with whom Justice Scalia joins as to Parts I-VII, concurring in part and dissenting in part.
Frederick Douglass, speaking to a group of abolitionists almost 140 years ago, delivered a message lost on today’s majority:
“[I]n regard to the colored people, there is always more that is benevolent, I perceive, than just, manifested towards us. What I ask for the negro is not benevolence, not pity, not sympathy, but simply justice. The American people have always been anxious to know what they shall do with us. ... I have had but one answer from the beginning. Do nothing with us! Your doing with us has already played the mischief with us. Do nothing with us! If the apples will not remain on the tree of *350their own strength, if they are worm-eaten at the core, if they are early ripe and disposed to fall, let them fall! . . . And if the negro cannot stand on his own legs, let him fall also. All I ask is, give him a chance to stand on his own legs! Let him alone!... [Y]our interference is doing him positive injury.” What the Black Man Wants: An Address Delivered in Boston, Massachusetts, on 26 January 1865, reprinted in 4 The Frederick Douglass Papers 59, 68 (J. Blassingame & J. McKivigan eds. 1991) (emphasis in original).
Like Douglass, I believe blacks can achieve in every avenue of American life without the meddling of university administrators. Because I wish to see all students succeed whatever their color, I share, in some respect, the sympathies of those who sponsor the type of discrimination advanced by the University of Michigan Law School (Law School). The Constitution does not, however, tolerate institutional devotion to the status quo in admissions policies when such devotion ripens into racial discrimination. Nor does the Constitution countenance the unprecedented deference the Court gives to the Law School, an approach inconsistent with the very concept of “strict scrutiny.”
No one would argue that a university could set up a lower general admissions standard and then impose heightened requirements only on black applicants. Similarly, a university may not maintain a high admissions standard and grant exemptions to favored races. The Law School, of its own choosing, and for its own purposes, maintains an exclusionary admissions system that it knows produces racially disproportionate results. Racial discrimination is not a permissible solution to the self-inflicted wounds of this elitist admissions policy.
The majority upholds the Law School’s racial discrimination not by interpreting the people’s Constitution, but by responding to a faddish slogan of the cognoscenti. Nevertheless, I concur in part in the Court’s opinion. First, I agree with the Court insofar as its decision, which approves of only *351one racial classification, confirms that further use of race in admissions remains unlawful. Second, I agree with the Court’s holding that racial discrimination in higher education admissions will be illegal in 25 years. See ante, at 343 (stating that racial discrimination will no longer be narrowly tailored, or “necessary to further” a compelling state interest, in 25 years). I respectfully dissent from the remainder of the Court’s opinion and the judgment, however, because I believe that the Law School’s current use of race violates the Equal Protection Clause and that the Constitution means the same thing today as it will in 300 months.
I
The majority agrees that the Law School’s racial discrimination should be subjected to strict scrutiny. Ante, at 326. Before applying that standard to this case, I will briefly revisit the Court’s treatment of racial classifications.
The strict scrutiny standard that the Court purports to apply in this case was first enunciated in Korematsu v. United States, 323 U. S. 214 (1944). There the Court held that “[pressing public necessity may sometimes justify the existence of [racial discrimination]; racial antagonism never can.” Id., at 216. This standard of “pressing public necessity” has more frequently been termed “compelling governmental interest,”1 see, e. g., Regents of Univ. of Cal. v. Bakke, 438 U. S. 265, 299 (1978) (opinion of Powell, J.). A majority of the Court has validated only two circumstances where “pressing public necessity” or a “compelling state interest” can possibly justify racial discrimination by state actors. First, the lesson of Korematsu is that national security constitutes a “pressing public necessity,” though the government’s use of race to advance that objective must be narrowly tailored. Second, the Court has recognized as a compelling state interest a government’s effort to remedy *352past discrimination for which it is responsible. Richmond v. J A. Croson Co., 488 U. S. 469, 504 (1989).
The contours of “pressing public necessity” can be further discerned from those interests the Court has rejected as bases for racial discrimination. For example, Wygant v. Jackson Bd. of Ed., 476 U. S. 267 (1986), found unconstitutional a collective-bargaining agreement between a school board and a teachers’ union that favored certain minority races. The school board defended the policy on the grounds that minority teachers provided “role models” for minority students and that a racially “diverse” faculty would improve the education of all students. See Brief for Respondents, O. T. 1984, No. 84-1340, pp. 27-28; 476 U. S., at 315 (Stevens, J., dissenting) (“[A]n integrated faculty will be able to provide benefits to the student body that could not be provided by an all-white, or nearly all-white, faculty”). Nevertheless, the Court found that the use of race violated the Equal Protection Clause, deeming both asserted state interests insufficiently compelling. Id., at 275-276 (plurality opinion); id., at 295 (White, J., concurring in judgment) (“None of the interests asserted by the [school board]... justify this racially discriminatory layoff policy”).2
An even greater governmental interest involves the sensitive role of courts in child custody determinations. In Palmore v. Sidoti, 466 U. S. 429 (1984), the Court held that even the best interests of a child did not constitute a compelling state interest that would allow a state court to award custody to the father because the mother was in a mixed-race marriage. Id., at 433 (finding the interest “substantial” but *353holding the custody decision could not be based on the race of the mother’s new husband).
Finally, the Court has rejected an interest in remedying general societal discrimination as a justification for race discrimination. See Wygant, supra, at 276 (plurality opinion); Croson, 488 U. S., at 496-498 (plurality opinion); id., at 520-521 (Scalia, J., concurring in judgment). “Societal discrimination, without more, is too amorphous a basis for imposing a racially classified remedy” because a “court could uphold remedies that are ageless in their reach into the past, and timeless in their ability to affect the future.” Wygant, supra, at 276 (plurality opinion). But see Gratz v. Bollinger, ante, p. 298 (Ginsburg, J., dissenting).
