with whom The Chief Justice, Justice Sc alia, and Justice Kennedy join, dissenting.
At the heart of the Constitution’s guarantee of equal protection lies the simple command that the Government must treat citizens “as individuals, not ‘as simply components of a racial, religious, sexual or national class.’ ” Arizona Governing Comm, for Tax Deferred Annuity and Deferred Compensation Plans v. Norris, 463 U. S. 1073, 1083 (1983). Social scientists may debate how peoples’ thoughts and behavior reflect their background, but the Constitution provides that the Government may not allocate benefits and burdens among individuals based on the assumption that race or ethnicity determines how they act or think. To uphold the challenged programs, the Court departs from these fundamental principles and from our traditional requirement that racial classifications are permissible only if necessary and narrowly tailored to achieve a compelling interest. This departure marks a renewed toleration of racial classifications and a repudiation of our recent affirmation that the Constitution’s equal protection guarantees extend equally to all citi*603zens. The Court’s application of a lessened equal protection standard to congressional actions finds no support in our cases or in the Constitution. I respectfully dissent.
I
As we recognized last Term, the Constitution requires that the Court apply a strict standard of scrutiny to evaluate racial classifications such as those contained in the challenged FCC distress sale and comparative licensing policies. See Richmond v. J. A. Croson Co., 488 U. S. 469 (1989); see also Bolling v. Sharpe, 347 U. S. 497 (1954). “Strict scrutiny” requires that, to be upheld, racial classifications must be determined to be necessary and narrowly tailored to achieve a compelling state interest. The Court abandons this traditional safeguard against discrimination for a lower standard of review, and in practice applies a standard like that applicable to routine legislation. Yet the Government’s different treatment of citizens according to race is no routine concern. This Court’s precedents in no way justify the Court’s marked departure from our traditional treatment of race classifications and its conclusion that different equal protection principles apply to these federal actions.
In both the challenged policies, the Federal Communications Commission (FCC) provides benefits to some members of our society and denies benefits to others based on race or ethnicity. Except in the narrowest of circumstances, the Constitution bars such racial classifications as a denial to particular individuals, of any race or ethnicity, of “the equal protection of the laws.” U. S. Const., Arndt. 14, § 1; cf. Croson, supra, at 493-494. The dangers of such classifications are clear. They endorse race-based reasoning and the conception of a Nation divided into racial blocs, thus contributing to an escalation of racial hostility and conflict. See Croson, supra, at 493-494; Korematsu v. United States, 323 U. S. 214, 240 (1944) (Murphy, J., dissenting) (upholding treatment of individual based on inference from race is “to destroy the *604dignity of the individual and to encourage and open the door to discriminatory actions against other minority groups in the passions of tomorrow”). Such policies may embody stereotypes that treat individuals as the product of their race, evaluating their thoughts and efforts — their very worth as citizens — according to a criterion barred to the Government by history and the Constitution. Accord, Mississippi University for Women v. Hogan, 458 U. S. 718, 725-726 (1982). Racial classifications, whether providing benefits to or burdening particular racial or ethnic groups, may stigmatize those groups singled out for different treatment and may create considerable tension with the Nation’s widely shared commitment to evaluating individuals upon their individual merit. Cf. Regents of University of California v. Bakke, 438 U. S. 265, 358-362 (1978) (opinion of Brennan, J.). “Because racial characteristics so seldom provide a relevant basis for disparate treatment, and because classifications based on race are potentially so harmful to the entire body politic, it is especially important that the reasons for any such classifications be clearly identified and unquestionably legitimate.” Fullilove v. Klutznick, 448 U. S. 448, 533-535 (1980) (Stevens, J., dissenting) (footnotes omitted).
The Constitution’s guarantee of equal protection binds the Federal Government as it does the States, and no lower level of scrutiny applies to the Federal Government’s use of race classifications. In Bolling v. Sharpe, supra, the companion case to Brown v. Board of Education, 347 U. S. 483 (1954), the Court held that equal protection principles embedded in the Fifth Amendment’s Due Process Clause prohibited the Federal Government from maintaining racially segregated schools in the District of Columbia: “[I]t would be unthinkable that the same Constitution would impose a lesser duty on the Federal Government.” Id., at 500. Consistent with this view, the Court has repeatedly indicated that “the reach of the equal protection guarantee of the Fifth Amendment is coextensive with that of the Fourteenth.” United States v. *605Paradise, 480 U. S. 149, 166, n. 16 (1987) (plurality opinion) (considering remedial race classification); id., at 196 (O’Con-nor, J., dissenting); see also, e. g., Buckley v. Valeo, 424 U. S. 1, 93 (1976); Weinberger v. Wiesenfeld, 420 U. S. 636, 638, n. 2 (1975).
Nor does the congressional role in prolonging the FCC’s policies justify any lower level of scrutiny. As with all instances of judicial review of federal legislation, the Court does not lightly set aside the considered judgment of a coordinate branch. Nonetheless, the respect due a coordinate branch yields neither less vigilance in defense of equal protection principles nor any corresponding diminution of the standard of review. In Weinberger v. Wiesenfeld, for example, the Court upheld a widower’s equal protection challenge to a provision of the Social Security Act, found the assertedly benign congressional purpose to be illegitimate, and noted that “[t]his Court’s approach to Fifth Amendment equal protection claims has always been precisely the same as to equal protection claims under the Fourteenth Amendment.” 420 U. S., at 638, n. 2. The Court has not varied its standard of review when entertaining other equal protection challenges to congressional measures. See, e. g., Heckler v. Mathews, 465 U. S. 728 (1984); Califano v. Webster, 430 U. S. 313 (1977) (per curiam); Califano v. Goldfarb, 430 U. S. 199, 210-211 (1977) (traditional equal protection standard applies despite deference to congressional benefit determinations) (opinion of Brennan, J.); Buckley v. Valeo, supra, at 93; Frontiero v. Richardson, 411 U. S. 677, 684-691 (1973) (opinion of Brennan, J.). And Bolling v. Sharpe, supra, itself involved extensive congressional regulation of the segregated District of Columbia public schools.
