dissenting.
I dissent from denial of the petition for writ of certiorari. In this challenge to Denver’s use of racial preferences in public contracting, the Tenth Circuit, overturning the findings of the District Court, held that Denver had demonstrated a compelling interest in remedying racial discrimination in the Denver construction industry. The decision rests on an inference of racial discrimination from evidence that patently does not measure up to the standards set forth in Richmond v. J. A Croson Co., 488 U. S. 469 (1989). Coming on the heels of our decision last Term in Grutter v. Bollinger, 539 U. S. 306 (2003), the Court’s decision to let this plain disregard of Croson stand invites speculation that that case has effectively been overruled.
*1028I
Denver’s use of racial preferences began a generation ago, m 1977. In 1989, after this Court invalidated the city of Richmond’s 30% racial set-aside program for public contracting because the city had “failed to demonstrate a compelling interest in apportioning public contracting opportunities on the basis of race,” Croson, supra, at 505, Denver commissioned a study to assess the appropriateness of its program. The study was completed in June 1990, and a few months later Denver continued its system of racial preferences’ by passing City Ordinance No. 513, which established the scheme at issue in the present case.
Under the 1990 Ordinance, Denver’s City Council sets annual goals for the participation of minority business enterprises (MBEs) and woman-owned business enterprises in city contracting. The Mayor’s Office of Contract Compliance director then sets contract-specific goals for each covered contract. A prime contractor whose bid does not meet the goals will be disqualified, unless the prime contractor can show that it made a “good-faith effort” to do so — which requires satisfaction of 10 specific steps prescribed in the Ordinance.
To qualify as an MBE under the 1990 Ordinance, a firm must pass four threshold requirements, a 51% minority-ownership test, and a minority-control test. One of the threshold requirements is that the firm certify either (1) that it has been a.victim of past discrimination or (2) that it was in the city construction industry before June 1, 1990 (subsequently changed to Mar. 31, 1996)— in which latter case it is presumed to-have been a victim of discrimination.
II
The last-mentioned presumption, if it is to be a true indication of past vietimhood, must rest upon evidence not merely that there was some racial discrimination in the Denver construction industry prior to 1990 (or 1996), but that racial discrimination was so pervasive that it is reasonable to assume that it affected all minority-owned and controlled firms. Absent such evidence of pervasive discrimination, Denver’s seeming limitation of the set-asides to victims of racial discrimination is a sham, and the only function of the preferences is to channel a fixed percentage of city contracting dollars to firms identified by race.
*1029The Tenth Circuit found that Denver’s evidence established that “discrimination was persistent in the local construction industry and that Denver was, at least, an indirect participant in that discrimination.” 321 F. 3d 950, 990 (2003). This conclusion contrasted sharply with the District Court’s conclusion that “what can be said about the statistical studies presented in evidence in this case is that the methodology was not designed to answer the relevant questions, the collection of data was flawed, important variables were not accounted for in the analyses and the conclusions were based on unreasonable assumptions.” 86 F. Supp. 2d 1042, 1071 (Colo. 2000). The District Court and the Tenth Circuit derived such divergent conclusions from the same evidence because they analyzed that evidence with differing levels of scrutiny and skepticism about the city’s justifications. The first stage of the Tenth Circuit’s analysis was a determination that “the district court’s framework imposed a greater burden on Denver than that required by applicable law.” 321 F. 3d, at 974. In the second stage, the court filled the resulting gap by reweighing the evidence on its own, leading it to find “that Denver has demonstrated that a strong basis in evidence supported its conclusion that remedial action was necessary to remedy racial discrimination in the Denver construction industry.” Id., at 992. The Tenth Circuit departed from the principles enunciated by this Court at both stages.
A
Our Equal Protection Clause jurisprudence establishes that “whenever the government treats any person unequally because of his or her race, that person has suffered an injury that falls squarely within the language and spirit of the Constitution’s guarantee of equal protection.” Adarand Constructors, Inc. v. Peña, 515 U. S. 200, 229-230 (1995). It follows that a proper plaintiff challenging governmental use of racial preferences can state a prima facie case simply by pointing to this practice and showing that he or she was treated “unequally because of his or her race,” Ibid. Thereupon, the burden of sustaining the constitutionality of the use of racial preferences passes to the government, which must establish that it is remedying “identified discrimination” and that it “had a ‘strong basis in evidence’ to conclude that remedial action was necessary.” Shaw v. Hunt, 517 U. S. 899, 909-910 (1996) (internal quotation marks omitted).
