(delivered the opinion of the Court, in which Roberts, C.J., and Scalia, Thomas, and Alito, JJ., joined)
In the fire department of New Haven, Connecticut – as in emergency-service agencies throughout the Nation – fire-fighters prize their promotion to and within the officer ranks. An agency’s officers command respect within the department and in the whole community; and, of course, added responsibilities command increased salary and benefits. Aware of the intense competition for promotions, New Haven, like many cities, relies on objective examinations to identify the best qualified candidates.
In 2003, 118 New Haven fire-fighters took examinations to qualify for promotion to the rank of lieutenant or captain. Promotion examinations in New Haven (or City) were infrequent, so the stakes were high. The results would determine which fire-fighters would be considered for promotions during the next two years, and the order in which they would be considered. Many fire-fighters studied for months, at considerable personal and financial cost.
When the examination results showed that white candidates had outperformed minority candidates, the mayor and other local politicians opened a public debate that turned rancorous. Some fire-fighters argued the tests should be discarded because the results showed the tests to be discriminatory. They threatened a discrimination lawsuit if the City made promotions based on the tests. Other fire-fighters said the exams were neutral and fair. And they, in turn, threatened a discrimination lawsuit if the City, relying on the statistical racial disparity, ignored the test results and denied promotions to the candidates who had performed well. In the end the City took the side of those who protested the test results. It threw out the examinations.
Certain white and Hispanic fire-fighters who likely would have been promoted based on their good test performance sued the City and some of its officials. Theirs is the suit now before us. The suit alleges that, by discarding the test results, the City and the named officials discriminated against the plaintiffs based on their race, in violation of both Title VII of the Civil Rights Act of 1964, 78 Stat. 253, as amended, 42 U. S. C. §2000e et seq., and the Equal Protection Clause of the Fourteenth Amendment. The City and the officials defended their actions, arguing that if they had certified the results, they could have faced liability under Title VII for adopting a practice that had a disparate impact on the minority fire-fighters. The District Court granted summary judgment for the defendants, and the Court of Appeals affirmed.
We conclude that race-based action like the City’s in this case is impermissible under Title VII unless the employer can demonstrate a strong basis in evidence that, had it not taken the action, it would have been liable under the disparate-impact statute. The respondents, we further determine, cannot meet that threshold standard. As a result, the City’s action in discarding the tests was a violation of Title VII. In light of our ruling under the statutes, we need not reach the question whether respondents’ actions may have violated the Equal Protection Clause.
This litigation comes to us after the parties’ cross-motions for summary judgment, so we set out the facts in some detail. As the District Court noted, although “the parties strenuously dispute the relevance and legal import of, and inferences to be drawn from, many aspects of this case, the underlying facts are largely undisputed.” 554 F. Supp. 2d 142, 145 (Conn. 2006).
When the City of New Haven undertook to fill vacant lieutenant and captain positions in its fire department (Department), the promotion and hiring process was governed by the city charter, in addition to federal and state law. The charter establishes a merit system. That system requires the City to fill vacancies in the classified civil-service ranks with the most qualified individuals, as determined by job-related examinations. After each examination, the New Haven Civil Service Board (CSB) certifies a ranked list of applicants who passed the test. Under the charter’s “rule of three,” the relevant hiring authority must fill each vacancy by choosing one candidate from the top three scorers on the list. Certified promotional lists remain valid for two years.
The City’s contract with the New Haven fire-fighters’ union specifies additional requirements for the promotion process. Under the contract, applicants for lieutenant and captain positions were to be screened using written and oral examinations, with the written exam accounting for 60 percent and the oral exam 40 percent of an applicant’s total score. To sit for the examinations, candidates for lieutenant needed 30 months’ experience in the Department, a high-school diploma, and certain vocational training courses. Candidates for captain needed one year’s service as a lieutenant in the Department, a high-school diploma, and certain vocational training courses.