Congress has considerable latitude, presenting special concerns for judicial review, when it exercises its “unique remedial powers . . . under §5 of the Fourteenth Amendment,” see Croson, supra, at 488 (opinion of O’Connor, J.), but this case does not implicate those powers. Section 5 empowers *606Congress to act respecting the States, and of course this case concerns only the administration of federal programs by federal officials. Section 5 provides to Congress the “power to enforce, by appropriate legislation, the provisions of this article,” which in part provides that “[n]o State shall . . . deny to any person within its jurisdiction the equal protection of the laws.” U. S. Const., Arndt. 14, §1. Reflecting the Fourteenth Amendment’s “dramatic change in the balance between congressional and state power over matters of race,” Croson, 488 U. S., at 490 (opinion of O’Connor, J.), that section provides to Congress a particular, structural role in the oversight of certain of the States’ actions. See id., at 488-491, 504; Hogan, swpra, at 732 (§ 5 grants power to enforce Amendment “‘to secure . . . equal protection of the laws against State denial or invasion,’ ” quoting Ex parte Virginia, 100 U. S. 339, 346 (1880)); Fullilove, supra, at 476-478, 483-484.
The Court asserts that Fullilove supports its novel application of intermediate scrutiny to “benign” race conscious measures adopted by Congress. Ante, at 564. Three reasons defeat this claim. First, Fullilove concerned an exercise of Congress’ powers under § 5 of the Fourteenth Amendment. In Fullilove, the Court reviewed an Act of Congress that had required States to set aside a percentage of federal construction funds for certain minority-owned businesses to remedy past discrimination in the award of construction contracts. Although the various opinions in Fullilove referred to several sources of congressional authority, the opinions make clear that it was § 5 that led the Court to apply a different form of review to the challenged program. See, e. g., 448 U. S., at 483 (opinion of Burger, C. J., joined by White, J., and Powell, J.) (“[I]n no organ of government, state or federal, does there repose a more comprehensive remedial power than in the Congress, expressly charged by the Constitution with competence and authority to enforce equal protection guarantees”); id., at 508-510, 516 (Powell, J., concur*607ring). Last Term, Croson resolved any doubt that might remain regarding this point. In Croson, we invalidated a local set-aside for minority contractors. We distinguished Fullilove, in which we upheld a similar set-aside enacted by Congress, on the ground that in Fullilove “Congress was exercising its powers under § 5 of the Fourteenth Amendment.” Croson, 488 U. S., at 504 (opinion of the Court); id., at 490 (opinion of O’Connor, J., joined by Rehnquist, C. J., and White, J.). Croson indicated that the decision in Fullilove turned on “the unique remedial powers of Congress under §5,” id., at 488 (opinion of O’Connor, J.), and that the latitude afforded Congress in identifying and redressing past discrimination rested on § 5’s “specific constitutional mandate to enforce the dictates of the Fourteenth Amendment.” Id., at 490. Justice Kennedy’s concurrence in Croson likewise provides the majority with no support, for it questioned whether the Court should, as it had in Fullilove, afford any particular latitude even to measures undertaken pursuant to §5. See id., at 518.
Second, Fullilove applies at most only to congressional measures that seek to remedy identified past discrimination. The Court upheld the challenged measures in Fullilove only because Congress had identified discrimination that had particularly affected the construction industry and had carefully constructed corresponding remedial measures. See Fullilove, 448 U. S., at 456-467, 480-489 (opinion of Burger, C. J.); id., at 498-499 (Powell, J., concurring). Fullilove indicated that careful review was essential to ensure that Congress acted solely for remedial rather than other, illegitimate purposes. See id., at 486-487 (opinion of Burger, C. J.); id., at 498-499 (Powell, J., concurring). The FCC and Congress are clearly not acting for any remedial purpose, see infra, at 611-612, and the Court today expressly extends its standard to racial classifications that are not remedial in any sense. See ante, at 564-565. This case does not present “a considered decision of the Congress and the President,” Fullilove, *608supra, at 473; nor does it present a remedial effort or exercise of § 5 powers.
Finally, even if Fullilove applied outside a remedial exercise of Congress’ §5 power, it would not support today’s adoption of the intermediate standard of review proffered by Justice Marshall, but rejected, in Fullilove. Under his suggested standard, the Government’s use of racial classifications need only be “ ‘substantially related to achievement’ ” of important governmental interests. Ante, at 565. Although the Court correctly observes that a majority did not apply strict scrutiny, six Members of the Court rejected intermediate scrutiny in favor of some more stringent form of review. Three Members of the Court applied strict scrutiny. See 448 U. S., at 496 (Powell, J., concurring) (challenged statute “employs a racial classification that is constitutionally prohibited unless it is a necessary means of advancing a compelling governmental interest”); id., at 498 (“means selected must be narrowly drawn”); id., at 523 (Stewart, J., joined by Rehnquist, J., dissenting). Chief Justice Burger’s opinion, joined by Justice White and Justice Powell, declined to adopt a particular standard of review but indicated that the Court must conduct “a most searching examination,” id., at 491, and that courts must ensure that “any congressional program that employs racial or ethnic criteria to accomplish the objective of remedying the present effects of past discrimination is narrowly tailored to the achievement of that goal.” Id., at 480. Justice Stevens indicated that “[rjacial classifications are simply too pernicious to permit any but the most exact connection between justification and classification.” Id., at 537-538 (dissenting opinion). Even Justice Marshall’s opinion concurring in the judgment, joined by Justice Brennan and Justice Blackmun, undermines the Court’s course today: That opinion expressly drew its lower standard of review from the plurality opinion in Regents of University of California v. Bakke, 438 U. S. 265 (1978), a case that did not involve con*609gressional action, and stated that the appropriate standard of review for the congressional measure challenged in Fullilove “is the same as that under the Fourteenth Amendment.” 448 U. S., at 517-518, n. 2 (internal quotation omitted). And, of course, Fullilove preceded our determination in Croson that strict scrutiny applies to preferences that favor members of minority groups, including challenges considered under the Fourteenth Amendment.