Where the Court has accepted only national security, and rejected even the best interests of a child, as a justification for racial discrimination, I conclude that only those measures the State must take to provide a bulwark against anarchy, or to prevent violence, will constitute a “pressing public necessity.” Cf. Lee v. Washington, 390 U. S. 333, 334 (1968) (per curiam) (Black, J., concurring) (indicating that protecting prisoners from violence might justify narrowly tailored racial discrimination); Croson, supra, at 521 (Scalia, J., concurring in judgment) (“At least where state or local action is at issue, only a social emergency rising to the level of imminent danger to life and limb . . . can justify [racial discrimination]”).
The Constitution abhors classifications based on race, not only because those classifications can harm favored races or are based on illegitimate motives, but also because every time the government places citizens on racial registers and makes race relevant to the provision of burdens or benefits, it demeans us all. “Purchased at the price of immeasurable human suffering, the equal protection principle reflects our Nation’s understanding that such classifications ultimately have a destructive impact on the individual and our society.” *354Adarand, Constructors, Inc. v. Peña, 515 U. S. 200, 240 (1995) (Thomas, J., concurring in part and concurring in judgment).
II
Unlike the majority, I seek to define with precision the interest being asserted by the Law School before determining whether that interest is so compelling as to justify racial discrimination. The Law School maintains that it wishes to obtain “educational benefits that flow from student body diversity,” Brief for Respondent Bollinger et al. 14. This statement must be evaluated carefully, because it implies that both “diversity” and “educational benefits” are components of the Law School’s compelling state interest. Additionally, the Law School’s refusal to entertain certain changes in its admissions process and status indicates that the compelling state interest it seeks to validate is actually broader than might appear at first glance.
Undoubtedly there are other ways to “better” the education of law students aside from ensuring that the student body contains a “critical mass” of underrepresented minority students. Attaining “diversity,” whatever it means,3 is the *355mechanism by which the Law School obtains educational benefits, not an end of itself. The Law School, however, apparently believes that only a racially mixed student body can lead to the educational benefits it seeks. How, then, is the Law School’s interest in these allegedly unique educational “benefits” not simply the forbidden interest in “racial balancing,” ante, at 330, that the majority expressly rejects?
A distinction between these two ideas (unique educational benefits based on racial aesthetics and race for its own sake) is purely sophistic — so much so that the majority uses them interchangeably. Compare ante, at 328 (“[T]he Law School has a compelling interest in attaining a diverse student body”), with ante, at 333 (referring to the “compelling interest in securing the educational benefits of a diverse student body” (emphasis added)). The Law School’s argument, as facile as it is, can only be understood in one way: Classroom aesthetics yields educational benefits, racially discriminatory admissions policies are required to achieve the right racial mix, and therefore the policies are required to achieve the educational benefits. It is the educational benefits that are the end, or allegedly compelling state interest, not “diversity.” But see ante, at 332 (citing the need for “openness and integrity of the educational institutions that provide [legal] training” without reference to any consequential educational benefits).
One must also consider the Law School’s refusal to entertain changes to its current admissions system that might produce the same educational benefits. The Law School adamantly disclaims any race-neutral alternative that would reduce “academic selectivity,” which would in turn “require the Law School to become a very different institution, and to sacrifice a core part of its educational mission.” Brief for Respondent Bollinger et al. 33-36. In other words, the Law School seeks to improve marginally the education it offers *356without sacrificing too much of its exclusivity and elite status.4
The proffered interest that the majority vindicates today, then, is not simply “diversity.” Instead the Court upholds the use of racial discrimination as a tool to advance the Law School’s interest in offering a marginally superior education while maintaining an elite institution. Unless each constituent part of this state interest is of pressing public necessity, the Law School’s use of race is unconstitutional. I find each of them to fall far short of this standard.
III
A
A close reading of the Court s opinion reveals that all of its legal work is done through one conclusory statement: The Law School has a “compelling interest in securing the educational benefits of a diverse student body.” Ante, at 33B. No serious effort is made to explain how these benefits fit with the state interests the Court has recognized (or rejected) as compelling, see Part I, supra, or to place any theoretical constraints on an enterprising court’s desire to discover still more justifications for racial discrimination. In the absence of any explanation, one might expect the Court to fall back on the judicial policy of stare decisis. But the Court eschews even this weak defense of its holding, shunning an analysis of the extent to which Justice Powell’s opinion in Regents of Univ. of Cal. v. Bakke, 438 U. S. 265 (1978), *357is binding, ante, at 325, in favor of an unfounded wholesale adoption of it.
Justice Powell’s opinion in Bakke and the Court’s decision today rest on the fundamentally flawed proposition that racial discrimination can be contextualized so that a goal, such as classroom aesthetics, can be compelling in one context but not in another. This “we know it when we see it” approach to evaluating state interests is not capable of judicial application. Today, the Court insists on radically expanding the range of permissible uses of race to something as trivial (by comparison) as the assembling of a law school class. I can only presume that the majority’s failure to justify its decision by reference to any principle arises from the absence of any such principle. See Part VI, infra.
B
Under the proper standard, there is no pressing public necessity in maintaining a public law school at all and, it follows, certainly not an elite law school. Likewise, marginal improvements in legal education do not qualify as a compelling state interest.
1
While legal education at a public university may be good policy or otherwise laudable, it is obviously not a pressing public necessity when the correct legal standard is applied. Additionally, circumstantial evidence as to whether a state activity is of pressing public necessity can be obtained by asking whether all States feel compelled to engage in that activity. Evidence that States, in general, engage in a certain activity by no means demonstrates that the activity constitutes a pressing public necessity, given the expansive role of government in today’s society. The fact that some fraction of the States reject a particular enterprise, however, creates a presumption that the enterprise itself is not a compelling state interest. In this sense, the absence of a public, American Bar Association (ABA) accredited, law school in *358Alaska, Delaware, Massachusetts, New Hampshire, and Rhode Island, see ABA-LSAC Official Guide to ABA-Approved Law Schools (W. Margolis, B. Gordon, J. Pus-karz, & D. Rosenlieb eds. 2004) (hereinafter ABA-LSAC Guide), provides further evidence that Michigan’s maintenance of the Law School does not constitute a compelling state interest.