*1030The Tenth Circuit interpreted the “strong basis in evidence” requirement in a miserly manner and ignored Croson’s requirement that the government prove that it is remedying identified discrimination. The District Court “believed Denver was required to prove the existence of discrimination.” 321 F. 3d, at 970. According to the Tenth Circuit, however, the District Court should only have asked “whether Denver had demonstrated strong evidence from which an inference of past or present discrimination could be drawn.” Ibid, (emphasis added). Instead of asking this easier question, the Tenth Circuit explained, the District Court was apparently misled by the plaintiff’s, “erroneous and unsupported statement... that Denver had the ‘burden of establishing by a preponderance that not only were there inferences of discrimination, but in fact that the inferences were correct.’ Denver, however, bore no such burden.” Ibid.
The District Court was correct, and the Tenth Circuit mistaken. “While the States and their subdivisions may take remedial action when they possess evidence that their own spending practices are exacerbating a pattern of prior discrimination, they must identify that discrimination, public or private, with some specificity before they may use race-conscious relief.” Croson, 488 U.S., at 504 (emphasis added). Quite obviously, “discrimination . . . identified] with some specificity” is discrimination that has been shown to have existed. It is inconsistent with Croson to permit racial preferences as a remedy for mere “might-have-been” racial discrimination, established by nothing more than evidence “from which an inference of past or present discrimination could be drawn.”
While holding Denver only to a watered-down “you don’t need to prove discrimination” standard, the Tenth Circuit simultaneously heightened the showing that the plaintiff contractor was required to make. According to the panel, “[o]nce Denver meets its burden, [the plaintiff] must introduce credible, particularized evidence to rebut [Denver’s] initial showing of the existence of a compelling interest.” 321 F. 3d, at 959 (internal quotation marks omitted). This is quite a daunting task, given how little Denver was required to show. Since Denver had to establish nothing more than the possibility of prior discrimination (evidence “from which an inference of past or present discrimination could be drawn”), the injured contractor was required to rebut *1031the possibility of discrimination in the Denver construction industry.
With regard to the burden of proof, then, the Tenth Circuit got it exactly backwards. It is not enough for a discriminating governmental entity to identify statistical disparities, assume (because it is a possible inference from the evidence) that these disparities were in fact caused by racial discrimination, implement racial preferences in public contracting on the basis of that assumption, and then require an injured contractor to demonstrate that the assumption of discrimination was incorrect. Rather, when the injured contractor has established the government’s use of racial preferences (a point conceded here), these preferences are presumed unconstitutional, and it then becomes the government’s burden to prove that it is acting on the basis of a compelling interest in remedying racial discrimination. To be crystal clear: Denver cannot meet its burden without proving that there was pervasive racial discrimination in the Denver construction industry.
B
The Tenth Circuit’s analysis also rests upon at least two serious errors that infect the statistical evidence submitted by the city. Croson requires the discriminating municipality to show a “significant statistical disparity” between the number of contractors hired and “the number of qualified minority contractors willing and able” to do the jobs. 488 U. S., at 509 (emphasis added). The statistical studies submitted by Denver did not meet this requirement because they did not measure the availability of minority firms — neither by use of actual bidding data nor by adjusting the raw data showing the total number of minority firms by means of some variable that would serve as a workable proxy for qualification, willingness, and ability. Instead, the city’s studies assumed that minority firms were on average as qualified, willing, and able as others.
In opposing certiorari, Denver argues that the use of actual contract bidding data should not be required (as the plaintiff’s expert suggested) because a study based on such data would be difficult to carry out, converting Croson scrutiny into scrutiny that is “‘strict in theory, but fatal in fact.’” Adarand Constructors, supra, at 237. This argument should be rejected. The scrutiny required by Croson should be fatal in fact when the statistical analysis put forward by the governmental entity de*1032fending intentional racial discrimination simply does not support its claim that the purpose and effect of its action was to compensate the victims of prior discrimination. If Denver found the use of contract bidding data too onerous, then it should have employed some other measure to make the statistical analysis valid — which obviously requires, as Croson said, that comparison be made, not with all minority firms, but with those that are qualified, willing, and able to undertake city contracts.
Secondly, even if it had been proper to assume that all minority firms were just as qualified, willing, and able to enter city contracts as other firms, it would still not have been proper to assume that all these minority firms had chances of obtaining city contracts equivalent to the chances of the nonminority firms that obtained them. Firms that are large and more experienced in performing big jobs will have more success in obtaining government contracts, and the uncontroverted evidence showed that MBEs were, on average, smaller and less experienced than their nonpreferred counterparts. In such circumstances, the government should have been required to produce a regression analysis controlling for these factors if it wished to rely on statistical disparities. See, e. g., Engineering Contractors Assn. of South Fla., Inc. v. Metropolitan Dade Cty., 122 F. 3d 895, 917 (CA11 1997) (after regression analysis to control for firm size was conducted, “most of the unfavorable disparities became statistically insignificant”).