The guarantee of equal protection extends to each citizen, regardless of race: The Federal Government, like the States, may not “deny to any person within its jurisdiction the equal protection of the laws.” As we observed only last Term in Croson, “[ajbsent searching judicial inquiry into the justification for such race-based measures, there is simply no way of determining what classifications are ‘benign’ or ‘remedial’ and what classifications are in fact motivated by illegitimate notions of racial inferiority or simple racial politics.” 488 U. S., at 493 (opinion of O’Connor, J.); see also id., at 500, 494 (“[TJhe standard of review under the Equal Protection Clause is not dependent on the race of those burdened or benefited by a particular classification”).
The Court’s reliance on “benign racial classifications,” ante, at 564, is particularly troubling. “‘Benign’ racial classification” is a contradiction in terms. Governmental distinctions among citizens based on race or ethnicity, even in the rare circumstances permitted by our cases, exact costs and carry with them substantial dangers. To the person denied an opportunity or right based on race, the classification is hardly benign. The right to equal protection of the laws is a personal right, see Shelley v. Kraemer, 334 U. S. 1, 22 (1948), securing to each individual an immunity from treatment predicated simply on membership in a particular racial or ethnic group. The Court’s emphasis on “benign racial classifications” suggests confidence in its ability to distinguish good from harmful governmental uses of racial criteria. History should teach greater humility. Untethered to nar*610rowly confined remedial notions, “benign” carries with it no independent meaning, but reflects only acceptance of the current generation’s conclusion that a politically acceptable burden, imposed on particular citizens on the basis of race, is reasonable. The Court provides no basis for determining when a racial classification fails to be “benevolent.” By expressly distinguishing “benign” from remedial race-conscious measures, the Court leaves the distinct possibility that any racial measure found to be substantially related to an important governmental objective is also, by definition, “benign.” See ante, at 564-565. Depending on the preference of the moment, those racial distinctions might be directed expressly or in practice at any racial or ethnic group. We are a Nation not of black and white alone, but one teeming with divergent communities knitted together by various traditions and carried forth, above all, by individuals. Upon that basis, we are governed by one Constitution, providing a single guarantee of equal protection, one that extends equally to all citizens.
This dispute regarding the appropriate standard of review may strike some as a lawyers’ quibble over words, but it is not. The standard of review establishes whether and when the Court and Constitution allow the Government to employ racial classifications. A lower standard signals that the Government may resort to racial distinctions more readily. The Court’s departure from our cases is disturbing enough, but more disturbing still is the renewed toleration of racial classifications that its new standard of review embodies.
h — I h-H
Our history reveals that the most blatant forms of discrimination have been visited upon some members of the racial and ethnic groups identified in the challenged programs. Many have lacked the opportunity to share in the Nation’s wealth and to participate in its commercial enterprises. It is undisputed that minority participation in the broadcasting industry falls markedly below the demographic representation *611of those groups, see, e. g., Congressional Research Service, Minority Broadcast Station Ownership and Broadcast Programming: Is There a Nexus? 42 (June 29, 1988) (minority owners possess an interest in 13.3 percent of stations and a controlling interest in 3.5 percent of stations), and this shortfall may be traced in part to the discrimination and the patterns of exclusion that have widely affected our society. As a Nation we aspire to create a society untouched by that history of exclusion, and to ensure that equality defines all citizens’ daily experience and opportunities as well as the protection afforded to them under law.
For these reasons, and despite the harms that may attend the Government’s use of racial classifications, we have repeatedly recognized that the Government possesses a compelling interest in remedying the effects of identified race discrimination. We subject even racial classifications claimed to be remedial to strict scrutiny, however, to ensure that the Government in fact employs any race-conscious measures to further this remedial interest and employs them only when, and no more broadly than, the interest demands. See, e. g., Croson, supra, at 493-495, 498-502; Wygant v. Jackson Bd. of Ed., 476 U. S. 267 (1986) (plurality opinion). The FCC or Congress may yet conclude after suitable examination that narrowly tailored race-conscious measures are required to remedy discrimination that may be identified in the allocation of broadcasting licenses. Such measures are clearly within the Government’s power.
Yet it is equally clear that the policies challenged in these cases were not designed as remedial measures and are in no sense narrowly tailored to remedy identified discrimination. The FCC appropriately concedes that its policies embodied no remedial purpose, Tr. of Oral Arg. 40-42, and has disclaimed the possibility that discrimination infected the allocation of licenses. The congressional action at most simply endorsed a policy designed to further the interest in achieving diverse programming. Even if the appropriations *612measure could transform the purpose of the challenged policies, its text reveals no remedial purpose, and the accompanying legislative material confirms that Congress acted upon the same diversity rationale that led the FCC to formulate the challenged policies. See S. Rep. No. 100-182, p. 76 (1987). The Court refers to the bare suggestion, contained in a Report addressing different legislation passed in 1982, that “past inequities” have led to “underrepresentation of minorities in the media of mass communications, as it has adversely affected their participation in other sectors of the economy as well.” H. R. Conf. Rep. No. 97-765, p. 43 (1982); ante, at 566. This statement indicates nothing whatever about the purpose of the relevant appropriations measures, identifies no discrimination in the broadcasting industry, and would not sufficiently identify discrimination even if Congress were acting pursuant to its § 5 powers. Cf. Fullilove, 448 U. S., at 456-467 (opinion of Burger, C. J.) (surveying identification of discrimination affecting contracting opportunities); id., at 502-506 (Powell, J., concurring). The Court evaluates the policies only as measures designed to increase programming diversity. Ante, at 566-568. I agree that the racial classifications cannot be upheld as remedial measures.
