Colin S. Diver’s speeches, letters, and articles
University of Illinois, October 2, 2003
From equality to diversity: The detour from Brown to Grutter
Remarks of Reed College President Colin Diver
III. Remediation of race-based discrimination
We are left, I believe, with the conclusion that "diversity" by itself is an inadequate basis upon which to satisfy the very heavy burden for justifying racially discriminatory admissions programs imposed by the Equal Protection Clause. Does that mean that affirmative action in higher education admission should be disbanded in favor of a purely colorblind system? In an ideal world, I believe that the answer is yes. Given our cultural history of racial divisiveness, and perhaps our genetic predisposition to tribalism, race should be viewed as inherently suspect as a criterion for distributing social benefits and burdens, and should be expunged from public life, even private life, insofar as possible. But this is not an ideal world. We still live in a post-Brown world. Only 50 years have passed since the American legal establishment began to take serious steps to declare that system illegal and to dismantle it. Many legally sanctioned forms of racial segregation or discrimination persist to this day. Most of us who lived through the Civil Rights era look back with a mixture of gratification at what has been accomplished to expunge racial discrimination from public institutions and public discourse, and heartsick disappointment at the huge gaps and barriers that persist between the African American population of this country and the non-black population.
What this means is that the remedial task triggered by Brown is still incomplete. Not only are blacks more likely than non-blacks to suffer from various educational, economic, social, health-related, and political disadvantages, but they continue to live, go to school, worship, play, and even work in highly segregated settings. Surely not all of these conditions are fairly and directly traceable to the systems of legally imposed or legally sanctioned disadvantage condemned by Brown and its progeny. In my opinion, the Supreme Court was right, in Bakke and its progeny, to reject the argument that societal remediation can be deployed as a constitutional justification for every form of race-conscious affirmative action practiced in post-Jim Crow America. But the Court was wrong, in my view, in restricting the domain of remedial race-consciousness so narrowly: namely, to those specific institutions that had themselves been adjudicated to have engaged in racially discriminatory practices.
The vestiges of legally mandated and legally sanctioned discrimination against blacks persist to this day in ways that make defensible some degree of race consciousness in higher education admissions. Perhaps the clearest example is the persistence of educationally deprived, predominantly minority urban school systems. Another example is the high concentration of minorities in economically and socially bankrupt rural and inner-city neighborhoods where crime, drug abuse, and social disintegration destroy incentives for academic achievement far more effectively than beleaguered schools can hope to create them. Another possible vestige is the growing body of evidence suggesting that teachers of all races tend to demand less of black students because of lowered expectations of black students' academic capabilities. The persistence of these phenomena strongly suggests that race consciousness may still be justified as one component of a strategy designed to address these deficits.
A. Fashioning a remedial rationale for race-conscious admissions
The strongest argument made by the Supreme Court against using societal remedy as a justification for reverse discrimination is that it is "too amorphous" to be subject to judicial control or limitation. As the Court famously said in Wygant:
In the absence of particularized findings, a court could uphold remedies that are ageless in their reach into the past, and timeless in their ability to affect the future.
There are, it seems to me, two responses to this argument. First, the diversity rationale embraced by the Court in Grutter is even more amorphous. As I have attempted to demonstrate above, the diversity argument is too conceptually muddled and inconsistently applied to sustain the demanding burden of proof applied to explicit racial classifications. Further, the diversity rationale is even more "timeless" than the remedial justification. The latter, at least, affords a theoretical basis upon which to terminate a race-conscious program-namely, when the structural vestiges of previous legalized discrimination have been dismantled and the remaining interracial disparities are primarily attributable to other factors. If diversity is the rationale, there is no inherent theoretical time limit on a race-conscious admissions program. So long as educators believe that racial mixing of students produces educational benefits, the use of race to achieve the desired mix continues to be justified. The goal of diversity per se tells us nothing about whether there is any logical ending point. As I explain later, Justice O'Connor's forlorn attempt, at the end of her Grutter opinion, to impose a time limit on Michigan's program demonstrates that she must really have remedial objectives in mind.
Second, the Court has never really tried to construct a set of standards that might discipline the use of a more targeted social remediation rationale. Without having done so, how can it know that such a rationale is ungovernable? In fact, if the Court had sanctioned its use in 1978, when Bakke was decided, by now we would have a well-developed jurisprudence of remediation in the context of higher education admissions. That jurisprudence would be driven, first, by the fact that educators would have a heavy burden of justifying the use of race to remedy societal discrimination. Intent, as educators obviously are, to admit racially diverse student bodies, and commanding, as they do, vast investigative resources, I have little doubt that the higher education establishment would have risen to the challenge with theoretical arguments and sociological studies of considerable sophistication.