2
As the foregoing makes clear, Michigan has no compelling interest in having a law school at all, much less an elite one. Still, even assuming that a State may, under appropriate circumstances, demonstrate a cognizable interest in having an elite law school, Michigan has failed to do so here.
This Court has limited the scope of equal protection review to interests and activities that occur within that State’s jurisdiction. The Court held in Missouri ex rel. Gaines v. Canada, 305 U. S. 337 (1938), that Missouri could not satisfy the demands of “separate but equal” by paying for legal training of blacks at neighboring state law schools, while maintaining a segregated law school within the State. The equal protection
“obligation is imposed by the Constitution upon the States severally as governmental entities, — each responsible for its own laws establishing the rights and duties of persons within its borders. It is an obligation the burden of which cannot be cast by one State upon another, and no State can be excused from performance by what another State may do or fail to do. That separate responsibility of each State within its own sphere is of the essence of statehood maintained under our dual system.” Id., at 350 (emphasis added).
The Equal Protection Clause, as interpreted by the Court in Gaines, does not permit States to justify racial discrimination on the basis of what the rest of the Nation “may do or fail to do.” The only interests that can satisfy the Equal *359Protection Clause’s demands are those found within a State’s jurisdiction.
The only cognizable state interests vindicated by operating a public law school are, therefore, the education of that State’s citizens and the training of that State’s lawyers. James Campbell’s address at the opening of the Law Department at the University of Michigan on October 3, 1859, makes this clear:
“It not only concerns the State that every one should have all reasonable facilities for preparing himself for any honest position in life to which he may aspire, but it also concerns the community that the Law should be taught and understood. . .. There is not an office in the State in which serious legal inquiries may not frequently arise.... In all these matters, public and private rights are constantly involved and discussed, and ignorance of the Law has frequently led to results deplorable and alarming.... [I]n the history of this State, in more than one instance, that ignorance has led to unlawful violence, and the shedding of innocent blood.” E. Brown, Legal Education at Michigan 1859-1959, pp. 404-406 (1959) (emphasis added).
The Law School today, however, does precious little training of those attorneys who will serve the citizens of Michigan. In 2002, graduates of the Law School made up less than 6% of applicants to the Michigan bar, Michigan Lawyers Weekly, available at http://www.michiganlawyersweekly.com/ barpassers0202.cfm,barpassers0702 .cfm (all Internet materials as visited June 13, 2003, and available in Clerk of Court’s case file), even though the Law School’s graduates constitute nearly 30% of all law students graduating in Michigan. Ibid. Less than 16% of the Law School’s graduating class elects to stay in Michigan after law school. ABA-LSAC Guide 427. Thus, while a mere 27% of the Law School’s 2002 entering class is from Michigan, see University of *360Michigan Law School Website, available at http://www.law. umich.edu/prospectivestudents/Admissions/index.htm, only half of these, it appears, will stay in Michigan.
In sum, the Law School trains few Michigan residents and overwhelmingly serves students, who, as lawyers, leave the State of Michigan. By contrast, Michigan’s other public law school, Wayne State University Law School, sends 88% of its graduates on to serve the people of Michigan. ABA-LSAC Guide 775. It does not take a social scientist to conclude that it is precisely the Law School’s status as an elite institution that causes it to be a waystation for the rest of the country’s lawyers, rather than a training ground for those who will remain in Michigan. The Law School’s decision to be an elite institution does little to advance the welfare of the people of Michigan or any cognizable interest of the State of Michigan.
Again, the fact that few States choose to maintain elite law schools raises a strong inference that there is nothing compelling about elite status. Arguably, only the public law schools of the University of Texas, the University of California, Berkeley (Boalt Hall), and the University of Virginia maintain the same reputation for excellence as the Law School.5 Two of these States, Texas and California, are so large that they could reasonably be expected to provide elite legal training at a separate law school to students who will, in fact, stay in the State and provide legal services to its citizens. And these two schools far outshine the Law School in producing in-state lawyers. The University of Texas, for example, sends over three-fourths of its graduates on to work in the State of Texas, vindicating the State’s interest (compelling or not) in training Texas’ lawyers. Id., at 691.
*3613
Finally, even if the Law School’s racial tinkering produces tangible educational benefits, a marginal improvement in legal education cannot justify racial discrimination where the Law School has no compelling interest either in its existence or in its current educational and admissions policies.
IV
The interest in remaining elite and exclusive that the majority thinks so obviously critical requires the use of admissions “standards” that, in turn, create the Law School’s “need” to discriminate on the basis of race. The Court validates these admissions standards by concluding that alternatives that would require “a dramatic sacrifice of . . . the academic quality of all admitted students,” ante, at 340, need not be considered before racial discrimination can be employed.6 In the majority’s view, such methods are not required by the “narrow tailoring” prong of strict scrutiny because that inquiry demands, in this context, that any race-neutral alternative work “'about as well.’” Ante, at 339 (quoting Wygant, 476 U. S., at 280, n. 6). The majority errs, however, because race-neutral alternatives must only be “workable,” ante, at 339, and do “about as well” in vindicating the compelling state interest. The Court never explicitly holds that the Law School’s desire to retain the status quo in “academic selectivity” is itself a compelling state interest, and, as I have demonstrated, it is not. See Part III-B, supra. Therefore, the Law School should be forced to choose between its classroom aesthetic and its exclusionary admissions system — it cannot have it both ways.