Ill
Under the appropriate standard, strict scrutiny, only a compelling interest may support the Government’s use of racial classifications. Modern equal protection doctrine has recognized only one such interest: remedying the effects of racial discrimination. The interest in increasing the diversity of broadcast viewpoints is clearly not a compelling interest. It is simply too amorphous, too insubstantial, and too unrelated to any legitimate basis for employing racial classifications. The Court does not claim otherwise. Rather, it employs its novel standard and claims that this asserted interest need only be, and is, “important.” This conclusion twice compounds the Court’s initial error of reducing its level *613of scrutiny of a racial classification. First, it too casually extends the justifications that might support racial classifications, beyond that of remedying past discrimination. We have recognized that racial classifications are so harmful that “[u]nless they are strictly reserved for remedial settings, they may in fact promote notions of racial inferiority and lead to a politics of racial hostility.” Croson, 488 U. S., at 493. As Chief Justice Burger warned in Fullilove: “The history of governmental tolerance of practices using racial or ethnic criteria for the purpose or with the effect of imposing an invidious discrimination must alert us to the deleterious effects of even benign racial or ethnic classifications when they stray from narrow remedial justifications.” 448 U. S., at 486-487. Second, it has initiated this departure by endorsing an insubstantial interest, one that is certainly insufficiently weighty to justify tolerance of the Government’s distinctions among citizens based on race and ethnicity. This endorsement trivializes the constitutional command to guard against such discrimination and has loosed a potentially far-reaching principle disturbingly at odds with our traditional equal protection doctrine.
An interest capable of justifying race-conscious measures must be sufficiently specific and verifiable, such that it supports only limited and carefully defined uses of racial classifications. In Croson, we held that an interest in remedying societal discrimination cannot be considered compelling. See 488 U. S., at 505 (because the city of Richmond had presented no evidence of identified discrimination, it had “failed to demonstrate a compelling interest in apportioning public contracting opportunities on the basis of race”). We determined that a “generalized assertion” of past discrimination “has no logical stopping point” and would support unconstrained uses of race classifications. See id., at 498 (internal quotation marks omitted). In Wygant, we rejected the asserted interest in “providing minority role models for [a public school system’s] minority students, as an attempt to allevi*614ate the effects of societal discrimination,” 476 U. S., at 274 (plurality opinion), because “[s]ocietal discrimination, without more, is too amorphous a basis for imposing a racially classified remedy” and would allow “remedies that are ageless in their reach into the past, and timeless in their ability to affect the future.” Id,., at 276. Both cases condemned those interests because they would allow distribution of goods essentially according to the demographic representation of particular racial and ethnic groups. See Croson, supra, at 498, 505-506, 507; Wygant, 476 U. S., at 276 (plurality opinion).
The asserted interest in these cases suffers from the same defects. The interest is certainly amorphous: The FCC and the majority of this Court understandably do not suggest how one would define or measure a particular viewpoint that might be associated with race, or even how one would assess the diversity of broadcast viewpoints. Like the vague assertion of societal discrimination, a claim of insufficiently diverse broadcasting viewpoints might be used to justify equally unconstrained racial preferences, linked to nothing other than proportional representation of various races. And the interest would support indefinite use of racial classifications, employed first to obtain the appropriate mixture of racial views and then to ensure that the broadcasting spectrum continues to reflect that mixture. We cannot deem to be constitutionally adequate an interest that would support measures that amount to the core constitutional violation of “outright racial balancing.” Croson, supra, at 507.
The asserted interest would justify discrimination against members of any group found to contribute to an insufficiently diverse broadcasting spectrum, including those groups currently favored. In Wygant, we rejected as insufficiently weighty the interest in achieving role models in public schools, in part because that rationale could as readily be used to limit the hiring of teachers who belonged to particular minority groups. See Wygant, supra, at 275-276 (plurality *615opinion). The FCC’s claimed interest could similarly justify limitations on minority members’ participation in broadcasting. It would be unwise to depend upon the Court’s restriction of its holding to “benign” measures to forestall this result. Divorced from any remedial purpose and otherwise undefined, “benign” means only what shifting fashions and changing politics deem acceptable. Members of any racial or ethnic group, whether now preferred under the FCC’s policies or not, may find themselves politically out of fashion and subject to disadvantageous but “benign” discrimination.
Under the majority’s holding, the FCC may also advance its asserted interest in viewpoint diversity by identifying what constitutes a “black viewpoint,” an “Asian viewpoint,” an “Arab viewpoint,” and so on; determining which viewpoints are underrepresented; and then using that determination to mandate particular programming or to deny licenses to those deemed by virtue of their race or ethnicity less likely to present the favored views. Indeed, the FCC has, if taken at its word, essentially pursued this course, albeit without making express its reasons for choosing to favor particular groups or for concluding that the broadcasting spectrum is insufficiently diverse. See Statement of Policy on Minority Ownership of Broadcasting Facilities, 68 F. C. C. 2d 979 (1978) (1978 Policy Statement).
We should not accept as adequate for equal protection purposes an interest unrelated to race, yet capable of supporting measures so difficult to distinguish from proscribed discrimination. The remedial interest may support race classifications because that interest is necessarily related to past racial discrimination; yet the interest in diversity of viewpoints provides no legitimate, much less important, reason to employ race classifications apart from generalizations im-permissibly equating race with thoughts and behavior. And it will prove impossible to distinguish naked preferences for members of particular races from preferences for members of particular races because they possess certain valued *616views: No matter what its purpose, the Government will be able to claim that it has favored certain persons for their ability, stemming from race, to contribute distinctive views or perspectives.
Even considered as other than a justification for using race classifications, the asserted interest in viewpoint diversity falls short of being weighty enough. The Court has recognized an interest in obtaining diverse broadcasting viewpoints as a legitimate basis for the FCC, acting pursuant to its “public interest” statutory mandate, to adopt limited measures to increase the number of competing licensees and to encourage licensees to present varied views on issues of public concern. See, e. g., FCC v. National Citizens Committee for Broadcasting, 436 U. S. 775 (1978); Red Lion Broadcasting Co. v. FCC, 395 U. S. 367 (1969); United States v. Storer Broadcasting Co., 351 U. S. 192 (1956); Associated Press v. United States, 326 U. S. 1 (1945); National Broadcasting Co. v. United States, 319 U. S. 190 (1943). We have also concluded that these measures do not run afoul of the First Amendment’s usual prohibition of Government regulation of the marketplace of ideas, in part because First Amendment concerns support limited but inevitable Government regulation of the peculiarly constrained broadcasting spectrum. See, e. g., Red Lion, supra, at 389-390. But the conclusion that measures adopted to further the interest in diversity of broadcasting viewpoints are neither beyond the FCC’s statutory authority nor contrary to the First Amendment hardly establishes the interest as important for equal protection purposes.