After reviewing bids from various consultants, the City hired Industrial/Organizational Solutions, Inc. (IOS) to develop and administer the examinations, at a cost to the City of $100,000. IOS is an Illinois company that specializes in designing entry-level and promotional examinations for fire and police departments. In order to fit the examinations to the New Haven Department, IOS began the test-design process by performing job analyses to identify the tasks, knowledge, skills, and abilities that are essential for the lieutenant and captain positions. IOS representatives interviewed incumbent captains and lieutenants and their supervisors. They rode with and observed other on-duty officers. Using information from those interviews and ride-alongs, IOS wrote job-analysis questionnaires and administered them to most of the incumbent battalion chiefs, captains, and lieutenants in the Department. At every stage of the job analyses, IOS, by deliberate choice, oversampled minority fire-fighters to ensure that the results – which IOS would use to develop the examinations – would not unintentionally favour white candidates.
With the job-analysis information in hand, IOS developed the written examinations to measure the candidates’ job-related knowledge. For each test, IOS compiled a list of training manuals, Department procedures, and other materials to use as sources for the test questions. IOS presented the proposed sources to the New Haven fire chief and assistant fire chief for their approval. Then, using the approved sources, IOS drafted a multiple-choice test for each position. Each test had 100 questions, as required by CSB rules, and was written below a 10th-grade reading level. After IOS prepared the tests, the City opened a 3-month study period. It gave candidates a list that identified the source material for the questions, including the specific chapters from which the questions were taken.
IOS developed the oral examinations as well. These concentrated on job skills and abilities. Using the job-analysis information, IOS wrote hypothetical situations to test incident-command skills, fire fighting tactics, interpersonal skills, leadership, and management ability, among other things. Candidates would be presented with these hypotheticals and asked to respond before a panel of three assessors.
IOS assembled a pool of 30 assessors who were superior in rank to the positions being tested. At the City’s insistence (because of controversy surrounding previous examinations), all the assessors came from outside Connecticut. IOS submitted the assessors’ resumes to City officials for approval. They were battalion chiefs, assistant chiefs, and chiefs from departments of similar sizes to New Haven’s throughout the country. Sixty-six percent of the panellists were minorities, and each of the nine three-member assessment panels contained two minority members. IOS trained the panellists for several hours on the day before it administered the examinations, teaching them how to score the candidates’ responses consistently using checklists of desired criteria.
Candidates took the examinations in November and December 2003. Seventy-seven candidates completed the lieutenant examination – 43 whites, 19 blacks, and 15 Hispanics. Of those, 34 candidates passed – 25 whites, 6 blacks, and 3 Hispanics. 554 F. Supp. 2d, at 145. Eight lieutenant positions were vacant at the time of the examination. As the rule of three operated, this meant that the top 10 candidates were eligible for an immediate promotion to lieutenant. All 10 were white. Ibid. Subsequent vacancies would have allowed at least 3 black candidates to be considered for promotion to lieutenant.
Forty-one candidates completed the captain examination – 25 whites, 8 blacks, and 8 Hispanics. Of those, 22 candidates passed – 16 whites, 3 blacks, and 3 Hispanics. Ibid. Seven captain positions were vacant at the time of the examination. Under the rule of three, 9 candidates were eligible for an immediate promotion to captain – 7 whites and 2 Hispanics. Ibid.
The City’s contract with IOS contemplated that, after the examinations, IOS would prepare a technical report that described the examination processes and methodologies and analyzed the results. But in January 2004, rather than requesting the technical report, City officials, including the City’s counsel, Thomas Ude, convened a meeting with IOS Vice President Chad Legel. (Legel was the leader of the IOS team that developed and administered the tests.) Based on the test results, the City officials expressed concern that the tests had discriminated against minority candidates. Legel defended the examinations’ validity, stating that any numerical disparity between white and minority candidates was likely due to various external factors and was in line with results of the Department’s previous promotional examinations.
Several days after the meeting, Ude sent a letter to the CSB purporting to outline its duties with respect to the examination results. Ude stated that under federal law, “a statistical demonstration of disparate impact,” standing alone, “constitutes a sufficiently serious claim of racial discrimination to serve as a predicate for employer-initiated, voluntar[y] remedies – even .... race-conscious remedies.” App. to Pet. for Cert. in No. 07–1428, p. 443a; see also 554 F. Supp. 2d, at 145 (issue of disparate impact “appears to have been raised by .... Ude”).
