Sandra Day O'Connor
Sandra Day O'Connor; drawing by David Levine

The Michigan affirmative action cases that the Supreme Court heard on April 1, and will probably decide by July, are among the most important in its history.1 Enormous crowds demonstrated in favor of affirmative action outside the Court during the oral argument and more amicus curiae (“friend of the court”) briefs were filed in these cases—by universities, colleges, students, political and military officials, corporations, political action groups, and other interested citizens—than in any preceding case.

The University of Michigan—the defendant in the cases—takes race into account in deciding which students to admit because more traditional criteria would admit only a very small number of black, Hispanic, and Native American students. Its undergraduate admissions office uses a point system. It awards points for test scores, high school record, unusual talents and experiences, and other factors, up to a maximum of 150 points. Twenty points are automatically included in the scores of stipulated minorities, as well as the scores of outstanding athletes and of students with socioeconomic disadvantages, though only one twenty-point boost is allowed for any combination of these features. The Michigan law school does not use a point system, and judges each applicant individually, but it counts minority race as one among many favorable factors (others include a higher-than-usual age and a poor economic background) in that competition.

White students who were denied admission to the undergraduate college and law school sued the university, arguing that these programs are unconstitutional because the Fourteenth Amendment declares that states must give all people “equal protection of the laws,” and the use of race in admissions decisions, even as one factor among many, denies that protection.2 Different federal district courts held, in contradictory decisions, that the undergraduate program is constitutional and that the law school program is unconstitutional.3 The Sixth Circuit Court of Appeals reversed the latter decision, and the Supreme Court agreed to review both the undergraduate and law school admissions policies.

The Court may conceivably distinguish between the two programs and hold that the law school’s policy is constitutionally permissible but the undergraduate admissions program is not, because adding a fixed number of points for race does not pay enough attention to the individual characteristics and situation of each candidate. That distinction would not be justified, because the undergraduate point system, as the oral argument made plain, allows tentative decisions based on points to be reviewed and altered on a more individualized basis. But a split decision would nevertheless not be catastrophic for affirmative action programs, because universities that use point systems could switch to more flexible plans like the law school’s even though these might be more unwieldy and expensive when used in admitting large undergraduate classes. But if the Court were to hold the law school program unconstitutional as well, that might well mean the end of effective affirmative action programs in American colleges and universities,4 and markedly fewer blacks and other minorities in positions of prestige and influence in the US. If so, the decision would prove to be among the most unfortunate and costly the Supreme Court has ever made.


Twenty-five years ago, in its famous Bakke decision, the Court struck down the affirmative action plan of the medical school of the University of California at Davis, which required a fixed quota of minority admissions. Five of the justices ruled, however, that a university could nonetheless take race into consideration as one of a number of factors.5 Since that time the great majority of the best colleges, universities, and professional schools have adopted admissions programs that take race into consideration in order to increase the representation of minorities in their student bodies. But affirmative action has remained very controversial, and voters in some states, including California, have outlawed it in their state universities.

It has been under successful attack in the courts as well. In a series of important decisions, a Supreme Court much more conservative than the Bakke court struck down a variety of racial-preference schemes in nonacademic institutions. In 1988, in the Croson case, for example, it declared unconstitutional a Richmond, Virginia, ordinance that gave preference to minority-owned firms in awarding municipal construction contracts.6 Then, in the 1996 Hopwood case, the US Court of Appeals for the Fifth Circuit applied these Supreme Court decisions to universities. It struck down the University of Texas law school’s affirmative action plan, and declared that race-sensitive admissions polices could not be used to seek racial diversity.7 The Fifth Circuit majority said that the Bakke case did not have the legal force most lawyers had attributed to it, and that even if it did, it would have been overruled by intervening Supreme Court decisions like Croson.