The FCC’s extension of the asserted interest in diversity of views in these cases presents, at the very least, an unsettled First Amendment issue. The FCC has concluded that the American broadcasting public receives the incorrect mix of ideas and claims to have adopted the challenged policies to supplement programming content with a particular set of views. Although we have approved limited measures de*617signed to increase information and views generally, the Court has never upheld a broadcasting measure designed to amplify a distinct set of views or the views of a particular class of speakers. Indeed, the Court has suggested that the First Amendment prohibits allocating licenses to further such ends. See National Broadcasting Co. v. United States, supra, at 226 (“But Congress did not authorize the Commission to choose among [license] applicants upon the basis of their political, economic or social views, or upon any other capricious basis. If it did, or if the Commission by these Regulations proposed a choice among applicants upon some such basis, the [First Amendment] issue before us would be wholly different”). Even if an interest is determined to be legitimate in one context, it does not suddenly become important enough to justify distinctions based on race.
i — H C
Our traditional equal protection doctrine requires, in addition to a compelling state interest, that the Government’s chosen means be necessary to accomplish, and narrowly tailored to further, the asserted interest. See Wygant, 476 U. S., at 274 (plurality opinion); Palmore v. Sidoti, 466 U. S. 429, 432-433 (1984). This element of strict scrutiny is designed to “ensur[e] that the means chosen ‘fit’ [the] compelling goal so closely that there is little or no possibility that the motive for the classification was illegitimate racial prejudice or stereotype.” Croson, 488 U. S., at 493 (opinion of O’Connor, J.). The chosen means, resting as they do on stereotyping and so indirectly furthering the asserted end, could not plausibly be deemed narrowly tailored. The Court instead finds the racial classifications to be “substantially related” to achieving the Government’s interest, ante, at 569, a far less rigorous fit requirement. The FCC’s policies fail even this requirement.
*618A
The FCC claims to advance its asserted interest in diverse viewpoints by singling out race and ethnicity as peculiarly linked to distinct views that require enhancement. The FCC’s choice to employ a racial criterion embodies the related notions that a particular and distinct viewpoint inheres in certain racial groups, and that a particular applicant, by virtue of race or ethnicity alone, is more valued than other applicants because “likely to provide [that] distinct perspective.” Brief for FCC in No. 89-453, p. 17; see 1978 Policy Statement, 68 F. C. C. 2d, at 981 (policies seek “representation of minority viewpoints in programming”); Brief for FCC in No. 89-700, p. 20 (current ownership structure creates programming deficient in “minorities[’]. . . tastes and viewpoints”). The policies directly equate race with belief and behavior, for they establish race as a necessary and sufficient condition of securing the preference. The FCC’s chosen means rest on the “premise that differences in race, or in the color of a person’s skin, reflect real differences that are relevant to a person’s right to share in the blessings of a free society. [T]hat premise is utterly irrational and repugnant to the principles of a free and democratic society.” Wygant, supra, at 316 (Stevens, J., dissenting) (internal quotation marks omitted; citation omitted). The policies impermissi-bly value individuals because they presume that persons think in a manner associated with their race. See Steele v. FCC, 248 U. S. App. D. C. 279, 285, 770 F. 2d 1192, 1198 (1985) (minority preference contrary to “one of our most cherished constitutional and societal principles . . . that an individual’s tastes, beliefs, and abilities should be assessed on their own merits rather than by categorizing that individual as a member of a racial group presumed to think and behave in a particular way”), vacated, No. 84-1176 (Oct. 31, 1985), remanded (CADC, Oct. 9, 1986).
The FCC assumes a particularly strong correlation of race and behavior. The FCC justifies its conclusion that insuffi*619ciently diverse viewpoints are broadcast by reference to the percentage of minority-owned stations. This assumption is correct only to the extent that minority-owned stations provide the desired additional views, and that stations owned by individuals not favored by the preferences cannot, or at least do not, broadcast underrepresented programming. Additionally, the FCC’s focus on ownership to improve programming assumes that preferences linked to race are so strong that they will dictate the owner’s behavior in operating the station, overcoming the owner’s personal inclinations and regard for the market. This strong link between race and behavior, especially when mediated by market forces, is the assumption that Justice Powell rejected in his discussion of health care service in Bakke. See 438 U. S., at 310-311. In that case, the state medical school argued that it could prefer members of minority groups because they were more likely to serve communities particularly needing medical care. Justice Powell rejected this rationale, concluding that the assumption was unsupported and that such individual choices could not be presumed from ethnicity or race. Ibid.
The majority addresses this point by arguing that the equation of race with distinct views and behavior is not “impermissible” in these particular cases. Ante, at 579. Apart from placing undue faith in the Government and courts’ ability to distinguish “good” from “bad” stereotypes, this reasoning repudiates essential equal protection principles that prohibit racial generalizations. The Court embraces the FCC’s reasoning that an applicant’s race will likely indicate that the applicant possesses a distinct perspective, but notes that the correlation of race to behavior is “not a rigid assumption about how minority owners will behave in every case.” Ibid. The corollary to this notion is plain: Individuals of unfavored racial and ethnic backgrounds are unlikely to possess the unique experiences and background that contribute to viewpoint diversity. Both the reasoning and its corollary reveal but disregard what is objectionable about a stereo*620type: The racial generalization inevitably does not apply to certain individuals, and those persons may legitimately claim that they have been judged according to their race rather than upon a relevant criterion. See Los Angeles Dept. of Water and Power v. Manhart, 435 U. S. 702, 708 (1978) (“Even a true generalization about the class is an insufficient reason for disqualifying an individual to whom the generalization does not apply”). Similarly disturbing is the majority’s reasoning that different treatment on the basis of race is permissible because efficacious “in the aggregate.” Ante, at 579. In Wiesenfeld, we rejected similar reasoning: “Obviously, the notion that men are more likely than women to be the primary supporters of their spouses and children is not entirely without empirical support. But such a gender-based generalization cannot suffice to justify the denigration of the efforts of women who do work and whose earnings contribute significantly to their families’ support. ” 420 U. S., at 645 (citation omitted). Similarly in these cases, even if the Court’s equation of race and programming viewpoint has some empirical basis, equal protection principles prohibit the Government from relying upon that basis to employ racial classifications. See Manhart, supra, at 709 (“Practices that classify employees in terms of religion, race, or sex tend to preserve traditional assumptions about groups rather than thoughtful scrutiny of individuals”). This reliance on the “aggregate” and on probabilities confirms that the Court has abandoned heightened scrutiny, which requires a direct rather than approximate fit of means to ends. We would not tolerate the Government’s claim that hiring persons of a particular race leads to better service “in the aggregate,” and we should not accept as legitimate the FCC’s claim in these cases that members of certain races will provide superior programming, even if “in the aggregate.” The Constitution’s text, our cases, and our Nation’s history foreclose such premises.
*621B
Moreover, the FCC’s selective focus on viewpoints associated with race illustrates a particular tailoring difficulty. The asserted interest is in advancing the Nation’s different “social, political, esthetic, moral, and other ideas and experiences,” Red Lion, 395 U. S., at 390, yet of all the varied traditions and ideas shared among our citizens, the FCC has sought to amplify only those particular views it identifies through the classifications most suspect under equal protection doctrine. Even if distinct views could be associated with particular ethnic and racial groups, focusing on this particular aspect of the Nation’s views calls into question the Government’s genuine commitment to its asserted interest. See Bakke, 438 U. S., at 314 (opinion of Powell, J.) (race-conscious measures might be employed to further diversity only if race were one of many aspects of background sought and considered relevant to achieving a diverse student body).
Our equal protection doctrine governing intermediate review indicates that the Government may not use race and ethnicity as “a ‘proxy for other, more germane bases of classification.’” Hogan, 458 U. S., at 726, quoting Craig v. Boren, 429 U. S. 190,198 (1976). The FCC has used race as a proxy for whatever views it believes to be underrepresented in the broadcasting spectrum. This reflexive or unthinking use of a suspect classification is the hallmark of an unconstitutional policy. See, e. g., Wengler v. Druggists Mutual Ins. Co., 446 U. S. 142, 151-152 (1980); Craig, supra, at 198-199; Wiesenfeld, supra, at 643-645. The ill fit of means to ends is manifest. The policy is over inclusive: Many members of a particular racial or ethnic group will have no interest in advancing the views the FCC believes to be underrepresented, or will find them utterly foreign. The policy is underinclu-sive: It awards no preference to disfavored individuals who may be particularly well versed in and committed to presenting those views. The FCC has failed to implement a case-by-case determination, and that failure is particularly unjus*622tified when individualized hearings already occur, as in the comparative licensing process. See Orr v. Orr, 440 U. S. 268, 281 (1979). Even in the remedial context, we have required that the Government adopt means to ensure that the award of a particular preference advances the asserted interest. In Fullilove, even reviewing an exercise of § 5 powers, the Court upheld the challenged set-aside only because it contained a waiver provision that ensured that the program served its remedial function in particular cases. See Fulli-love, 448 U. S., at 487-488 (opinion of Burger, C. J.); see also Croson, 488 U. S., at 488-489 (opinion of O’Connor, J.).
Moreover, the FCC’s programs cannot survive even intermediate scrutiny because race-neutral and untried means of directly accomplishing the governmental interest are readily available. The FCC could directly advance its interest by requiring licensees to provide programming that the FCC believes would add to diversity. The interest the FCC asserts is in programming diversity, yet in adopting the challenged policies, the FCC expressly disclaimed having attempted any direct efforts to achieve its asserted goal. See 1978 Policy Statement, 68 F. C. C. 2d, at 981; ante, at 584-685, n. 36. The Court suggests that administrative convenience excuses this failure, ibid., yet intermediate scrutiny bars the Government from relying upon that excuse to avoid measures that directly further the asserted interest. See, e. g., Orr v. Orr, swpra, at 281; Craig v. Boren, supra,.at 198. The FCC and the Court suggest that First Amendment interests in some manner should exempt the FCC from employing this direct, race-neutral means to achieve its asserted interest. They essentially argue that we may bend our equal protection principles to avoid more readily apparent harm to our First Amendment values. But the FCC cannot have it both ways: Either the First Amendment bars the FCC from seeking to accomplish indirectly what it may not accomplish directly; or the FCC may pursue the goal, but must do so in a manner that comports with equal protection principles. And if the *623FCC can direct programming in any fashion, it must employ that direct means before resorting to indirect race-conscious means.
Other race-neutral means also exist, and all are at least as direct as the FCC’s racial classifications. The FCC could evaluate applicants upon their ability to provide, and commitment to offer, whatever programming the FCC believes would reflect underrepresented viewpoints. If the FCC truly seeks diverse programming rather than allocation of goods to persons of particular racial backgrounds, it has little excuse to look to racial background rather than programming to further the programming interest. Additionally, if the FCC believes that certain persons by virtue of their unique experiences will contribute as owners to more diverse broadcasting, the FCC could simply favor applicants whose particular background indicates that they will add to the diversity of programming, rather than rely solely upon suspect classifications. Also, race-neutral means exist to allow access to the broadcasting industry for those persons excluded for financial and related reasons. The Court reasons that various minority preferences, including those reflected in the distress sale, overcome barriers of information, experience, and financing that inhibit minority ownership. Ante, at 593-594. Race-neutral financial and informational measures most directly reduce financial and informational barriers.
The FCC could develop an effective ascertainment policy, one guaranteeing programming that reflects underrepresented viewpoints. The Court’s discussion of alternatives nearly exclusively focuses on the FCC’s ascertainment policy. Ante, at 585-589. Yet that policy applied only to existing licensees, addressed not viewpoints but issues of concern to often relatively homogeneous local communities, and, by the FCC’s own admission, was toothless and ineffective. According to the FCC, the ascertainment policies altered programming little more than the market already did, and provided “no guarantee that once a concern is ascertained by *624formal or informal means, programming responsive to that concern will be presented.” Commercial TV Stations, 98 F. C. C. 2d 1076, 1098 (1984), reconsideration denied, 104 F. C. C. 2d 358 (1986), remanded on other grounds sub nom. Action for Children’s Television v. FCC, 261 U. S. App. D. C. 253, 821 F. 2d 741 (1987); see also 98 F. C. C. 2d, at 1098-1101. Unsurprisingly, the FCC has concluded that this limited ascertainment policy has not proved to be effective, and has eliminated it throughout most media. See id., at 1097-1101; id., at 1099, and nn. 78-80 (surveying proceedings abandoning ascertainment requirements).