The CSB first met to consider certifying the results on January 22, 2004. Tina Burgett, director of the City’s Department of Human Resources, opened the meeting by telling the CSB that “there is a significant disparate impact on these two exams.” App. to Pet. for Cert. in No. 07–1428, at 466a. She distributed lists showing the candidates’ races and scores (written, oral, and composite) but not their names. Ude also described the test results as reflecting “a very significant disparate impact,” id., at 477a, and he outlined possible grounds for the CSB’s refusing to certify the results.
Although they did not know whether they had passed or failed, some firefighter-candidates spoke at the first CSB meeting in favour of certifying the test results. Michael Blatchley stated that “[e]very one” of the questions on the written examination “came from the [study] material. .... [I]f you read the materials and you studied the material, you would have done well on the test.” App. in No. 06–4996–cv (CA2), pp. A772–A773 (hereinafter CA2 App.). Frank Ricci stated that the test questions were based on the Department’s own rules and procedures and on “nationally recognized” materials that represented the “accepted standard[s]” for fire fighting. Id., at A785–A786. Ricci stated that he had “several learning disabilities,” including dyslexia; that he had spent more than $1,000 to purchase the materials and pay his neighbour to read them on tape so he could “give it [his] best shot”; and that he had studied “8 to 13 hours a day to prepare” for the test. Id., at A786, A789. “I don’t even know if I made it,” Ricci told the CSB, “[b]ut the people who passed should be promoted. When your life’s on the line, second best may not be good enough.” Id., at A787–A788.
Other fire-fighters spoke against certifying the test results. They described the test questions as outdated or not relevant to fire fighting practices in New Haven. Gary Tinney stated that source materials “came out of New York .... Their makeup of their city and everything is totally different than ours.” Id., at A774–A775; see also id., at A779, A780–A781. And they criticized the test materials, a full set of which cost about $500, for being too expensive and too long.
At a second CSB meeting, on February 5, the president of the New Haven fire-fighters’ union asked the CSB to perform a validation study to determine whether the tests were job-related. Petitioners’ counsel in this action argued that the CSB should certify the results. A representative of the International Association of Black Professional Fire-fighters, Donald Day from neighbouring Bridgeport, Connecticut, “beseech[ed]” the CSB “to throw away that test,” which he described as “inherently unfair” because of the racial distribution of the results. Id., at A830–A831. Another Bridgeport-based representative of the association, Ronald Mackey, stated that a validation study was necessary. He suggested that the City could “adjust” the test results to “meet the criteria of having a certain amount of minorities get elevated to the rank of Lieutenant and Captain.” Id., at A838. At the end of this meeting, the CSB members agreed to ask IOS to send a representative to explain how it had developed and administered the examinations. They also discussed asking a panel of experts to review the examinations and advise the CSB whether to certify the results.
At a third meeting, on February 11, Legel addressed the CSB on behalf of IOS. Legel stated that IOS had previously prepared entry-level firefighter examinations for the City but not a promotional examination. He explained that IOS had developed examinations for departments in communities with demographics similar to New Haven’s, including Orange County, Florida; Lansing, Michigan; and San Jose, California.
Legel explained the exam-development process to the CSB. He began by describing the job analyses IOS performed of the captain and lieutenant positions – the interviews, ride-alongs, and questionnaires IOS designed to “generate a list of tasks, knowledge, skills and abilities that are considered essential to performance” of the jobs. Id., at A931–A932. He outlined how IOS prepared the written and oral examinations, based on the job-analysis results, to test most heavily those qualities that the results indicated were “critica[l]” or “essentia[l].” Id., at A931. And he noted that IOS took the material for each test question directly from the approved source materials. Legel told the CSB that third-party reviewers had scrutinized the examinations to ensure that the written test was drawn from the source material and that the oral test accurately tested real-world situations that captains and lieutenants would face. Legel confirmed that IOS had selected oral-examination panellists so that each three-member assessment panel included one white, one black, and one Hispanic member.
Near the end of his remarks, Legel “implor[ed] anyone that had .... concerns to review the content of the exam. In my professional opinion, it’s facially neutral. There’s nothing in those examinations .... that should cause somebody to think that one group would perform differently than another group.” Id., at A961.