These were unfavorable political and judicial portents for the future of affirmative action in education. But two extraordinary developments in the new Michigan cases may signal a significant shift in public attitudes: a developing conviction that racial diversity in colleges and professional schools is necessary not as compensation to minorities for past discrimination against them, but as a crucial practical contribution to the entire community’s future. First, the Bush administration, which is surely among the most socially conservative in our history, filed an amicus brief that, as expected, asked the Court to declare both Michigan plans unconstitutional. But that brief does not deny the importance of racial diversity in student bodies—on the contrary, it declares that it is “laudable” for universities to do all they can to ensure “that student bodies are experientially diverse and broadly representative of the public.” It argues only that racial diversity can be achieved through admissions strategies that do not appeal explicitly to race, like the Texas plan, adopted after the Hopwood decision, which admits the top 10 percent of graduates of all Texas high schools to that state’s universities. The plan guarantees some black and Hispanic admissions because some of those high schools have an almost exclusively minority student body. The administration’s brief concedes that if, contrary to its view, race-neutral alternatives would not ensure that minorities are represented in universities “then the question whether race could ever be a consideration would arise.”8


The second development is even more striking: the enthusiastic support for affirmative action among America’s largest and most powerful businesses and within the military. Sixty-five of the best-known American corporations, including General Electric, Microsoft, and Coca-Cola, filed an amicus brief urging the Court to uphold both Michigan plans.9 “In the practical experience of the amici businesses, the need for diversity is indeed compelling,” it says, and adds that because of “the increasingly global reach of American business,” these businesses “need the talent and creativity of a workforce that is as diverse as the world around it.” Many other corporations, including IBM and General Motors, filed their own briefs to the same effect.

The most arresting amicus brief of all, in view of the war in Iraq, was filed by a group of twenty-nine retired military and civilian leaders, including General H. Norman Schwarzkopf, who directed the allied forces in the 1991 Gulf War, Robert McFarlane, who was President Reagan’s national security adviser, Admiral William T. Crowe, who was chairman of the Joint Chiefs of Staff from 1985 to 1989, and General Wesley Clark, who was Supreme Allied Commander in Europe from 1997 to 2000.10 The brief declares that racial imbalance in the war in Vietnam—over 10 percent of the servicemen but only 3 percent of the officers were black—caused “increased racial polarization, pervasive disciplinary problems, and racially motivated incidents” that sometimes “reached a point where there was an inability to fight.”

The service academies and ROTC have since instituted “race-conscious recruiting and admissions policies,” and the brief concludes, “At present, the military cannot achieve an officer corps that is both highly qualified and racially diverse” without such policies.11 Several of the justices referred to the military brief in oral argument: they pressed the solicitor general and the other lawyers opposing Michigan to say whether they thought the armed forces’ affirmative action programs are unconstitutional. The lawyers were evasive.

So there is now growing support for the view that affirmative action is of great and general value to the country, because it attacks the economic racial imbalances that have proved so harmful. There is also less apparent support, at least among leading American institutions, for the once-popular view that affirmative action is unfair to white applicants. It is sometimes said that college and university applicants have a right to be judged only on narrow academic criteria, but that cannot be seriously maintained.12 Places in selective universities are not merit badges or prizes for some innate talent or for past performance or industry: they are opportunities that are properly offered to those who show the most promise of future contribution to goals the university rightfully seeks to advance. These goals can be, and historically have been, social as well as more narrowly academic.

Universities say they are training the nation’s and the world’s future leaders: if it is best for the nation that its leaders more closely match the diversity of its citizens, then no one is cheated by universities who include that goal among their aims. Of course, they must not act unjustly or in violation of anyone’s moral rights. In particular, they must not exclude students out of racial or any other form of prejudice or stereotype. But Michigan’s programs do not reflect racial or other prejudices—on the contrary the scholars who choose such programs aim, among other goals, to reduce prejudice in the classroom and in the country.


Even if it is beneficial for universities to use race-sensitive tests in admission, however, and even if no moral principle forbids this, we must still ask whether the Supreme Court’s own precedents require it to strike down the University of Michigan’s plans. The equal protection clause does not, of course, forbid government to make distinctions or classifications among citizens. But the Supreme Court has ruled over many decades that classifications according to race are inherently “suspect” and must therefore be subject to a “strict” scrutiny that imposes the following three tests. Race-sensitive admissions plans must serve some “compelling” goal, universities must not be able to pursue that goal adequately without them, and they must be “narrowly tailored” to achieving that goal.


Do racially sensitive admission plans serve a compelling goal? The briefs supporting the University of Michigan cite two different goals as compelling, and it is important to distinguish between them. The first is the social goal that the corporate and military briefs I mentioned emphasize: equipping more minority students for leadership in order to attack damaging racial stratification in politics, business, the professions, and the military. The second is the educational goal of classroom diversity: universities argue that a racially diverse student body helps them to fulfill their most basic pedagogic functions because students with different backgrounds, experiences, and perspectives can contribute to each other’s education.