The FCC has posited a relative absence of “minority viewpoints,” yet it has never suggested what those views might be or what other viewpoints might be absent from the broadcasting spectrum. It has never identified any particular deficiency in programming diversity that should be the subject of greater programming or that necessitates racial classifications.
The FCC has never attempted to assess what alternatives to racial classifications might prove effective. The 1978 Policy Statement referred to only two alternatives that the Commission had undertaken: a minority hiring policy and the ascertainment policy. 68 F. C. C. 2d, at 979-980. Relying on ownership statistics and cursory evaluations of what viewpoints the broadcasting spectrum contained, the FCC asserted that insufficient programming diversity existed and that racial classifications were necessary. Id., at 980-981. Not until 1986 did the FCC attempt to determine the nature of the viewpoints that might be underrepresented or to determine whether effective race-neutral measures might achieve the FCC’s asserted interest. See, e. g., Notice of Inquiry on Racial, Ethnic, or Gender Classifications, 1 F. C. C. Red 1315 (1986), modified, 2 F. C. C. Red 2377 (1987). The FCC solicited comment about a range of potential race-neutral alternatives: It asked what race-neutral means might effectively increase program diversity, *625whether it should require an individualized showing of ability to contribute to program diversity, whether it should allow nonminority members to demonstrate their ability to contribute to diverse programming, and whether it should select applicants based on demonstrated commitment to particular issues rather than according to race. See 1 F. C. C. Red, at 1318. It was this inquiry, of course, that the congressional appropriations measures halted. See Continuing Appropriations Act for Fiscal Year 1988, Pub. L. 100-202, 101 Stat. 1329. Thus the record is clear: The FCC has never determined that it has any need to resort to racial classifications to achieve its asserted interest, and it has employed race-conscious means before adopting readily available race-neutral, alternative means.
The FCC seeks to avoid the tailoring difficulties by focusing on minority ownership rather than the asserted interest in diversity of broadcast viewpoints. The Constitution clearly prohibits allocating valuable goods such as broadcast licenses simply on the basis of race. See Bakke, 438 U. S., at 307 (opinion of Powell, J.). Yet the FCC refers to the lack of minority ownership of stations to support the existence of a lack of diversity of viewpoints, and has fitted its programs to increase ownership. See 1978 Policy Statement, supra; Commission Policy Regarding Advancement of Minority Ownership in Broadcasting, 92 F. C. C. 2d 849 (1982). This repeated focus on ownership supports the inference that the FCC seeks to allocate licenses based on race, an impermissible end, rather than to increase diversity of viewpoints, the asserted interest. And this justification that links the use of race preferences to minority ownership rather than to diversity of viewpoints ensures that the FCC’s programs, like that at issue in Croson, “cannot be said to be narrowly tailored to any goal, except perhaps outright racial balancing.” Croson, 488 U. S., at 507.
*626C
Even apart from these tailoring defects in the FCC’s policies, one particular flaw underscores the Government’s ill fit of means to ends. The FCC’s policies assume, and rely upon, the existence of a tightly bound “nexus” between the owners’ race and the resulting programming. The Court’s lengthy discussion of this issue, ante, at 569-579, purports to establish only that some relation exists between owners’ race and programming: i. e., that the FCC’s choice to focus on allocation of licenses is rationally related to the asserted end. The Court understandably makes no stronger claims, because the evidence provides no support and because the requisite deference would so obviously abandon heightened scrutiny. For argument’s sake, we can grant that the Court’s review of congressional hearings and social science studies establishes the existence of some rational nexus. But even assuming that to be true, the Court’s discussion does not begin to establish that the programs are directly and substantially related to the interest in diverse programming. That equal protection issue turns on the degree owners’ race is related to programming, rather than whether any relation exists. To the extent that the FCC cannot show the nexus to be nearly complete, that failure confirms that the chosen means do not directly advance the asserted interest, that the policies rest instead upon illegitimate stereotypes, and that individualized determinations must replace the FCC’s use of race as a proxy for the desired programming.
Three difficulties suggest that the nexus between owners’ race and programming is considerably less than substantial. First, the market shapes programming to a tremendous extent. Members of minority groups who own licenses might be thought, like other owners, to seek to broadcast programs that will attract and retain audiences, rather than programs that reflect the owner’s tastes and preferences. See Winter Park Communications, Inc. v. FCC, 277 U. S. App. D. C. 134, 145-148, 873 F. 2d 347, 358-361 (1989) (case below) (Wil*627liams, J., concurring in part and dissenting in part) (surveying evidence suggesting programming geared to audience taste). Second, station owners have only limited control over the content of programming. The distress sale presents a particularly acute difficulty of this sort. Unlike the comparative licensing program, the distress sale policy provides preferences to minority owners who neither intend nor desire to manage the station in any respect. See ante, at 557-558; Commission Policy Regarding Advancement of Minority Ownership in Broadcasting, supra. Whatever distinct programming may attend the race of an owner actively involved in managing the station, an absentee owner would have far less effect on programming.