At the next meeting, on March 11, the CSB heard from three witnesses it had selected to “tell us a little bit about their views of the testing, the process, [and] the methodology.” Id., at A1020. The first, Christopher Hornick, spoke to the CSB by telephone. Hornick is an industrial/organizational psychologist from Texas who operates a consulting business that “direct[ly]” competes with IOS. Id., at A1029. Hornick, who had not “stud[ied] the test at length or in detail” and had not “seen the job analysis data,” told the CSB that the scores indicated a “relatively high adverse impact.” Id., at A1028, A1030, A1043. He stated that “[n]ormally, whites outperform ethnic minorities on the majority of standardized testing procedures,” but that he was “a little surprised” by the disparity in the candidates’ scores – although “[s]ome of it is fairly typical of what we’ve seen in other areas of the countr[y] and other tests.” Id., at A1028–A1029. Hornick stated that the “adverse impact on the written exam was somewhat higher but generally in the range that we’ve seen professionally.” Id., at A1030–A1031.
When asked to explain the New Haven test results, Hornick opined in the telephone conversation that the collective-bargaining agreement’s requirement of using written and oral examinations with a 60/40 composite score might account for the statistical disparity. He also stated that “[b]y not having anyone from within the [D]epartment review” the tests before they were administered – a limitation the City had imposed to protect the security of the exam questions – “you inevitably get things in there” that are based on the source materials but are not relevant to New Haven. Id., at A1034–A1035. Hornick suggested that testing candidates at an “assessment centre” rather than using written and oral examinations “might serve [the City’s] needs better.” Id., at A1039–A1040. Hornick stated that assessment centres, where candidates face real-world situations and respond just as they would in the field, allow candidates “to demonstrate how they would address a particular problem as opposed to just verbally saying it or identifying the correct option on a written test.” Ibid.
Hornick made clear that he was “not suggesting that [IOS] somehow created a test that had adverse impacts that it should not have had.” Id., at A1038. He described the IOS examinations as “reasonably good test[s].” Id., at A1041. He stated that the CSB’s best option might be to “certify the list as it exists” and work to change the process for future tests, including by “[r]ewriting the Civil Service Rules.” Ibid. Hornick concluded his telephonic remarks by telling the CSB that “for the future,” his company “certainly would like to help you if we can.” Id., at A1046.
The second witness was Vincent Lewis, a fire program specialist for the Department of Homeland Security and a retired fire captain from Michigan. Lewis, who is black, had looked “extensively” at the lieutenant exam and “a little less extensively” at the captain exam. He stated that the candidates “should know that material.” Id., at A1048, A1052. In Lewis’s view, the “questions were relevant for both exams,” and the New Haven candidates had an advantage because the study materials identified the particular book chapters from which the questions were taken. In other departments, by contrast, “you had to know basically the .... entire book.” Id., at A1053. Lewis concluded that any disparate impact likely was due to a pattern that “usually whites outperform some of the minorities on testing,” or that “more whites .... take the exam.” Id., at A1054.
The final witness was Janet Helms, a professor at Boston College whose “primary area of expertise” is “not with fire-fighters per se” but in “race and culture as they influence performance on tests and other assessment procedures.” Id., at A1060. Helms expressly declined the CSB’s offer to review the examinations. At the outset, she noted that “regardless of what kind of written test we give in this country .... we can just about predict how many people will pass who are members of under-represented groups. And your data are not that inconsistent with what predictions would say were the case.” Id., at A1061. Helms nevertheless offered several “ideas about what might be possible factors” to explain statistical differences in the results. Id., at A1062. She concluded that because 67 percent of the respondents to the job-analysis questionnaires were white, the test questions might have favoured white candidates, because “most of the literature on fire-fighters shows that the different groups perform the job differently.” Id., at A1063. Helms closed by stating that no matter what test the City had administered, it would have revealed “a disparity between blacks and whites, Hispanics and whites,” particularly on a written test. Id., at A1072.