With the adoption of different admissions methods, such as accepting all students who meet minimum qualifications, *362see Brief for United States as Amicus Curiae 13-14, the Law School could achieve its vision of the racially aesthetic student body without the use of racial discrimination. The Law School concedes this, but the Court holds, implicitly and under the guise of narrow tailoring, that the Law School has a compelling state interest in doing what it wants to do. I cannot agree. First, under strict scrutiny, the Law School’s assessment of the benefits of racial discrimination and devotion to the admissions status quo are not entitled to any sort of deference, grounded in the First Amendment or anywhere else. Second, even if its “academic selectivity” must be maintained at all costs along with racial discrimination, the Court ignores the fact that other top law schools have succeeded in meeting their aesthetic demands without racial discrimination.
A
The Court bases its unprecedented deference to the Law School — a deference antithetical to strict scrutiny — on an idea of “educational autonomy” grounded in the First Amendment. Ante, at 329. In my view, there is no basis for a right of public universities to do what would otherwise violate the Equal Protection Clause.
The constitutionalization of “academic freedom” began with the concurring opinion of Justice Frankfurter in Sweezy v. New Hampshire, 354 U. S. 234 (1957). Sweezy, a Marxist economist, was investigated by the Attorney General of New Hampshire on suspicion of being a subversive. The prosecution sought, inter alia, the contents of a lecture Sweezy had given at the University of New Hampshire. The Court held that the investigation violated due process. Id., at 254.
Justice Frankfurter went further, however, reasoning that the First Amendment created a right of academic freedom that prohibited the investigation. Id., at 256-267 (opinion concurring in result). Much of the rhetoric in Justice Frankfurter’s opinion was devoted to the personal right of Sweezy to free speech. See, e. g., id., at 265 (“For a citizen to be *363made to forego even a part of so basic a liberty as his political autonomy, the subordinating interest of the State must be compelling”). Still, claiming that the United States Reports "need not be burdened with proof,” Justice Frankfurter also asserted that a “free society” depends on “free universities” and “[t]his means the exclusion of governmental intervention in the intellectual life of a university.” Id., at 262. According to Justice Frankfurter: “It is the business of a university to provide that atmosphere which is most conducive to speculation, experiment and creation. It is an atmosphere in which there prevail ‘the four essential freedoms’ of a university — to determine for itself on academic grounds who may teach, what may be taught, how it shall be taught, and who may be admitted to study.” Id., at 263 (citation omitted).
In my view, “[i]t is the business” of this Court to explain itself when it cites provisions of the Constitution to invent new doctrines — including the idea that the First Amendment authorizes a public university to do what would otherwise violate the Equal Protection Clause. The majority fails in its summary effort to prove this point. The only source for the Court’s conclusion that public universities are entitled to deference even within the confines of strict scrutiny is Justice Powell’s opinion in Bakke. Justice Powell, for his part, relied only on Justice Frankfurter’s opinion in Sweezy and the Court’s decision in Keyishian v. Board of Regents of Univ. of State of N. Y., 385 U. S. 589 (1967), to support his view that the First Amendment somehow protected a public university’s use of race in admissions. Bakke, 438 U. S., at 312. Keyishian provides no answer to the question whether the Fourteenth Amendment’s restrictions are relaxed when applied to public universities. In that case, the Court held that state statutes and regulations designed to prevent the “appointment or retention of ‘subversive’ persons in state employment,” 385 U. S., at 592, violated the First Amendment for vagueness. The statutes covered all public employees and were not invalidated only as applied to uni*364versity faculty members, although the Court appeared sympathetic to the notion of academic freedom, calling it a “special concern of the First Amendment.” Id., at 603. Again, however, the Court did not relax any independent constitutional restrictions on public universities.
I doubt that when Justice Frankfurter spoke of governmental intrusions into the independence of universities, he was thinking of the Constitution’s ban on racial discrimination. The majority’s broad deference to both the Law School’s judgment that racial aesthetics leads to educational benefits and its stubborn refusal to alter the status quo in admissions methods finds no basis in the Constitution or decisions of this Court.
B
1
The Court’s deference to the Law School’s conclusion that its racial experimentation leads to educational benefits will, if adhered to, have serious collateral consequences. The Court relies heavily on social science evidence to justify its .deference. See ante, at 330-332; but see also Rothman, Lipet, & Nevitte, Racial Diversity Reconsidered, 151 Public Interest 25 (2003) (finding that the racial mix of a student body produced by racial discrimination of the type practiced by the Law School in fact hinders students’ perception of academic quality). The Court never acknowledges, however, the growing evidence that racial (and other sorts) of heterogeneity actually impairs learning among black students. See, e. g., Flowers & Pascarella, Cognitive Effects of College Racial Composition on African American Students After 3 Years of College, 40 J. of College Student Development 669, 674 (1999) (concluding that black students experience superior cognitive development at Historically Black Colleges (HBCs) and that, even among blacks, “a substantial diversity moderates the cognitive effects of attending an HBC”); Allen, The Color of Success: African-American College Stu*365dent Outcomes at Predominantly White and Historically Black Public Colleges and Universities, 62 Harv. Edue. Rev. 26, 35 (1992) (finding that black students attending HBCs report higher academic achievement than those attending predominantly white colleges).
At oral argument in Gratz v. Bollinger, ante, p. 244, counsel for respondents stated that “most every single one of [the HBCs] do have diverse student bodies.” Tr. of Oral Arg. in No. 02-516, p. 52. What precisely counsel meant by “diverse” is indeterminate, but it is reported that in 2000 at Morehouse College, one of the most distinguished HBCs in the Nation, only 0.1% of the student body was white, and only 0.2% was Hispanic. College Admissions Data Handbook 2002-2003, p. 613 (43d ed. 2002) (hereinafter College Admissions Data Handbook). And at Mississippi Valley State University, a public HBC, only 1.1% of the freshman class in- 2001 was white. Id., at 603. If there is a “critical mass” of whites at these institutions, then “critical mass” is indeed a very small proportion.