The opponents of the Michigan plan challenge both of these goals, but on different grounds. They say that past Supreme Court decisions rule out the first, social, goal because the Court has repeatedly declared that remedying past socioeconomic injustice cannot count as a compelling goal justifying racial classifications. But this objection confuses two different ideas: the backward-looking claim that affirmative action is justified in order to compensate minority students for past injustice to their race, which the Court has rejected, and the forward-looking claim that it is justified in order to improve society in ways that benefit practically everyone.

Several justices have declared, in past cases, that though an institution may use racial classifications to compensate for its own past discrimination, it may not do so to compensate for discrimination by others or in the community as a whole. It is, in fact, doubtful that affirmative action can ever be justified as compensation, because compensation is a matter of individual, not group, entitlement, and allowing black applicants to have preference now cannot compensate generations of blacks who suffered injustice in the past. But the forward-looking social goal is very different: it justifies sensitivity to race not on the basis of any compensatory theory, but on the pragmatic assumption that securing a better racial balance in positions of prestige and influence benefits the community as a whole.13 Universities may properly serve that goal through their admissions procedures, just as they may legitimately choose students in order to provide a better balance in the community between corporate and civil rights lawyers, for example, or between specialist and primary care doctors.

Michigan’s own briefs emphasize the second goal—the educational benefits of classroom diversity—because that goal was unambiguously recognized as compelling by Justice Lewis Powell in his crucial opinion in Bakke. Opponents in the cases also challenge that goal, however. They deny that Powell’s swing vote was a ruling by the Court,14 and they add that though it is indeed desirable that students be different from one another in various ways—in economic and social background, talent, interest, and conviction, for example—there is no further pedagogic value in students being diverse in race as well.

But race matters independently of the other background factors with which it may or may not be correlated, for two reasons. First, the experience of a black person in American society is special, and cannot be duplicated by the experience of a white person of similar economic or social background; it is pedagogically important that the perspectives of that distinct experience be available to students of history, politics, and society who would not otherwise encounter it. Second, learning that race is not correlated with stereotypical perspectives—learning, for example, as Justice Stephen Breyer put it during the oral argument, that a black student may be a rich Exeter graduate and a conservative Republican—is itself of educational value, and that lesson about stereotyping may well be best learned from direct experience. It is controversial whether racial diversity in classrooms produces other pedagogical benefits as well, and how far its overall benefits are offset by disadvantages.15 But in the judgment of the great majority of selective American universities and colleges, the educational benefits of affirmative action outweigh any educational costs.

The second constitutional test asks whether these goals, assuming they are compelling, could be achieved in other ways, that is, without racial classifications. As I have said, the Bush administration argued that Michigan could achieve racial diversity through devices like the Texas plan, which automatically admits a given percentage of the top students of the graduating class of each high school in the state. But these devices are patent subter-fuges. They are as much based on racial criteria, though covertly, as plans that appeal to race openly like Michigan’s; as Justice David Souter said during the oral argument, if the purpose is to increase the percentage of minorities, “then whatever it is, it’s not a race-neutral measure.” Such programs depend on and therefore presuppose the continuation of de facto racial segregation in high schools, moreover, which is in itself objectionable, and they are perverse because they require universities to pass up the more qualified black applicants that openly race-sensitive criteria would admit in favor of less qualified ones from segregated schools.

In any case, the percentage schemes have not produced the racial diversity in those states that have tried them that their universities produced by affirmative action before it was banned,16 and it is unclear (the administration’s brief does not take up the issue) how such a plan could be used for admission to a graduate or professional school rather than an undergraduate program that takes students from high school, or whether it could sensibly be applied to private universities that recruit nationally. The Bush administration’s suggestion is nevertheless revealing because it concedes that racial diversity in education is a compelling goal, and that its only genuine objection to affirmative action plans like Michigan’s is that they are honest.

The third constitutional test I have mentioned asks whether Michigan’s plans are “narrowly tailored” to the goals it cites as compelling. The Court has used that phrase to insist that “suspect” classifications, such as those based on race, should be limited in their scope and duration, and based on narrowly defined strategies rather than what Justice Sandra Day O’Connor once called “amorphous” concerns that might li- cense an indiscriminate use.17 Opponents argue that plans like Michigan’s could be in place indefinitely because black and other minority applicants may continue to score lower on the traditional academic tests that selective universities use. It would be wrong to assume that they will. The number of black applicants to universities whose test scores are as high as white candidates who have been admitted is growing, even though it is not yet high enough to allow many of these universities a significant minority presence. Moreover, the gap between the scores of blacks admitted under affirmative action programs and admitted white candidates is decreasing. In any case, however, the fact that damaging racial imbalances may persist is hardly a reason for prohibiting one of the most effective strategies for reducing those imbalances. As an amicus brief filed by eight of the most prominent American universities and colleges pointed out, “The proper constitutional concern would thus seem to be whether the selected means outlast the interest they are designed to serve, not whether they go on ‘too long’ in some abstract, undefined sense.”18

So Michigan’s admissions policies seem to pass the technical, doctrinal tests of judicial precedent as well as to meet the moral and practical standards I have described. But we must ask a further question: Do they respect the underlying legal principles that the doctrinal tests are meant to enforce? Why are racially sensitive admissions programs subject to such “strict” scrutiny? The most helpful answer was given by Justice O’Connor.19 She said that identifying all uses of race as suspicious, and subjecting them to a strict scrutiny, allows judges to “smoke out” illegitimate uses of race that masquerade as helpful to disadvantaged groups. “Absent searching judicial inquiry into the justification for such race-based measures,” she said, “there is simply no way of determining what classifications are ‘benign’ or ‘remedial’ and what classifications are in fact motivated by illegitimate notions of racial inferiority or simple racial politics.”20

Unconstitutional racial classifications, that is, are those whose genesis and features do not dispel all suspicion of improper personal or institutional motives. The Richmond municipal council that preferred black contractors was itself dominated by black officials, and promoting diversity among contractors is not a traditional goal of city administrations. So suspicion could not be eliminated that the scheme’s purpose was political rather than to secure some benefit for the community as a whole. But there are no such grounds for suspicion about the University of Michigan’s motives. The most prestigious universities have for many decades taken the two goals I cited—providing a student body diverse in many important dimensions, and training students to participate effectively in the community’s political, social, and economic life—as essential parts of their justification and aims. The FCC’s decision to seek a particular racial balance among broadcasters, which the Court also struck down, was suspicious for a different reason: it imposed a racial structure fixed by a federal agency on a particular industry.21 But university affirmative action programs only improve the mix of talented graduates and professionals that government, business, and private individuals may employ; the racial balance in the economy that results is not fixed in advance by any agency, but determined by the ordinary processes of individual choice.22

Michigan’s admissions programs seem models of useful and safe racial sensitivity. They use race in a carefully modulated way that allows individual judgment of candidates one by one, respects all pertinent rights, imposes no racial structure on any section of industry or the community, and helps to achieve crucial national objectives. Nothing in the development and refinement of the programs gives ground for the slightest suspicion that they are guided by racial prejudices or racial politics. If this example of using racial distinctions for important practical goals is not permitted by the Constitution, it is hard to imagine any that would be. Justice Ruth Ginsburg asked, in the oral argument, whether a state could deliberately hire minority guards in order to have a significant presence of such guards in a prison of mainly minority inmates.23 One of the lawyers attacking Michigan’s plans said no, and he was only being consistent.

Of course it is deplorable that America is still plagued by racial inequality so many decades after it committed itself to ending it. It would compound our failure, however, to forbid what so many of our academic, economic, and political leaders, after a quarter-century of reflective experience, think is our best weapon against that inequality. Colorblindness that has no basis in moral principle and helps only to perpetuate racial stratification is worse than pointless.

Our colleges and universities have, on the whole, served the country brilliantly: the US is widely acknowledged to have the best institutions of higher education in the world. They are committed to academic excellence and the common good, and they have, as institutions, no political goals that compete with these aims. They each decide on admissions strategies for themselves, they impose no strategies on other institutions, and they are supervised by courts alert to the slightest evidence of improper motives. It would be foolish not to allow them to do what they think urgent for their students and their country.

This Issue

May 15, 2003