Applying the high burden of proof implicit in the compelling state interest standard, and lessons learned from institution-specific remedial orders, the courts could have fashioned a set of presumptions that would discipline such offers of proof. Thus, for example, racial preferences afforded in the admission process would need to be conditioned on a showing that the applicant comes from an educational background or social context still substantially shaped by a regime of legally mandated or legally sanctioned discrimination. An example might be a black student from a school district which had been found to engage in discriminatory practices aimed at blacks and in which educational performance by black students continue to lag significantly behind similarly situated majority students in that district or similarly situated black students in other districts. Another example might be a black student from a school or district in which the adverse educational effects of "lowered expectation" discrimination had been credibly demonstrated.
Beyond demonstrating some plausible link between past patterns of race-based discrimination and current educational deficit, practitioners of racial preferences would also be required to justify the weight given to race and explain what they are doing to mitigate the adverse consequences of academic mismatch, such as problems of academic underperformance. This requirement would impose on educators a degree of candor about the costs and benefits of their programs that has been noticeably missing from the subject. But it would also focus attention on what is really important-not the quantitative racial composition of the entering class, but the (net) educational benefits afforded to those to whom preferences have been given. The much-debated concept of "critical mass" provides an illustration. In Grutter, the University of Michigan Law School argued that it sought to enroll a sufficient number of black students to dispel white stereotypes that all blacks think or act alike. Given the undergraduate educational backgrounds of the white students who enroll at Michigan's law school, and their educational exposure to the views of black jurists ranging from Thurgood Marshall to Clarence Thomas, the premise for this rationalization seems implausible on its face. Worse still, however, the law school's argument illustrates the unattractive tendency of diversity justifications to sanction the use of minority students for the education of whites. By contrast, remediation can provide a far more satisfying, and constitutionally compelling, rationale for the "critical mass" theory. Availing itself of such a rationale, an educational institution would need to show that effectively combating the adverse educational effects of social isolation requires the enrollment of a minimum number of black students.
B. Does the choice of rationale make a difference?
One might ask whether it really matters whether educational preferences are justified by reference to diversity or social remediation. The cynic might argue that, either way, the clever folks at elite educational institutions will justify doing whatever they want,-or perhaps that ideologically conservative judges will find a way to strike down whatever they want. The cynic would be wrong. The rationale does make a difference, for several important reasons.
1. Moral Justification: First, because the remedial rationale rests on a stronger moral foundation, it is more likely to command the degree of public acceptance necessary to sustain affirmative action programs over time. The remedial justification invokes a widely shared notion of compensatory justice: namely, that those who have been wrongfully disadvantaged should be compensated. In his Bakke opinion, Justice Powell argued that the societal remedy theory fails the compensatory justice test because it breaks the connection between the wrongdoer and the persons burdened with the obligation to remedy that wrong. It is true that, as conventionally understood, the philosophical notion of corrective justice requires such a connection. That is, corrective justice claims are "correlative," in the sense that the right of the unjustly harmed person to compensation is correlative with a compensatory duty of the person responsible for the unjust harm.
There are several answers to this objection. First, as a purely pragmatic matter, the connection between the persons unjustly harmed and those forced to bear the burden of compensation is attenuated or broken in many cases of institutional wrongdoing. For example, many of those who bore the burdens of school desegregation, following Brown, had little if anything to do with the creation and enforcement of segregation. Further, there is no necessary incompatibility between the theory of corrective justice and the notion of third-party remediation. As Richard Wright argues: "There is nothing in corrective justice which prevents that [compensatory] duty from being discharged voluntarily, on behalf of the party with the duty, by someone else-e.g., that party's insurer or rich aunt." In a sense, institutions of higher education that practice affirmative action are placing themselves in the position of the "rich aunts" of the actors (for example, school districts) whose discriminatory behavior produced the primary educational injury.