The majority grants deference to the Law School’s “assessment that diversity will, in fact, yield educational benefits,” ante, at 328. It follows, therefore, that an HBC’s assessment that racial homogeneity will yield educational benefits would similarly be given deference.7 An HBC’s rejection of white applicants in order to maintain racial homogeneity seems permissible, therefore, under the majority’s view of the Equal Protection Clause. But see United States v. Fordice, 505 U. S. 717, 748 (1992) (Thomas, J., concurring) (“Obviously, a State cannot maintain . . . traditions by closing particular institutions, historically white or historically black, to particular racial groups”). Contained within today’s majority opinion is the seed of a new constitutional *366justification for a concept I thought long and rightly rejected — racial segregation.
2
Moreover one would think, in light of the Court’s decision in United States v. Virginia, 518 U. S. 515 (1996), that before being given license to use racial discrimination, the Law School would be required to radically reshape its admissions process, even to the point of sacrificing some elements of its character. In Virginia, a majority of the Court, without a word about academic freedom, accepted the all-male Virginia Military Institute’s (VMI) representation that some changes in its “adversative” method of education would be required with the admission of women, id., at 540, but did not defer to VMI’s judgment that these changes would be too great. Instead, the Court concluded that they were “manageable.” Id., at 551, n. 19. That case involved sex discrimination, which is subjected to intermediate, not strict, scrutiny. Id., at 533; Craig v. Boren, 429 U. S. 190, 197 (1976). So in Virginia, where the standard of review dictated that greater flexibility be granted to VMI’s educational policies than the Law School deserves here, this Court gave no deference. Apparently where the status quo being defended is that of the elite establishment — here the Law School — rather than a less fashionable Southern military institution, the Court will defer without serious inquiry and without regard to the applicable legal standard.
C
Virginia is also notable for the fact that the Court relied on the “experience” of formerly single-sex institutions, such as the service academies, to conclude that admission of women to VMI would be “manageable.” 518 U. S., at 544-545. Today, however, the majority ignores the “experience” of those institutions that have been forced to abandon explicit racial discrimination in admissions.
*367The sky has not fallen at Boalt Hall at the University of California, Berkeley, for example. Prior to Proposition 209’s adoption of Cal. Const., Art. 1, § 31(a), which bars the State from “granting] preferential treatment ... on the basis of race ... in the operation of... public education,”8 Boalt Hall enrolled 20 blacks and 28 Hispanics in its first-year class for 1996. In 2002, without deploying express racial discrimination in admissions, Boalt’s entering class enrolled 14 blacks and 36 Hispanics.9 University of California Law and Medical School Enrollments, available at http://www.ucop.edu/ acadadv/datamgmt/lawmed/law-enrolls-eth2.html. Total underrepresented minority student enrollment at Boalt Hall now exceeds 1996 levels. Apparently the Law School cannot be counted on to be as resourceful. The Court is willfully blind to the very real experience in California and elsewhere, which raises the inference that institutions with “reputation[s] for excellence,” ante, at 339, rivaling the Law School’s have satisfied their sense of mission without resorting to prohibited racial discrimination.
V
Putting aside the absence of any legal support for the majority’s reflexive deference, there is much to be said for the view that the use of tests and other measures to “predict” academic performance is a poor substitute for a system that gives every applicant a chance to prove he can succeed in the study of law. The rallying cry that in the absence of racial discrimination in admissions there would be a true *368meritocracy ignores the fact that the entire process is poisoned by numerous exceptions to “merit.” For example, in the national debate on racial discrimination in higher education admissions, much has been made of the fact that elite institutions utilize a so-called “legacy” preference to give the children of alumni an advantage in admissions. This, and other, exceptions to a “true” meritocracy give the lie to protestations that merit admissions are in fact the order of the day at the Nation’s universities. The Equal Protection Clause does not, however, prohibit the use of unseemly legacy preferences or many other kinds of arbitrary admissions procedures. What the Equal Protection Clause does prohibit are classifications made on the basis of race. So while legacy preferences can stand under the Constitution, racial discrimination cannot.10 I will not twist the Constitution to invalidate legacy preferences or otherwise impose my vision of higher education admissions on the Nation. The majority should similarly stay its impulse to validate faddish racial discrimination the Constitution clearly forbids.
In any event, there is nothing ancient, honorable, or constitutionally protected about “selective” admissions. The University of Michigan should be well aware that alternative methods have historically been used for the admission of students, for it brought to this country the German certificate system in the late-19th century. See H. Wechsler, The Qualified Student 16-39 (1977) (hereinafter Qualified Student). Under this system, a secondary school was certified by a university so that any graduate who completed the course offered by the school was offered admission to the university. The certification regime supplemented, and later virtually replaced (at least in the Midwest), the prior regime of rigor*369ous subject-matter entrance examinations. Id., at 57-58. The facially race-neutral “percent plans” now used in Texas, California, and Florida, see ante, at 340, are in many ways the descendents of the certificate system.
Certification was replaced by selective admissions in the beginning of the 20th century, as universities sought to exercise more control over the composition of their student bodies. Since its inception, selective admissions has been the vehicle for racial, ethnic, and religious tinkering and experimentation by university administrators. The initial driving force for the relocation of the selective function from the high school to the universities was the same desire to select racial winners and losers that the Law School exhibits today. Columbia, Harvard, and others infamously determined that they had “too many” Jews, just as today the Law School argues it would have “too many” whites if it could not discriminate in its admissions process. See Qualified Student 155-168 (Columbia); H. Broun & G. Britt, Christians Only: A Study in Prejudice 53-54 (1931) (Harvard).
Columbia employed intelligence tests precisely because Jewish applicants, who were predominantly immigrants, scored worse on such tests. Thus, Columbia could claim (falsely) that “‘[w]e have not eliminated boys because they were Jews and do not propose to do so. We have honestly attempted to eliminate the lowest grade of applicant [through the use of intelligence testing] and it turns out that a good many of the low grade men are New York City Jews.’ ” Letter from Herbert E. Hawkes, dean of Columbia College, to E. B. Wilson, June 16,1922 (reprinted in Qualified Student 160-161). In other words, the tests were adopted with full knowledge of their disparate impact. Cf. DeFunis v. Odegaard, 416 U. S. 312, 335 (1974) (per curiam) (Douglas, J., dissenting).