At the final CSB meeting, on March 18, Ude (the City’s counsel) argued against certifying the examination results. Discussing the City’s obligations under federal law, Ude advised the CSB that a finding of adverse impact “is the beginning, not the end, of a review of testing procedures” to determine whether they violated the disparate-impact provision of Title VII. Ude focused the CSB on determining “whether there are other ways to test for .... those positions that are equally valid with less adverse impact.” Id., at A1101. Ude described Hornick as having said that the written examination “had one of the most severe adverse impacts that he had seen” and that “there are much better alternatives to identifying [fire fighting] skills.” Ibid. Ude offered his “opinion that promotions .... as a result of these tests would not be consistent with federal law, would not be consistent with the purposes of our Civil Service Rules or our Charter[,] nor is it in the best interests of the fire-fighters .... who took the exams.” Id., at A1103–A1104. He stated that previous Department exams “have not had this kind of result,” and that previous results had not been “challenged as having adverse impact, whereas we are assured that these will be.” Id., at A1107, A1108.
CSB Chairman Segaloff asked Ude several questions about the Title VII disparate-impact standard, Id., at A1108–A1109.
Karen DuBois-Walton, the City’s chief administrative officer, spoke on behalf of Mayor John DeStefano and argued against certifying the results. DuBois-Walton stated that the results, when considered under the rule of three and applied to then-existing captain and lieutenant vacancies, created a situation in which black and Hispanic candidates were disproportionately excluded from opportunity. DuBois-Walton also relied on Hornick’s testimony, asserting that Hornick “made it extremely clear that .... there are more appropriate ways to assess one’s ability to serve” as a captain or lieutenant. Id., at A1120.
Burgett (the human resources director) asked the CSB to discard the examination results. She, too, relied on Hornick’s statement to show the existence of alternative testing methods, describing Hornick as having “started to point out that alternative testing does exist” and as having “begun to suggest that there are some different ways of doing written examinations.” Id., at A1125, A1128.
Other witnesses addressed the CSB. They included the president of the New Haven fire-fighters’ union, who supported certification. He reminded the CSB that Hornick “also concluded that the tests were reasonable and fair and under the current structure to certify them.” Id., at A1137. Firefighter Frank Ricci again argued for certification; he stated that although “assessment centres in some cases show less adverse impact,” id., at A1140, they were not available alternatives for the current round of promotions. It would take several years, Ricci explained, for the Department to develop an assessment-centre protocol and the accompanying training materials. Id., at A1141. Lieutenant Matthew Marcarelli, who had taken the captain’s exam, spoke in favour of certification.
At the close of witness testimony, the CSB voted on a motion to certify the examinations. With one member recused, the CSB deadlocked 2 to 2, resulting in a decision not to certify the results. Explaining his vote to certify the results, Chairman Segaloff stated that “nobody convinced me that we can feel comfortable that, in fact, there’s some likelihood that there’s going to be an exam designed that’s going to be less discriminatory.” Id., at A1159–A1160.
The CSB’s decision not to certify the examination results led to this lawsuit. The plaintiffs – who are the petitioners here – are 17 white fire-fighters and 1 Hispanic firefighter who passed the examinations but were denied a chance at promotions when the CSB refused to certify the test results. They include the named plaintiff, Frank Ricci, who addressed the CSB at multiple meetings.
Petitioners sued the City, Mayor DeStefano, DuBois-Walton, Ude, Burgett, and the two CSB members who voted against certification. Petitioners also named as a defendant Boise Kimber, a New Haven resident who voiced strong opposition to certifying the results. Those individuals are respondents in this Court. Petitioners filed suit under Rev. Stat. §§1979 and 1980, 42 U. S. C. §§1983 and 1985, alleging that respondents, by arguing or voting against certifying the results, violated and conspired to violate the Equal Protection Clause of the Fourteenth Amendment. Petitioners also filed timely charges of discrimination with the Equal Employment Opportunity Commission (EEOC); upon the EEOC’s issuing right-to-sue letters, petitioners amended their complaint to assert that the City violated the disparate-treatment prohibition contained in Title VII of the Civil Rights Act of 1964, as amended. See 42 U. S. C. §§2000e–2(a).
The parties filed cross-motions for summary judgment. Respondents asserted they had a good-faith belief that they would have violated the disparate-impact prohibition in Title VII, §2000e–2(k), had they certified the examination results. It follows, they maintained, that they cannot be held liable under Title VII’s disparate-treatment provision for attempting to comply with Title VII’s disparate-impact bar. Petitioners countered that respondents’ good-faith belief was not a valid defence to allegations of disparate treatment and unconstitutional discrimination.