It is true, of course, that the burden of remediation is borne not only by the educational institution adopting a race-conscious admission policy, but also by the non-minority students who are denied admission in order to make a place for the remedially admitted minority applicants. Many defenders of affirmative action have argued that such injuries are too indirect or statistically insignificant to exert much of a moral counterweight. My response to this argument is based on the somewhat different rationale of unjust enrichment. An alternative account of corrective justice holds that one who benefits from a wrong inflicted on another has a moral obligation to relinquish that benefit to the person from whom it was wrongfully denied. By this logic, one could reasonably argue that those who currently benefit from the system of racial oppression outlawed by Brown have no moral entitlement to exclude those who were harmed by that system. If, in absence of racial discrimination in her elementary and secondary education, applicant X would have qualified on her own merits for admission to Michigan, applicant Y, unburdened by discrimination, who was admitted instead, has no compelling moral claim to be admitted in place of X. While the university, itself innocent of past racial oppression, may not be morally obliged to compensate X, it is morally permitted to do so. And that moral permission, moreover, is sufficient to overcome the general moral objection to using race as a classificatory device.
By contrast, the diversity argument rests on a weaker moral foundation. From a moral perspective, the claim that a certain racial mix enhances educational outcomes is derivative on the moral weight to be attached to the value of education. If we strip remedial considerations away, we are left with general dignity or automony claims, to the effect that education increases a person's capacity for self-expression, self-fulfillment, empathy, social cooperation, and the like. The moral case for diversity, then, rests on the premise that a more racially diverse education is even more effective than a less racially diverse education at promoting these goals. This is not a trivial moral claim, but it pales in significance, it seems to me, when set against the corrective justice claim on which the remedial justification rests.
Because the use of race is so freighted with negative moral energy, the strength of the moral justification for a practice like affirmative action is important in its own right. It is also important for political reasons. Public support for affirmative action in higher education is, at best, very fragile. Respondents to public opinion surveys have typically favored such programs when the question is phrased to emphasize the remedial purpose of such programs. By contrast, however, the public has consistently expressed opposition to such programs to the extent that the question emphasizes the use of preferential standards of admission for favored minority groups. This latter fact shows that even remedially justified programs must swim against public opinion. But the removal of the remedial rationale in favor of a diversity rationale exacerbates the problem by stripping the programs of their most popularly acceptable justification.
As Peter Schuck points out, educators in elite universities and colleges seem determined to pursue racially preferential programs despite popular opposition. But lack of public acceptance can undermine race-conscious admissions practices in a variety of ways. The most obvious, and extreme, way is illustrated by legislative measures, such as California's Proposition 209, which forbids public agencies, including state universities, from taking race into account for any purpose. Opponents of affirmative action have promised similar legislative or initiative measures in the wake of Grutter. More subtly, but pervasively, lack of public acceptance can undermine the purposes of such programs, by exacerbating racial relations and strengthening negative racial stereotypes.
2. Candor: A second reason for preferring the remedial justification lies in the virtue of candor. Educational affirmative action programs were developed to serve remedial purposes, and by and large that is still the motivation of most educators who implement such programs. Unfortunately, the rush to embrace the diversity rationale has exacerbated a culture of obfuscation and dissembling that has characterized race-conscious admissions programs since their inception. As if to compound the problem, the Court in Grutter and Gratz has adopted a rule that penalizes candor. Whatever else one might say about the undergraduate point system struck down in Gratz, it had the advantage of making clear the extent to which the university was favoring race, and for that matter other factors, in its admission calculus. By comparison, the Law School's so-called "holistic" process upheld in Grutter covers that question under a shroud penetrable only by exhaustive statistical analysis based on data obtainable only by subpoena. Based on the statistics unearthed by the plaintiffs, and summarized in the dissenting opinion of Chief Justice Rehnquist, it appears that the advantage conferred on minority applicants by the Law School was, in relative terms, as weighty as that conferred by the undergraduate point system. But now, following the Grutter and Gratz decisions, admissions officers are scrambling to make their formulas less quantitative and more "holistic." Whatever the Court means by "holistic," on thing is certain: admissions criteria will be more, not less, obscure. Lack of candor may be justifiable, even beneficial in some contexts. In the early years of educational affirmative action, many educators believed that lack of candor was appropriate, in the interest of avoiding stigmatization of the minority students benefited by preferential admissions programs. That culture bred a whole lexicon of euphemisms about the weight given to race. But it is now-thanks largely to the spate of litigation culminating in Grutter and Gratz-long past the point that one can continue pretending that most elite institutions treat race simply as a tie-breaker or a gentle boost afforded in a handful of cases. Even if it were possible to maintain what Meir Dan-Cohen calls "acoustic separation" between those who practice affirmative action admissions and those whom it affects, dissimulation is particularly inappropriate-and demoralizing-in a social institution like higher education, whose function is, supposedly, the unbiased transmission of knowledge and the relentless pursuit of truth.