Similarly no modern law school can claim ignorance of the poor performance of blacks, relatively speaking, on the Law School Admission Test (LSAT). Nevertheless, law schools *370continue to use the test and then attempt to “correct” for black underperformance by using racial discrimination in admissions so as to obtain their aesthetic student body. The Law School’s continued adherence to measures it knows produce racially skewed results is not entitled to deference by this Court. See Part IV, supra. The Law School itself admits that the test is imperfect, as it must, given that it regularly admits students who score at or below 150 (the national median) on the test. See App. 156-203 (showing that, between 1995 and 2000, the Law School admitted 37 students — 27 of whom were black; 31 of whom were “underrepresented minorities” — with LSAT scores of 150 or lower). And the Law School’s amici cannot seem to agree on the fundamental question whether the test itself is useful. Compare Brief for Law School Admission Council as Amicus Curiae 12 (“LSAT scores . . . are an effective predictor of students’ performance in law school”) with Brief for Harvard Black Law Students Association et al. as Amici Curiae 27 (“Whether [the LSAT] measure[s] objective merit... is certainly questionable”).
Having decided to use the LSAT, the Law School must accept the constitutional burdens that come with this decision. The Law School may freely continue to employ the LSAT and other allegedly merit-based standards in whatever fashion it likes. What the Equal Protection Clause forbids, but the Court today allows, is the use of these standards hand-in-hand with racial discrimination. An infinite variety of admissions methods are available to the Law School. Considering all of the radical thinking that has historically occurred at this country’s universities, the Law School’s intractable approach toward admissions is striking.
The Court will not even deign to make the Law School try other methods, however, preferring instead to grant a 25-year license to violate the Constitution. And the same Court that had the courage to order the desegregation of all public schools in the South now fears, on the basis of plati*371tudes rather than principle, to force the Law School to abandon a decidedly imperfect admissions regime that provides the basis for racial discrimination.
VI
The absence of any articulated legal principle supporting the majority’s principal holding suggests another rationale. I believe what lies beneath the Court’s decision today are the benighted notions that one can tell when racial discrimination benefits (rather than hurts) minority groups, see Adarand, 515 U. S., at 239 (Scalia, J., concurring in part and concurring in judgment), and that racial discrimination is necessary to remedy general societal ills. This Court’s precedents supposedly settled both issues, but clearly the majority still cannot commit to the principle that racial classifications are per se harmful and that almost no amount of benefit in the eye of the beholder can justify such classifications.
Putting aside what I take to be the Court’s implicit rejection of Adarand’s holding that beneficial and burdensome racial classifications are equally invalid, I must contest the notion that the Law School’s discrimination benefits those admitted as a result of it. The Court spends considerable time discussing the impressive display of amicus support for the Law School in this case from all corners of society. Ante, at 330-331. But nowhere in any of the filings in this Court is any evidence that the purported “beneficiaries” of this racial discrimination prove themselves by performing at (or even near) the same level as those students who receive no preferences. Cf. Thernstrom & Thernstrom, Reflections on the Shape of the River, 46 UCLA L. Rev. 1583, 1605-1608 (1999) (discussing the failure of defenders of racial discrimination in admissions to consider the fact that its “beneficiaries” are underperforming in the classroom).
The silence in this case is deafening to those of us who view higher education’s purpose as imparting knowledge and skills to students, rather than a communal, rubber-stamp, *372credentialing process. The Law School is not looking for those students who, despite a lower LSAT score or undergraduate grade point average, will succeed in the study of law. The Law School seeks only a facade — it is sufficient that the class looks right, even if it does not perform right.
The Law School tantalizes unprepared students with the promise of a University of Michigan degree and all of the opportunities that it offers. These overmatched students take the bait, only to find that they cannot succeed in the cauldron of competition. And this mismatch crisis is not restricted to elite institutions. See T. Sowell, Race and Culture 176-177 (1994) (“Even if most minority students are able to meet the normal standards at the ‘average’ range of colleges and universities, the systematic mismatching of minority students begun at the top can mean that such students are generally overmatched throughout all levels of higher education”). Indeed, to cover the tracks of the aestheticists, this cruel farce of racial discrimination must continue— in selection for the Michigan Law Review, see University of Michigan Law School Student Handbook 2002-2003, pp. 39-40 (noting the presence of a “diversity plan” for admission to the review), and in hiring at law firms and for judicial clerkships — until the “beneficiaries” are no longer tolerated. While these students may graduate with law degrees, there is no evidence that they have received a qualitatively better legal education (or become better lawyers) than if they had gone to a less “elite” law school for which they were better prepared. And the aestheticists will never address the real problems facing “underrepresented minorities,” 11 instead continuing their social experiments on other people’s children.
*373Beyond the harm the Law School’s racial discrimination visits, upon its test subjects, no social science has disproved the notion that this discrimination “engenders] attitudes of superiority or, alternatively, provoke[s] resentment among those who believe that they have been wronged by the government’s use of race.” Adarand, 515 U. S., at 241 (Thomas, J., concurring in part and concurring in judgment). “These programs stamp minorities with a badge of inferiority and may cause them to develop dependencies or to adopt an attitude that they are ‘entitled’ to preferences.” Ibid.
It is uncontested that each year, the Law School admits a handful of blacks who would be admitted in the absence of racial discrimination. See Brief for Respondent Bollinger et al. 6. Who can differentiate between those who belong and those who do not? The majority of blacks are admitted to the Law School because of discrimination, and because of this policy all are tarred as undeserving. This problem of stigma does not depend on determinacy as to whether those stigmatized are actually the “beneficiaries” of racial discrimination. When blacks take positions in the highest places of government, industry, or academia, it is an open question today whether their skin color played a part in their advancement. The question itself is the stigma — because either racial discrimination did play a role, in which case the person may be deemed “otherwise unqualified,” or it did not, in which case asking the question itself unfairly marks those blacks who would succeed without discrimination. Is this what the Court means by “visibly open”? Ante, at 332.