Third, the FCC had absolutely no factual basis for the nexus when it adopted the policies and has since established none to support its existence. Until the mid-1970’s, the FCC believed that its public interest mandate and 1965 Policy Statement precluded it from awarding preference based on race and ethnicity, and instead required applicants to demonstrate particular entitlement to an advantage in a comparative hearing. Policy Statement on Comparative Broadcast Hearings, 1 F. C. C. 2d 393 (1965). See, e. g., Mid-Florida Television Corp., 33 F. C. C. 2d 1 (Rev. Bd.), review denied, 37 F. C. C. 2d 559 (1972), rev’d, TV 9, Inc. v. FCC, 161 U. S. App. D. C. 349, 495 F. 2d 929 (1973), cert. denied, 419 U. S. 986 (1974). The Court of Appeals for the District of Columbia Circuit rejected the FCC’s position on statutory grounds. See TV 9, 161 U. S. App. D. C., at 356-358, 495 F. 2d, at 936-938. The court rejected the FCC’s arguments that “the Communications Act, like the Constitution, is color-blind,” and that a race preference was incompatible with the FCC’s governing statute. Ibid. Instead, based on nothing other than its conception of the public interest, that court required that an applicant’s membership in a minority group be presumed to lead to greater diversity of programming. Id., at 357-358, 495 F. 2d, at *628937-938; see Garrett v. FCC, 168 U. S. App. D. C. 266, 272-273, 513 F. 2d 1056, 1062-1063 (1975). Principally relying on the panel’s presumed nexus between race and programming, the FCC in its 1978 Policy Statement acquiesced and established the policies challenged in these cases. See 1978 Policy Statement, supra, at 981-982. In the mid-1980’s, the FCC, prompted by this Court’s decisions indicating that a factual predicate must be established to support use of race classifications, unanimously sought to examine whether, and to what extent, any nexus existed between an owner’s race and programming. See Notice of Inquiry on Racial, Ethnic, or Gender Classifications, 1 F. C. C. Red 1315 (1986), modified, 2 F. C. C. Red 2377 (1987). As the Chairman of the FCC explained to Congress:
“To the extent that heightened scrutiny requires certain factual predicates, we discovered that notwithstanding our statements in the past regarding the assumed nexus between minority or female ownership and program diversity, a factual predicate has never been established.
“For example, the Commission has at no time examined whether there is a nexus between a broadcast owner’s race or gender and program diversity, either on a case-by-case basis or generically. We had no reason to, because the court in TV 9 told us we could, indeed must, assume such a nexus.” Minority-Owned Broadcast Stations, Hearing on H. R. 5373 before the Subcommittee on Telecommunications, Consumer Protection, and Finance of the House Committee on Energy and Commerce, 99th Cong., 2d Sess., 16 (1986).
Through the appropriations measures, Congress barred the FCC’s attempt to initiate that examination. See Continuing Appropriations Act for Fiscal Year 1988, 101 Stat. 1329-31.
Even apart from the limited nature of the Court’s claims, little can be discerned from the congressional action. First, the Court’s survey does not purport to establish that the *629FCC or Congress has identified any particular deficiency in the viewpoints contained in the broadcast spectrum. Second, no degree of congressional endorsement may transform the equation of race with behavior and thoughts into a permissible basis of governmental action. Even the most express and lavishly documented congressional declaration that members of certain races will as owners produce distinct and superior programming would not allow the Government to employ such reasoning to allocate benefits and burdens among citizens on that basis. Third, we should hesitate before accepting as definitive any declaration regarding even the existence of a nexus. The two legislative Reports that claim some nexus to exist refer to sources that provide no support for the proposition. See S. Rep. No. 100-182, p. 76 (1987); H. R. Conf. Rep. No. 97-765, p. 43 (1982). Congress, through appropriations measures, sought to foreclose examination of an issue that the FCC believed to be entirely unresolved. See Continuing Appropriations Act for Fiscal Year 1988, supra. Especially where Congress rejects the considered judgment of the executive officials possessing particular expertise regarding the matter in issue, courts are hardly bound to accept the congressional declaration. See, e. g., Rostker v. Goldberg, 453 U. S. 57, 83-85 (1981) (White, J., dissenting). Additionally, the FCC created the challenged policies. Congress has, through the appropriations process, frozen those policies in place by preventing the FCC from reexamining or altering them. That congressional action does not amount to an endorsement of the reasoning and empirical claims originally asserted and then abandoned by the FCC, and does not reflect the same considered judgment embodied in measures crafted through the legislative process and subject to the hearings and deliberation accompanying substantive legislation. Cf. TVA v. Hill, 437 U. S. 153 (1978); Andrus v. Sierra Club, 442 U. S. 347, 359-361 (1979).
*630D
Finally, the Government cannot employ race classifications that unduly burden individuals who are not members of the favored racial and ethnic groups. See, e. g., Wygant, 476 U. S., at 280-281 (plurality opinion). The challenged policies fail this independent requirement, as well as the other constitutional requirements. The comparative licensing and distress sale programs provide the eventual licensee with an exceptionally valuable property and with a rare and unique opportunity to serve the local community. The distress sale imposes a particularly significant burden. The FCC has at base created a specialized market reserved exclusively for minority controlled applicants. There is no more rigid quota than a 100% set-aside. This fact is not altered by the observation, see ante, at 598-599, that the FCC and the seller have some discretion over whether stations may be sold through the distress program. For the would-be purchaser or person who seeks to compete for the station, that opportunity depends entirely upon race or ethnicity. The Court’s argument that the distress sale allocates only a small percentage of all license sales, ante, at 599, also misses the mark. This argument readily supports complete preferences and avoids scrutiny of particular programs: It is no response to a person denied admission at one school, or discharged from one job, solely on the basis of race, that other schools or employers do not discriminate.
The comparative licensing program, too, imposes a significant burden. The Court’s emphasis on the multifactor process should not be confused with the claim that the preference is in some sense a minor one. It is not. The basic nonrace criteria are not difficult to meet, and, given the sums at stake, applicants have every incentive to structure their ownership arrangement to prevail in the comparative process. Applicants cannot alter their race, of course, and race is clearly the dispositive factor in a substantial percentage of comparative proceedings. Petitioner Metro asserts that *631race is overwhelmingly the dispositive factor. In reply, the FCC admits that it has not assessed the operation of its own program, Brief for FCC in No. 89-453, p. 39, and the Court notes only that “minority ownership does not guarantee that an applicant will prevail.” Ante, at 597-598, n. 50.
In sum, the FCC has not met its burden even under the Court’s test that approves of racial classifications that are substantially related to an important governmental objective. Of course, the programs even more clearly fail the strict scrutiny that should be applied. The Court has determined, in essence, that Congress and all federal agencies are exempted, to some ill-defined but significant degree, from the Constitution’s equal protection requirements. This break with our precedents greatly undermines equal protection guarantees and permits distinctions among citizens based on race and ethnicity which the Constitution clearly forbids. I respectfully dissent.