3. Disciplining Racial Preferences: Third, the remedial justification provides a better basis for disciplining of the practice of race-conscious admissions in higher education. As discussed earlier, however morally attractive its purposes or educationally beneficial its effects, affirmative action is no less susceptible than other institutional practices to distortion through "capture" by those who administer and benefit from it. The diversity rationale, especially as articulated in Grutter, is too confused, imprecise, and deferential to facilitate effective judicial discipline. The Court has, in effect, given higher education a virtual blank slate to practice discrimination in favor of nonmajoritarian races in the name of educational philosophy.
The remedial justification, by contrast, cuts through the mantle of academic freedom in which the Court has wrapped educational diversity and insists that every practitioner of racial classifications-educator or otherwise-tie its practices to the social imperative of dismantling an immoral social system. Correctly formulated and implemented, the remedial principle obliges those practitioners to formulate and defend theories of social causation that can convincingly connect current racial disparities to that discredited system. Further, the remedial justification would require educators to be much more explicit about the way in which, and the weight with which, it deploys racial characteristics than does the diversity justification. If, as Justice Brandeis famously asserted, "sunlight is the greatest disinfectant," clear articulation should itself serve as a valuable restraint against excess.
The requirement of explicitness should also place pressure on educators to diminish the educational costs that beneficiaries of affirmative action programs often bear in the interest of so-called diversity. Schools that give racial groups substantial preferences in terms of academic predictors would be obliged to measure the academic performance-and particularly the extent of academic underperformance-of the recipients of those preferences. To the extent that such studies document troubling differentials in performance, those schools would be obligated to undertake remedial programs or else to discontinue the granting of such preferences.
A further advantage of the remedial justification is that it forces higher education to examine the cultural and racial biases in the criteria it now uses to allocate positions in its entering classes. The diversity rationale permits higher education simply to ignore the notorious racial disparities in standard aptitude tests and grade-point averages by, in effect, setting different parameters for different racial groups. The remedial strategy requires that the racial differentials in the measures most widely used by higher education be examined and explained. To the extent that those measures reflect social or cultural differences fairly traceable to the segregationist regime, they should be either adjusted in particular cases, or rejected wholesale in favor of cleaner measures.
4. Compatibility with Brown: Finally, the remedial justification, unlike the diversity justification, is faithful to the spirit of Brown. Brown is unquestionably the greatest judicial document on the meaning of equality in post-slavery America. As such, the decision-and, indeed, the aura that has grown around the decision over time-stands as a beacon through the treacherous shoals of tribalism. We need such a beacon precisely because we, as a nation, and probably as a species, are continuously pulled back toward tribalism. Brown can serve that function only to the extent that it remains a symbol of moral clarity, or, in Richard Kluger's words, an embodiment of "simple justice." The moral clarity of Brown stems from its universalism-its denial that people may be separated, or treated separately, simply because of their birthright. The remedial justification keeps our eyes on that principle, even as we use birthright preferences to overcome birthright disadvantages. The diversity rationale now adopted into law by a majority of the Supreme Court severs that connection. It announces a different ideal world, one in which the path to justice, or at least a good career, lies in mixing by birthright. That rhetorical move, it seems to me, is the tragedy of Grutter.
I have criticized the opinion of the Court in Grutter as if it were the product of a single intelligent actor, unencumbered by the constraint of precedent and the need to cobble together a consensus. By that standard, in my judgment, it fails to satisfy the very test-strict scrutiny-that the Court applies to racial classifications. The diversity rationale is ambiguous in concept and unconvincingly executed in practice; it sanctions the use of race for instrumental reasons nominally divorced from the historical conditions that made race so toxic as a criterion for allocating governmental benefits and burdens. In fact, however, the majority opinion in Grutter is the product of compromise among five Justices with differing views on the subject. Some-surely Ginsburg, Stevens, and Souter-would adopt a frankly remedial strategy if they felt that they could garner the necessary votes. Others-most notably O'Connor, and probably Breyer and Kennedy-resist the general remedial objective, consistently rejected by the Court over the past 25 years. But they are prepared to embrace the use of race for some nonremedial purposes, at least (and perhaps only) in the educational context.