Finally, the Court’s disturbing reference to the importance of the country’s law schools as training grounds meant to cultivate “a set of leaders with legitimacy in the eyes of the citizenry,” ibid., through the use of racial discrimination deserves discussion. As noted earlier, the Court has soundly *374rejected the remedying of societal discrimination as a justification for governmental use of race. Wygant, 476 U. S., at 276 (plurality opinion); Croson, 488 U. S., at 497 (plurality opinion); id., at 520-521 (Scalia, J., concurring in judgment). For those who believe that every racial disproportionality in our society is caused by some kind of racial discrimination, there can be no distinction between remedying societal discrimination and erasing racial disproportionalities in the country’s leadership caste. And if the lack of proportional racial representation among our leaders is not caused by societal discrimination, then “fixing” it is even less of a pressing public necessity.
The Court’s civics lesson presents yet another example of judicial selection of a theory of political representation based on skin color — an endeavor I have previously rejected. See Holder v. Hall, 512 U. S. 874, 899 (1994) (Thomas, J., concurring in judgment). The majority appears to believe that broader utopian goals justify the Law School’s use of race, but “[t]he Equal Protection Clause commands the elimination of racial barriers, not their creation in order to satisfy our theory as to how society ought to be organized.” DeFunis, 416 U. S., at 342 (Douglas, J., dissenting).
VII
As the foregoing makes clear, I believe the Court’s opinion to be, in most respects, erroneous. I do, however, find two points on which I agree.
A
First, I note that the issue of unconstitutional racial discrimination among the groups the Law School prefers is not presented in this case, because petitioner has never argued that the Law School engages in such a practice, and the Law School maintains that it does not. See Brief for Respondent Bollinger et al. 32, n. 50, and 6-7, n. 7. I join the Court’s opinion insofar as it confirms that this type of racial discrimination remains unlawful. Ante, at 326-327. Under today’s *375decision, it is still the case that racial discrimination that does not help a university to enroll an unspecified number, or “critical mass,” of underrepresented minority students is unconstitutional. Thus, the Law School may not discriminate in 'admissions between similarly situated blacks and Hispanics, or between whites and Asians. This is so because preferring black to Hispanic applicants, for instance, does nothing to further the interest recognized by the majority today.12 Indeed, the majority describes such racial balancing as “patently unconstitutional.” Ante, at 330. Like the Court, ante, at 336, I express no opinion as to whether the Law School’s current admissions program runs afoul of this prohibition.
B
The Court also holds that racial discrimination in admissions should be given another 25 years before it is deemed no longer narrowly tailored to the Law School’s fabricated compelling state interest. Ante, at 343. While I agree that in 25 years the practices of the Law School will be illegal, they are, for the reasons I have given, illegal now. The majority does not and cannot rest its time limitation on any evidence that the gap in credentials between black and white *376students is shrinking or will be gone in that timeframe.13 In recent years there has been virtually no change, for example, in the proportion of law school applicants with LSAT scores of 165 and higher who are black.14 In 1993 blacks constituted 1.1% of law school applicants in that score range, though they represented 11.1% of all applicants. Law School Admission Council, National Statistical Report (1994) (hereinafter LSAC Statistical Report). In 2000 the comparable numbers were 1.0% and 11.3%. LSAC Statistical Report (2001). No one can seriously contend, and the Court does not, that the racial gap in academic credentials will disappear in 25 years. Nor is the Court’s holding that racial discrimination will be unconstitutional in 25 years made contingent on the gap closing in that time.15
*377Indeed, the very existence of racial discrimination of the type practiced by the Law School may impede the narrowing of the LSAT testing gap. An applicant’s LSAT score can improve dramatically with preparation, but such preparation is a cost, and there must be sufficient benefits attached to an improved score to justify additional study. Whites scoring between 163 and 167 on the LSAT are routinely rejected by the Law School, and thus whites aspiring to admission at the Law School have every incentive to improve their score to levels above that range. See App. 199 (showing that in 2000, 209 out of 422 white applicants were rejected in this scoring range). Blacks, on the other hand, are nearly guaranteed admission if they score above 155. Id., at 198 (showing that 63 out of 77 black applicants are accepted with LSAT scores above 155). As admission prospects approach certainty, there is no incentive for the black applicant to continue to prepare for the LSAT once he is reasonably assured of achieving the requisite score. It is far from certain that the LSAT test-taker’s behavior is responsive to the Law School’s admissions policies.16 Nevertheless, the possibility remains that this racial discrimination will help fulfill the bigot’s prophecy about black underperformance — just as it confirms the conspiracy theorist’s belief that “institutional racism” is at fault for every racial disparity in our society.
I therefore can understand the imposition of a 25-year time limit only as a holding that the deference the Court pays to the Law School’s educational judgments and refusal to change its admissions policies will itself expire. At that point these policies will clearly have failed to “ ‘eliminate] *378the [perceived] need for any racial or ethnic’ ” discrimination because the academic credentials gap will still be there. Ante, at 343 (quoting Nathanson & Bartnik, The Constitutionality of Preferential Treatment for Minority Applicants to Professional Schools, 58 Chicago Bar Rec. 282, 293 (May-June 1977)). The Court defines this time limit in terms of narrow tailoring, see ante, at 343, but I believe this arises from its refusal to define rigorously the broad state interest vindicated today. Cf. Part II, supra. With these observations, I join the last sentence of Part III of the opinion of the Court.
* * *
For the immediate future, however, the majority has placed its imprimatur on a practice that can only weaken the principle of equality embodied in the Declaration of Independence and the Equal Protection Clause. “Our Constitution is color-blind, and neither knows nor tolerates classes among citizens.” Plessy v. Ferguson, 163 U. S. 537, 559 (1896) (Harlan, J., dissenting). It has been nearly 140 years since Frederick Douglass asked the intellectual ancestors of the Law School to “[d]o nothing with us!” and the Nation adopted the Fourteenth Amendment. Now we must wait another 25 years to see this principle of equality vindicated. I therefore respectfully dissent from the remainder of the Court’s opinion and the judgment.