The Tenth Circuit accepted the city’s contention that there was no need to control for size and experience because those are not race-neutral variables. MBEs, the court said, “are generally smaller and less experienced because of industry discrimination.” 321 F. 3d, at 981. The argument fails because it rests on nothing but speculation. Little is known about the relationship between minority ownership and size-and-experience in the Denver construction industry; one of the defects of the city’s disparity studies is precisely that they did not address those variables.
Denver did introduce studies identifying racial disparities in business formation rates and in access to capital. But if disparities in those more general areas sufficed to render size and experience impermissible explanations of racial disparities in construction contracting, they would similarly invalidate those explanations (and permit racial preferences) in every field of enterprise. Croson spoke specifically to this point, pointing out that reliance *1033upon such general societal discrimination “ ‘has no logical stopping point.’ . . . ‘Relief’ for such . . . ill-defined wrong[s] could extend until the percentage of public contracts awarded to MBE’s . . . mirrored the percentage of minorities in the population as a whole.” 488 U. S., at 498 (quoting Wygant v. Jackson Bd. of Ed., 476 U. S. 267, 275 (1986)). Discrimination in access to capital can be remedied directly — for example, by “prohibiting] discrimination in the provision of credit ... by local suppliers and banks,” 488 U. S., at 510 — but does not give rise to a compelling state interest to discriminate by race in construction contracting. No more so than does the existence of other larger social forces, such as disparities in education, that may similarly have some indirect effect on construction contracting.
Ill
Apart from the questions this case raises about faithful application of Croson, the case is worthy of the Court’s review because it presents a clear Circuit split on the standard of appellate review for the “strong basis in evidence” requirement. The genesis of this requirement is Wygant, supra, at 277, an affirmative-action-in-employment case in which the Court stated that “the trial court must make a factual determination that the employer had a strong basis in evidence for its conclusion that remedial action was necessary.” Croson carried the “strong basis in evidence” standard over to the use of racial classifications in public contracting. See 488 U. S., at 510.
The Tenth Circuit and three other Courts of Appeals view the question whether a governmental unit has demonstrated a “strong basis in evidence” sufficient to support its use of racial classifications as a question of law to be reviewed de novo. See Rothe Development Corp. v. United States Dept. of Defense, 262 F. 3d 1306, 1322-1323 (CA Fed. 2001); Majeske v. Chicago, 218 F. 3d 816, 820 (CA7 2000); Contractors Assn. of Eastern Pa., Inc. v. Philadelphia, 91 F. 3d 586, 596 (CA3 1996); Concrete Works of Colo., Inc. v. City and County of Denver, 36 F. 3d 1513, 1522 (CA10 1994). The Eleventh Circuit, in contrast, draws on the “factual determination” language in Wygant to interpret the “strong basis in evidence” question as one of fact, whose disposition is to be reviewed for clear error. See Engineering Contractors Assn. of South Fla., Inc., supra, at 903.
*1034The Court should resolve this significant and unsettled question. Any doubts about the question’s practical importance dissolve when one considers the manner in which the Tenth Circuit’s application of de novo review in this case permitted it to rule as it did notwithstanding the factual determinations made by the District Court after trial.
* * *
One of the primary functions of the requirement that governmental entities identify discrimination with specificity before using racial preferences is to implement the demand of the Equal Protection Clause “that the deviation from the norm of equal treatment of all racial and ethnic groups [be] a temporary matter.” Croson, supra, at 510. Governmental use of racial preferences must be “limited in time” because “racial classifications, however compelling their goals, are potentially so dangerous that they may be employed no more broadly than the interest demands.” Grutter, 539 U. S., at 342. Yet Denver has been using racial preferences in public contracting for á generation, and there is no indication that this will be anything other than business as usual for the foreseeable future.
Perhaps more than for any other reason, denial of certiorari in this case is important because of what it signals about this Court’s ongoing commitment to exacting judicial review of race-conscious policies. If the evidence relied upon by governmental units to justify their use of racial classifications can be as inconclusive as Denver’s evidence in this case, our former insistence upon a “strong basis in evidence” has been abandoned, to be replaced by what amounts to an “apparent-good-faith” requirement — that is, in the words of the Tenth Circuit, the existence of “evidence from which an inference of past or present discrimination could be drawn.” 321 F. 3d, at 970 (emphasis added). Some language in our recent radal-preferences-in-law-school-admissions case suggests a new willingness to rely upon good faith. See, e. g., Grutter, supra, at 343 (“We take the Law School at its word that it would ‘like nothing better than to find a race-neutral admissions formula’ and will terminate its race-conscious admissions program as soon as practicable”). We should grant certiorari to make clear that we stand by Croson’s insistence that “[rjadal classifications are suspect,” that “simple legislative assurances of good *1035intention cannot suffice,” and that the courts will employ “searching judicial inquiry into the justification for such race-based measures ... to ‘smoke out’ illegitimate uses of race.” 488 U. S., at 500, 493.