The District Court granted summary judgment for respondents. 554 F. Supp. 2d 142. It described petitioners’ argument as “boil[ing] down to the assertion that if [respondents] cannot prove that the disparities on the Lieutenant and Captain exams were due to a particular flaw inherent in those exams, then they should have certified the results because there was no other alternative in place.” Id., at 156. The District Court concluded that, “[n]otwithstanding the shortcomings in the evidence on existing, effective alternatives, it is not the case that [respondents] must certify a test where they cannot pinpoint its deficiency explaining its disparate impact .... simply because they have not yet formulated a better selection method.” Ibid. It also ruled that respondents’ “motivation to avoid making promotions based on a test with a racially disparate impact .... does not, as a matter of law, constitute discriminatory intent” under Title VII. Id., at 160. The District Court rejected petitioners’ equal protection claim on the theory that respondents had not acted because of “discriminatory animus” toward petitioners. Id., at 162. It concluded that respondents’ actions were not “based on race” because “all applicants took the same test, and the result was the same for all because the test results were discarded and nobody was promoted.” Id., at 161.
After full briefing and argument by the parties, the Court of Appeals affirmed in a one-paragraph, unpublished summary order; it later withdrew that order, issuing in its place a nearly identical, one-paragraph per curiam opinion adopting the District Court’s reasoning. 530 F. 3d 87 (CA2 2008). Three days later, the Court of Appeals voted 7 to 6 to deny rehearing en banc, over written dissents by Chief Judge Jacobs and Judge Cabranes. 530 F. 3d 88.
This action presents two provisions of Title VII to be interpreted and reconciled, with few, if any, precedents in the courts of appeals discussing the issue. Depending on the resolution of the statutory claim, a fundamental constitutional question could also arise. We found it prudent and appropriate to grant certiorari. 555 U. S. ___ (2009). We now reverse.
Petitioners raise a statutory claim, under the disparate-treatment prohibition of Title VII, and a constitutional claim, under the Equal Protection Clause of the Fourteenth Amendment. A decision for petitioners on their statutory claim would provide the relief sought, so we consider it first. See Atkins v Parker, 472 U. S. 115, 123 (1985); Escambia County v McMillan, 466 U. S. 48, 51 (1984) (per curiam) (“[N]ormally the Court will not decide a constitutional question if there is some other ground upon which to dispose of the case”).
Title VII of the Civil Rights Act of 1964, 42 U. S. C. §2000e et seq., as amended, prohibits employment discrimination on the basis of race, colour, religion, sex, or national origin. Title VII prohibits both intentional discrimination (known as “disparate treatment”) as well as, in some cases, practices that are not intended to discriminate but in fact have a disproportionately adverse effect on minorities (known as “disparate impact”).
As enacted in 1964, Title VII’s principal non-discrimination provision held employers liable only for disparate treatment. That section retains its original wording today. It makes it unlawful for an employer “to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s race, colour, religion, sex, or national origin.” §2000e–2(a)(1); see also 78 Stat. 255. Disparate-treatment cases present “the most easily understood type of discrimination,” Teamsters v United States, 431 U. S. 324, 335, n. 15 (1977), and occur where an employer has “treated [a] particular person less favourably than others because of” a protected trait. Watson v Fort Worth Bank & Trust, 487 U. S. 977, 985–986 (1988). A disparate-treatment plaintiff must establish “that the defendant had a discriminatory intent or motive” for taking a job-related action. Id., at 986.
The Civil Rights Act of 1964 did not include an express prohibition on policies or practices that produce a disparate impact. But in Griggs v Duke Power Co., 401 U. S. 424 (1971), the Court interpreted the Act to prohibit, in some cases, employers’ facially neutral practices that, in fact, are “discriminatory in operation.” Id., at 431. The Griggs Court stated that the “touchstone” for disparate-impact liability is the lack of “business necessity”: “If an employment practice which operates to exclude [minorities] cannot be shown to be related to job performance, the practice is prohibited.” Ibid.; see also id., at 432 (employer’s burden to demonstrate that practice has “a manifest relationship to the employment in question”); Albemarle Paper Co. v Moody, 422 U. S. 405, 425 (1975). Under those precedents, if an employer met its burden by showing that its practice was job-related, the plaintiff was required to show a legitimate alternative that would have resulted in less discrimination. