not to be racially discriminated against was not reserved for members of particular groups, but ascribed equally to every person in the United States.

Was the moral principle behind this legislation mistaken? For Blacks it can be seen as a two?edged sword, banning adverse discrimination to be sure, but also prohibiting any discrimination in their favor. The antidiscrimination statutes left Blacks with two important disadvantages. They were still held back by deficiencies in ability, training and motivation attributable at least in part to past discrimination; and they faced the prospect that discrimination against them in the future, though illegal, would often occur. No one doubts that the social and economic condition of American Blacks would be better, absent their history of racial oppression. A plausible remedy would be racial preference, until both the effects of past, and the practice of current, anti?Black discrimination had dissipated. But such a remedy would require important exceptions to the general ban on racial discrimination.

Any society that decides to end an era of discrimination faces the same moral dilemma. If everyone is granted the right not to be discriminated against on account of race, the possibility of helping the victims of past discrimination through racial preference is lost. If members of the previously excluded groups are favored on the basis of race, the right of others not to suffer racial discrimination is denied.

There is a simple way to slice through the dilemma, which would assist many disadvantaged individuals. Instead of racial preference, a program could assist those who had suffered specific, oppressive treatment, such as chronic and substantial racial discrimination. Any person, regardless of race, who could demonstrate such treatment in his own case would be eligible for the assistance. Such a program would satisfy the racial nondiscrimination rule, since the basis for assistance would be individual injury, not racial identity. But it would help only a fraction of those who currently benefit from race-based affirmative action.

Rights or Necessities That Trump Racial Nondiscrimination?

Are there superior rights or public necessities that might supersede the right to racial nondiscrimination? The right to racial nondiscrimination, though momentous, is not the only care of the republic. Other (sometimes conflicting) rights and interests must also be protected. The moral dilemma of racial preference for some versus racial nondiscrimination for all might be avoided if, in certain circumstances, the right to racial nondiscrimination were trumped by a higher right or by public necessity.

A Superior Right to Equal Racial Success? Some argue that there is a right to equal participation for racial groups, which overrides the individual right to nondiscrimination. According to this view, 'equal participation' means equal success in wealth, status, and achievement, not for every individual, but for the average person in each group, as compared with the average American. A belief in this right is often the moral basis for affirmative-action goals, adopted for the purpose of increasing the percentage of "underrepresented" minorities in the higher echelons of education and employment, to match their share of the general population. If such a right exists, it would conflict with the right to racial nondiscrimination, and might overrule it.

Individuals in this country do have the right to equal respect as members of the body politic, but this has never implied the right to individual equality of wealth, social status or achievement. If individuals have no right to equal success, how would the groups that individuals belong to acquire it? Moreover, if each racial group-Whites, Blacks, Asians, Native Americans-possesses this right, why not each ethnic group-Poles, Irish, Italians, Slavs, Arabs, Jews, Mexicans, Cubans, Puerto Ricans, Dominicans, Argentines, Colombians, Sioux, Navaho, Aleuts, Pacific Islanders, Chinese, Japanese, Koreans, to name a few? Moreover, if these groups have a right to equal attainment, why not other groups-religious, sexual, geographical, political, occupational, athletic, artistic, commercial, and so on? The claim to a right of equal success for racial or ethnic groups does not survive inspection.

A Superior Right to "Just Compensation"? Does the right of some people to compensation supersede the right of others not to be discriminated against? The right to just compensation is certainly a powerful right: by requiring the payment of damages, for example, it can overrule the right not to be deprived of one's property involuntarily. But the nature of the right to compensation, properly applied, prevents it from requiring racial preference. Compensation is awarded as a remedy for injury, to the persons injured and occasionally to their immediate families. It is regarded as just, if its response to injury is reasonable or proportionate; having that basis, it needs no other.

Various ways have been or could be enacted, by which the right to compensation can redress individual injury to members of previously excluded groups, without involving racial preference. For example, under present law, a person may be compensated if injured by racial discrimination, and the compensation may take the form of preferential treatment. But since this preference is based on individual injury rather than racial identity, it is not preferential treatment based on race. If unlawful injury is shown, that is sufficient basis for compensation; if no injury has occurred, no basis exists.

In extraordinary circumstances, the legislature may compensate for injustice that was lawful when it occurred. Congress might have awarded preference, after the Civil War, to persons formerly enslaved; or, after the overthrow of de jure segregation, to persons who had been educated in substandard, segregated schools. The recipients in both instances, though injured because of their race, would have been compensated because of their injury. Such compensation would not have constituted special treatment based on race.

Moreover, Congress is free at any time to authorize a program of special assistance for people whom the law has failed to protect from discrimination. The program could require that an applicant present clear and convincing evidence of having suffered racial discrimination that was chronic and substantial, not fleeting or insignificant, and experienced in this country. Adverse discrimination in education and employment would be especially relevant. The right to be heard, the standards of proof, and the scope of the remedy would be the same, regardless of the race of the persons involved. In such a program, the right of just compensation would be wholly consistent with the racial nondiscrimination rule.

If individuals who have been subjected to racial discrimination can be given compensatory help without running afoul of the nondiscrimination rule, why not an entire racial group? Could we thus escape from our moral dilemma? Is it possible that all we need is a finding by the national legislature that discrimination against certain racial groups has been and continues to be so pervasive that every member of the group is entitled to compensatory preference? Many proponents of affirmative action proceed as if such a finding had occurred, in their own minds if not in the legislative process. This helps them to think of racial preference as compensation, rather than discrimination.

A legislative finding of this sort, though based on evidence of injury to some, would be mere supposition concerning others. But the right of just compensation requires proof of specific injury to the person who invokes it. A legislative decision to compensate an entire racial group could not meet this criterion; it would be discrimination masquerading as compensation. Moreover, a legislature permitted to stereotype racial groups sympathetically would be free to do the contrary. Based on data that discrimination against Blacks is much more frequent than against Whites, it could declare every Black a victim. Based on statistics that crime by Blacks is much more frequent than by Whites, it could declare every Black a criminal.

A Public Necessity to Prevent Racial Warfare? Everyone would agree that in time of extraordinary peril, a right may have to be suspended. Is this such a time? Is there now a clear and present danger to the republic, say, of racial warfare, to be avoided only by racial preference in education and employment? The urban riots of the 1960s may have aroused this fear. But the supporters of affirmative action do not ordinarily justify their programs in these terms. I know of no convincing evidence that the danger of racial warfare is present or clear, whether racial preference goes or stays.

A Public Necessity to Achieve Racial Diversity? Some, giving a broader definition to public necessity, uphold two propositions, (a) that racial diversity in education and employment is a public necessity, and (b) that racial preference is essential to achieve such diversity. If by "diversity" they simply mean difference or variety, proposition (a) may be true, but proposition (b) is manifestly untrue. In a society composed of many different groups, all one needs in order to ensure racial and ethnic variety in colleges and workplaces is not to discriminate. But among supporters of race-based affirmative action, "diversity" usually means having a larger number from "underrepresented groups" than would occur without racial preference. Using this definition, proposition (b) is true, but proposition (a) is false. So far as I am aware, no one has demonstrated a public necessity that racial groups be represented in education or employment in proportions higher than warranted by the fitness of their members, individually and impartially assessed.

A Necessity To Prevent Discrimination against Minorities? Some people argue, paradoxically, that racial preference prevents racial discrimination. They assume that unlawful discrimination against non-Whites in education and employment is common, since most educators and employers are White. According to the argument, when decision?makers have to meet goals for increasing minority participation, anti-minority discrimination is effectively prevented. Racial goals and quotas are therefore imposed, by institutions over their officials or by courts over institutions, to ensure that people who might discriminate will not do so.

This policy prevents violations of the right to racial nondiscrimination-by making certain that they occur. It purchases freedom from possible misconduct with the coin of assured misconduct. It enforces the rule by breaking it. Absent public necessity, this exchange is unacceptable. The public benefit alleged in justification is an additional reduction in discrimination against people of color. But this goal, though desirable, is not a public necessity. There is indeed a public necessity that the laws, including those against racial discrimination, be generally enforced, but there is no necessity that the rights of some be enforced by violating the rights of others. Take for example the problem of crime prevention. Crime rates in our society are egregiously high; they could perhaps be reduced by licensing citizens to steal the property of thieves, or by permitting policemen to violate the rights of suspects. Nevertheless, we (wisely) insist that the reduction of crime does not justify such means.

We have found that, if a general moral right not to be racially discriminated against is recognized, racial preference cannot be justified as serving a superior right or a public necessity. The supposed rights and necessities either do not exist, or do not conflict with the right to nondiscrimination. Is there another approach that might clear the way for racial preference?

Repeal or Limit The Right to Racial Nondiscrimination?

The moral right to racial nondiscrimination could be expunged or limited. One could (1) scrap the right altogether, (2) define the right more narrowly, (3) exempt education and employment from the nondiscrimination rule, (4) permit discrimination favorable to blacks, or (5) permit discrimination favorable to all "underrepresented" minorities. Should the United States have chosen (or now choose) one of these options?

1. Should The Right to Racial Nondiscrimination Be Scrapped? This option would require us to repeal our antidiscrimination laws and to reject the moral principle on which they are based. No one advocates this. The dispute concerning race-based affirmative action reaches only certain aspects of education, employment, government contracting and elections; access to other goods, services and opportunities remains under the antidiscrimination rule. With respect to housing, shopping, credit, transportation, health care, law enforcement, restaurants, motels, theatres, athletic events, public parks, and so on, there is an overwhelming consensus that racial discrimination should be forbidden.

Is this consensus warranted? Recall some of the considerations that produced it. First, a person's racial or ethnic identity ordinarily has nothing to do with the merits of the case: whether one should be rented an apartment, served at a restaurant, given medical treatment, convicted of a crime, etc. Second, one's ancestry is something over which one has no control. Third, unlike discrimination based on local or idiosyncratic factors, discrimination based on race is readily imitated throughout a society.

Still, why not leave people free to practice racial discrimination when there seems to be a good reason? The problem is that, when race or ethnicity is on the scale, reason is extremely corruptible. Those who practice the unequal treatment, and those whom it favors, almost always think there is good reason; those who are discriminated against, almost never. When racial or ethnic advantage is at stake, prejudice and partisanship transform reason from an impartial judge into a biased and sophistical advocate.

Wherever practiced, racial discrimination generates racial oppression, hostility and violence. Nondiscrimination is not easy, but it is the only standard to which members of every racial and ethnic group might agree, since it is the only standard that places no one at a disadvantage because of his group membership. Rationales can be developed for organizing a society on the basis of racial or ethnic discrimination: favoring some groups as original inhabitants or successful conquerors, as biologically or culturally superior, as victims of past oppression or present poverty. But these rationales are never lastingly persuasive to groups whose status they reduce. Those discriminated against, feeling it oppressive and unjust, become resentful, bitter, alienated, sullen, hostile and violent. The favored groups, unwilling to admit that they receive any more than their share, respond in kind. It is foolish to imagine that relationships among diverse racial or ethnic groups in a democratic society can be cooperative or even peaceful, where racial discrimination is widespread.

2. Should wrongful discrimination be redefined? Instead of forbidding all unequal treatment based on race, we might bar such treatment only when it is motivated by racial prejudice or hostility. This would clear the way for "benign" discrimination in behalf of a previously excluded group, without sacrificing anyone's right to be free from "wrongful" discrimination.

A principal disadvantage to this approach is the extensive harm that it would legalize. A major reason for antidiscrimination laws is to protect people from being deprived of products, services, and opportunities by discriminatory acts. But this deprivation is just as great, whether the discrimination is motivated by prejudice or not. Discrimination is not benign to the person it injures.

Moreover, by making the process of determining guilt more difficult and less reliable, we would impede the law's enforcement. When the offense is defined simply as unequal treatment, its core is an act, for which there are objective criteria, such as whether the defendant employed a double standard. But if the offense is unequal treatment motivated by prejudice, its core is a state of mind; the court must examine the reasons claimed by the discriminator, trying to judge their validity and his honesty. More important, it would be harder for the person who makes a discriminatory choice, and for the person discriminated against, to know whether the choice is unlawful. Since people generally underestimate the influence of prejudice on their own motives, and exaggerate its influence on others', they would tend to assume that their own discriminatory acts were lawful, while discrimination against them was not. Whether for these or other reasons, there has been no serious effort to amend the statutes by this redefinition.

3. Should education and employment be exempted from the racial nondiscrimination rule? No one contends that racial discrimination should be outlawed in every kind of decision: to bar it in choosing a friend, a spouse, or a legislative representative would be invasive or unenforceable. Why not, then, withdraw the prohibition from the two areas in which preferential treatment might be most helpful for members of a previously excluded group, by bringing them more quickly into prestigious occupations and encouraging their fellows to aim higher and work harder?

A decision to exempt education and employment from the ban on discrimination would place both society and government in moral contradiction with themselves. The society, having decided that racial discrimination in general is wrong, would nevertheless be treating it in crucial areas as beneficial. The government, in its roles as educator and employer, would freely practice here that which elsewhere it must prosecute and punish. Such broad contradictions are fatal to the public consensus that racial discrimination is ordinarily unjust, a consensus that is necessary for general adherence to antidiscrimination laws.

Moreover, this option is overwhelmingly opposed by supporters of affirmative action. They fear that if everyone in education and employment may discriminate, and be discriminated against, most of the victims in a predominantly White society would be people of color. They insist that the right of Blacks, Hispanics and Native Americans not to be discriminated against in education and employment be held inviolate.

4. Should racial discrimination be allowed, but only in favor of Blacks? This would respond forthrightly to the moral dilemma posed early in this essay, by making discrimination in behalf of Blacks an exception to the nondiscrimination rule. Such an exception could apply to all areas of life that are covered by the rule, including housing, business, finance, voter registration, shopping, entertainment, criminal and civil justice, etc., as well as education, employment, and government contracting. But an exception this broad, which could easily sink the rule, has no champions. What is proposed instead is to limit the exception primarily to employment and higher education.

The exception faces two ways: Blacks would gain the privilege of favorable discrimination, by themselves or in their behalf; while all others would lose the right not to be racially discriminated against when Blacks are the beneficiaries.

A major argument against this option is the absence of a principled basis for making blacks the only beneficiaries of racial discrimination. If, when the nation decided to ban racial discrimination, Blacks were the only group to have suffered it in the past, a basis for this exception would be clear. But Blacks were not alone. American Indians; Mexicans, Puerto Ricans, and other Latinos; Japanese, Chinese, and other Asians; Poles, Italians, Slavs, Arabs, Jews, and other Whites could all point to group wounds from past discrimination. The historical experiences differed in form and severity, in the duration of effects and in the proportion of members involved, but they all have a place in the spectrum of past injustice. If historical disadvantage warrants discrimination in favor of Blacks, it supports discrimination in favor of other groups as well.

One could try to confine affirmative action to Blacks by emphasizing the unique aspects of their oppression: their abduction and life here as slaves. But American Indians could cite the conquest and obliteration of tribes; Asians, long-lasting restrictions on immigration and citizenship, and so on. Other groups, even if willing to concede that Blacks had suffered most, would scarcely think their own histories worth nothing in the bidding for preferential treatment. For these and other reasons, a policy of confining race-based affirmative action to Blacks has never garnered widespread support.

5. Should racial discrimination be permitted, at least in education and employment, whenever it favors "underrepresented" minorities? It may be argued that this, in effect, is the option we have chosen, not by explicitly revising the nondiscrimination statutes, but by adding race-based affirmative-action programs. Under these, Blacks, Latinos, and Native Americans receive racial preference and are supposedly not discriminated against; Whites do not receive preference and are discriminated against; Asians are sometimes beneficiaries, sometimes victims. That Whites and Asians have lost their right to racial nondiscrimination in these areas is not made explicit. But it is surely implied, by the view that racial preference at their expense is morally permissible when serving a good purpose, and by the argument that they have no more reason to complain when disadvantaged by racial preference, than if the preference had been based on place of residence or family connections.

By rejecting Option 3, which would exempt education and employment from the nondiscrimination rule, we have confirmed that the right to racial nondiscrimination in these two areas is essential. Nevertheless, Option 5 would void this right for all persons who are not "underrepresented" minorities. This change is so drastic that no supporter dares to speak its name. No person advocating, no institution practicing, race-based affirmative action, calls it racial discrimination.

What would justify sacrificing the personal right not to be racially discriminated against in education and employment? I suggest, first, that the arguments for the sacrifice should be overwhelming. Second, the sacrifice should be favored by popular majorities within the groups who are required to make it.

Supporters of racial preference for "underrepresented" minorities in education and employment typically invoke principles of racial justice, such as the right to compensation for past injury and/or a right to equal racial success. We have argued above that the latter right does not exist and the former right, properly applied, does not require special treatment based on race. We have argued also that the plea of public necessity is unfounded.

If arguments from racial justice and public necessity are set aside, there remain arguments based on what might be desirable for the country. Would not America be a better place, one such argument contends, if Black, Hispanic and Native Americans, often ill-treated in the past, were helped through preferential treatment to reduce, perhaps remove, the achievement gaps that separates them from the rest of the population? Not because such help is a group entitlement, or a public necessity, but because it is a good thing to do.

Many Americans, including those whose own right to nondiscrimination is abrogated by racial preference, have nevertheless upheld race-based affirmative action in education and employment as serving a generous and noble cause. Let them set forth-honestly, clearly, realistically-the benefits from race-based affirmative action in education and employment.

Over against these benefits, let them weigh harms like the following:
the rejection of better qualified Whites and Asians in admissions and hiring;
the growing assumption of a right to racial preference among its recipients;
the discrediting of minority attainment as due to favoritism;
the spread of racial discrimination in education from the admission of students to the selection of faculty, and to the content of courses;
the spread of racial discrimination in employment from the hiring of workers to selection for training and promotion;
the reduction of standards in order to accommodate less-qualified applicants and participants;
the growth in racial consciousness, partisanship and resentment in colleges and workplaces.

To these harms, add the deterioration in the moral authority of the racial nondiscrimination rule, spreading beyond education and employment to other social relationships. Also, the increase in racial partisanship, tension, conflict and hostility that results from racial discrimination, especially when practiced by governmental and powerful private institutions. Having never seen a realistic set of benefits from race-based affirmative action in education and employment that would match, let alone overwhelm, the harms identified here and elsewhere in this essay, I believe that Option 5 fails the first part of its justification test.

Certainly, it fails the second part. After long exposure to the policy, White and Asian Americans are emphatically unconvinced that it warrants the revocation of their own right not to be racially discriminated against. They are never asked in so many words whether they approve this sacrifice. But their view of Option 5 is clear enough, in public opinion polls. For example, in 2001, the Washington Post put the question in these words: "In order to give minorities more opportunity, do you believe race or ethnicity should be a factor when deciding who is hired, promoted, or admitted to college, or that hiring, promotions, and college admissions should be based strictly on merit and qualifications other than race or ethnicity?" From 94% of White and 84% of Asian Americans, the answer was that such decisions should be based on merit, not race or ethnicity. [Question #50, Washington Post Racial Attitudes Survey, 7/11/01]


Our inquiry began with a moral dilemma. If all have the right not to be subject to racial discrimination, no one may be assisted via racial preference; if racial preference is authorized for some, the right not to suffer racial discrimination is thereby denied to others. Two ways out of the dilemma were examined.

May the right to racial nondiscrimination, especially in education and employment, though belonging to everyone, be overridden by a superior right or a public necessity? By a right to equal success for racial groups, or to just compensation for past discrimination? Or by a public necessity for racial preference as a means to racial peace, to racial diversity, or to the prevention of discrimination? These supposed rights and necessities were found to be either non-existent, or not in conflict with the right to racial nondiscrimination, and therefore incapable of superseding it.

Should the right to racial nondiscrimination be rescinded or limited, in order to make racial preference available? Five options were considered. The nondiscrimination rule could be scrapped altogether, redefined to cover only prejudiced or hostile acts, dropped from education and employment, or modified in these areas to allow preference for blacks only, or for all "underrepresented" minorities. The arguments against the first four options are so strong that they lack any significant public support. The fifth option commands widespread support, but fails to meet the burden of justification for the sacrifice it requires.

We cannot have the individual and social benefits of the nondiscrimination rule if we decline to obey it. We cannot teach our children that racial discrimination is wrong if we doggedly discriminate. We cannot preserve the right to nondiscrimination by systematically violating it. But, without breaking or bending the rule, we can respond to many people who need and deserve help. The racial nondiscrimination rule does not preclude compensation for specific injury. It does not bar special assistance, by the public or private sector, to persons who labor under social, cultural, educational or economic disadvantages, provided that the purpose of the help and the criteria for eligibility are color-blind.

Besides excluding racial preference, there are other important respects in which a desirable assistance program would not imitate current affirmative action. It would help people increase their ability to meet regular standards, instead of lowering standards to accommodate lack of ability. The role of government would be primarily determined by the legislative branch, not the bureaucracy or the judiciary. The participation of the private sector would be voluntary or contractual, not compulsory. The rules and operation of the program would be honestly described and freely accessible to public scrutiny. These guidelines are not mandates of the nondiscrimination rule, just counsels of good sense. They will be easier to follow in a racial policy that we really believe is right.


*Note: (An earlier version of this essay was published in the journal, Society, March-April 2004, under the title, "Racial Preference versus Nondiscrimination." The present version replaces a much earlier and quite different effort, posted on this website in December 2002 under the title, "Affirmative Action vs Equal Treatment: The Moral Dilemma." ) [return to top]

The moral issues in this essay can be explored further in debates on this website, e.g.:

Is there a moral right to equal racial achievement?
Does Racial Justice Require Racial Equality in Achievement?

Is there a moral right to racial preference as just compensation for past wrong?
See Should Racial Preference Be Granted to Blacks and Other Minorities as A Remedy for Disadvantages Due to Past Discrimination?

Is racial discrimination wrong when not motivated by prejudice or hostility?
See Should Racial Discrimination Be Permitted against Persons Whose Race or Ethnic Group Is Not The Object of Prejudice or Contempt?

Is racial preference for Non-Whites less wrong than racial preference for Whites?
See Does The Constitution Treat Racial Discrimination against Whites and Non-Whites Differently?

Does the goal of more Blacks in elite colleges justify racial preference in their behalf?
See The Facts Show That Racial Preference in College Admissions (Benefits, Harms) America.

Does the goal of more "underrepresented" minorities in the professions justify racial preference in their behalf?
See Should We Strive for Racial Balance in The Learned Professions?