If one reads the Grutter majority opinion carefully, however, it seems to me that Justice O'Connor and her colleagues are really embracing a remedial notion. Without some underlying remedial conception, most of what the majority opinion says would be incomplete at best or incoherent at worst. The highly selective way in which the Michigan Law School sought to achieve educational "diversity" cannot satisfy a "narrowly tailoring" test, at least as conventionally understood and previously applied. Only as an effort to correct for the vestiges of de jure racial discrimination does the law school's particular set of choices and methods even begin to make sense. Likewise, the Court's "legitimacy" argument must be predicated on some notion that-regardless of what people think-institutions like the military, corporations, law firms, and universities cannot be truly "legitimate" so long as their composition is still bleached by the racist regime struck down in Brown and its progeny Similarly, the Court's dictum at the end of the end of the opinion regarding time limits can only be reconciled with a remedial perspective.
So we have, on the eve of Brown's fiftieth anniversary, yet another reminder of how deeply the issue of affirmative action divides the contemporary Supreme Court. As a sociological fact, we should hardly be surprised. This is a Court deeply divided on many issues. And there are few issues in America today that are as divisive as the use of racial preferences. Grutter, by its strained logic, reminds us of the path not taken back in 1978, when the Court, sensing a choice between the extremes of racial favoritism and colorblindness, opted for the latter. In so doing, it failed to seek out a middle path, one that would have left open the possibility of developing a forthright rationale for more targeted remedial measures. The diversity justification can be viewed as a fumbling effort to rediscover that pathway. One might have wanted a cleaner justification for the use of racial preferences in educational admissions. But in this imperfect world, an imperfect justification is perhaps the best we can realistically hope for.
 See Richard Dawkins, The Selfish Gene 100 (1989 ed.):"Conceivably, racial prejudice could be interpreted as an irrational generalization of a kin-selected tendency to identify with individuals physically resembling oneself, and to be nasty to individuals different in appearance."
 I focus primarily on the situation of African Americans because they were the object of the segregationist regime whose dismantling was initiated by Brown. In so doing, I do not mean to suggest that other racial and ethnic have not been subjected to legalized discrimination that can trigger legitimate remedial claims.
 The sociological evidence is surveyed in Ronald F. Ferguson, "Teachers' Perceptions and Expectations and the Black-White Test Score Gap," 38 Urban Education 460 (2003).
 Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 270 (1986).
 See TAN 115, infra.
 539 U.S. at (slip op. at 30-31).
 I include here not only the financial resources, faculty, and students of the elite institutions, but also foundations like the Mellon Foundation, which funded the Bowen and Bok study and the Cole and Barber studies. See Bowen & Bok, supra n.27; Cole & Barber, supra n.67.
 Indeed, it was not until the "diversity" rationale came under sustained attack in the courts, beginning with the Hopwood litigation, that academia produced most of the growing body of sociological evidence on the educational benefits of diversity. See Douglas Lederman, "Backers of Affirmative Action Struggle to Find Research that Will Help in Court," Chronicle of Higher Educ., May 23, 1997, at A28. A prominent example is the Gurin et al. study, which was produced precisely for the purpose of helping Michigan to defend itself in Grutter and Gratz. See Gurin et al., supra n. 37, cited approvingly by the Court in Grutter, 539 U.S. at (slip op. at 18).
 See TAN 103-11, infra.
 See TAN 64-68, supra.
This is, in essence, the argument made by Jeffrey Rosen for changing his view of the appropriateness of race-conscious admissions practices like Michigan's. Jeffrey Rosen, "How I Learned to Love Quotas," N.Y. Times Magazine, June 1, 2003, at 52. The prospect of widespread evasion and sabotage troubled members of the Brown Court, see n. 14, supra, but fortunately did not deter them. The extent of the subversion anticipated by Rosen is paltry by comparison to the resistance stirred by court-ordered desegregation in the wake of Brown.
 See Justice Powell's opinion in Bakke, 438 U.S. at 298: "there is a measure of inequity in forcing innocent persons in respondent's position to bear the burdens of redressing grievances not of their making."
 See, e.g., Jules L. Coleman, "The Practice of Corrective Justice," in David G. Owen, ed., Philosophical Foundations of Tort Law 53, 66-67 (1997 ed.). In this essay, Coleman recants a previous position arguing that corrective justice does not require correlativity. See Jules L. Coleman, "Tort Law and the Demands of Corrective Justice," 67 Ind. L.J. 349 (1992). The correlativity requirement is usually traced to Aristotle's notion of equalizing gains and losses. Aristotle, Nichomachean Ethics V.iv.
 Conversely, the remedial impact of the desegregation orders came too late to help many of the victims.
 Richard W. Wright, "Right, Justice, and Tort Law," in David G. Owen, ed., supra n.97, at 159, 178.
 E.g., Goodwin Liu, "The Causation Fallacy: Bakke and the Basic Arithmetic of Selective Admissions," 100 Mich. L. Rev. 1045 (2002).
 See Feinberg, supra n. 19, at 76-82 ; Schuck, supra n. 19, at 34.
 See, e.g., Chronicle of Higher Education poll (circa March 2003). For summaries of the literature on public attitudes toward affirmative action in admissions, see Heriot, supra n. 66, at 225-28; Rothman et al., supra n. 61, at 32-33; Schuck, supra n. 19, at 55.
 See Schuck, supra n. 19, at 58.
 [Cite post-Grutter statements by Center for Individual Rights. Also California Proposition 54, which would prohibit the state from collecting information on a person's racial or ethnic group.]
 See, e.g., Krieger, supra n. 60, at 1331.
See Gewirtz, supra n.19, at 127 (diversity discourse "can obscure the degree of some of the real trade-offs regardikng other qualifications that affirmative action sometimes does require . . ."); Schuck, supra n. 19, at 71 (lamenting the "pervasive dissimulation and deformation of thought on all sides due to the felt need to deny or ignore the fact that racial preferences play a large, often decisive role in many admissions decisions"). Cf. George Kateb, ____, in Austin Sarat, ed., Race, Law and Culture: Reflections on Brown v. Board of Education ___, 106 (1997) (noting how the Jim Crow regime forced Southern whites to "live the lie" that a segregated society is politically legitimate).
 This concern was also noted by Justice Blackmun, dissenting in Bakke: "The cynical, of course, may say that under a program such as Harvard's, one may accomplish covertly what Davis concedes it does openly." 438 U.S. at 406.
 539 U.S. at (Rehnquist, C.J., dissenting) (Rehnquist slip op. at 3-5.
 [Cite to news articles about Michigan's adoption of a new undergraduate admission system in the wake of Gratz. August 2003.]
 See, e.g, Guido Calabresi & Philip Bobbit, Tragic Choices (19__); Michael Fitts, [Penn Law Rev. article, circa 1991, on the virtues of silence].
See, e.g., Bowen & Bok, supra n. 27, at 265.
 See, e.g., Stephan Thernstrom & Abigail Thernstrom, "Book Review: Reflections on the Shape of the River," 46 U.C.L.A. L. Rev. 1583, 1591 (1999) (noting the euphemisms used by Bowen & Bok to describe racial preferences).
 See n. 66, supra.
 Meir Dan-Cohen, "Decision Rules and Conduct Rules: On Acoustic Separation in Criminal Law," 97 Harv. L. Rev. 625 (1984).
 See TAN 77, supra.
 Justice Scalia virtually invites critics to challenge universities and colleges that "talk the talk of multiculturalism and racial diversity in the courts but walk the walk of tribalism and racial segregation on their campuses . . ." 539 U.S. at (Scalia, J., concurring in part and dissenting in part) (Scalia slip op. at 3). But, once having expunged anything that smacks of point systems and quotas, educators should have little trouble rationalizing most of their current practices and even the use of more widespread or aggressive racial preferences in the future.
 Kluger, supra n. 2.
 Breyer and Kennedy voted with the majority in Gratz, suggesting that they do not embrace the societal remediation position, at least in the broad form espoused by the Court's liberal wing. Breyer (by his vote with the Grutter majority) and Kennedy (in his Grutter dissent) both acknowledge, however, that racial diversity can justify the use of racial preferences, so long as they are sufficiently narrowly tailored.
 I use the term "dictum" advisedly. Justice Thomas, joined by Justice Scalia, does his best to characterize the 25-year time limit as a "holding," in the obvious hope that a future Court will interpret it as such. See 539 U.S. at (Thomas, J., concurring in part and dissenting in part) (Thomas slip op. at 2). But it is hard to read the majority's statement-that "[w]e expect that 25 years from now, the use of racial preferences will no longer be necessary . . ."-as anything more than just that, an expectation.