Throughout I will use the two phrases interchangeably.
The Court’s refusal to address Wygant’s rejection of a state interest virtually indistinguishable from that presented by the Law School is perplexing. If the Court defers to the Law School’s judgment that a racially mixed student body confers educational benefits to all, then why would the Wygant Court not defer to the school board’s judgment with respect to the benefits a racially mixed faculty confers?
“[D]iversity,” for all of its devotees, is more a fashionable catehphrase than it is a useful term, especially when something as serious as racial discrimination is at issue. Because the Equal Protection Clause renders the color of one’s skin constitutionally irrelevant to the Law School’s mission, I refer to the Law School’s interest as an “aesthetic.” That is, the Law School wants to have a certain appearance, from the shape of the desks and tables in its classrooms to the color of the students sitting at them.
I also use the term “aesthetic” because I believe it underlines the ineffectiveness of racially discriminatory admissions in actually helping those who are truly underprivileged. Cf. Orr v. Orr, 440 U. S. 268, 283 (1979) (noting that suspect classifications are especially impermissible when “the choice made by the State appears to redound ... to the benefit of those without need for special solicitude”). It must be remembered that the Law School’s racial discrimination does nothing for those too poor or uneducated to participate in elite higher education and therefore presents only an illusory solution to the challenges facing our Nation.
The Law School believes both that the educational benefits of a racially engineered student body are large and that adjusting its overall admissions standards to achieve the same racial mix would require it to sacrifice its elite status. If the Law School is correct that the educational benefits of “diversity” are so great, then achieving them by altering admissions standards should not compromise its elite status. The Law School’s reluctance to do this suggests that the educational benefits it alleges are not significant or do not exist at all.
Cf. U. S. News & World Report, America’s Best Graduate Schools 28 (2004 ed.) (placing these schools in the uppermost 15 in the Nation).
The Court refers to this component of the Law School’s compelling state interest variously as “academic quality,” avoiding “sacrifice [of] a vital component of its educational mission,” and “academic selectivity.” Ante, at 340.
For example, North Carolina A&T State University, which is currently 5.4% white, College Admissions Data Handbook 643, could seek to reduce the representation of whites in order to gain additional educational benefits.
Were this Court to have the courage to forbid the use of racial discrimination in admissions, legacy preferences (and similar practices) might quickly become less popular — a possibility not lost, I am certain, on the elites (both individual and institutional) supporting the Law School in this case.
Cal. Const., Art. 1, § 31(a), states in full:
“The state shall not discriminate against, or grant preferential treatment to, any individual or group on the basis of race, sex, color, ethnicity, or national origin in the operation of public employment, public education, or public contracting.” See Coalition for Economic Equity v. Wilson, 122 F. 3d 692 (CA9 1997).
Given the incredible deference the Law School receives from the Court, I think it appropriate to indulge in the presumption that Boalt Hall operates without violating California law.
For example, there is no recognition by the Law School in this ease that even with their racial discrimination in place, black men are “underrepresented” at the Law School. See ABA-LSAC Guide 426 (reporting that the Law School has 46 black women and 28 black men). Why does the Law School not also discriminate in favor of black men over black *373women, given this underrepresentation? The answer is, again, that all the Law School cares about is its own image among know-it-all elites, not solving real problems like the crisis of black male underperformance.
That interest depends on enrolling a “critical mass” of underrepresented minority students, as the majority repeatedly states. Ante, at 316, 318, 319, 330, 333, 335, 340; cf. ante, at 333 (referring to the unique experience of being a “racial minority,” as opposed to being black, or Native American); ante, at 335-336 (rejecting argument that the Law School maintains a disguised quota by referring to the total number of enrolled underrepresented minority students, not specific races). As it relates to the Law School’s racial discrimination, the Court clearly approves of only one use of race — the distinction between underrepresented minority applicants and those of all other races. A relative preference awarded to a black applicant over, for example, a similarly situated Native American applicant, does not lead to the enrollment of even one more underrepresented minority student, but only balances the races within the “critical mass.”
I agree with Justice Ginsburg that the Court’s holding that racial discrimination in admissions will be illegal in 25 years is not based upon a “forecast,” post, at 346 (concurring opinion). I do not agree With Justice Ginsburg’s characterization of the Court’s holding as an expression of “hope.” Ibid.
I use a score of 165 as the benchmark here because the Law School feels it is the relevant score range for applicant consideration (absent race discrimination). See Brief for Respondent Bollinger et al. 5; App. to Pet. for Cert. 309a (showing that the median LSAT score for all accepted applicants from 1995-1998 was 168); id., at 310a-311a (showing the median LSAT score for accepted applicants was 167 for the years 1999 and 2000); University of Michigan Law School Website, available at http://www. law.umich.edu/prospectivestudents/Admissions/index.htm (showing that the median LSAT score for accepted applicants in 2002 was 166).
The majority’s non sequitur observation that since 1978 the number of blacks that have scored in these upper ranges on the LSAT has grown, ante, at 343, says nothing about current trends. First, black participation in the LSAT until the early 1990’s lagged behind black representation in the general population. For instance, in 1984 only 7.3% of law school applicants were black, whereas in 2000 11.3% of law school applicants were black. See LSAC Statistical Reports (1984 and 2000). Today, however, unless blacks were to begin applying to law school in proportions greater than their representation in the general population, the growth in absolute numbers of high scoring blacks should be expected to plateau, and it has. *377In 1992, 63 black applicants to law school had LSAT scores above 165. In 2000, that number was 65. See LSAC Statistical Reports (1992 and 2000).
I use the LSAT as an example, but the same incentive structure is in place for any admissions criteria, including undergraduate grades, on which minorities are consistently admitted at thresholds significantly lower than whites.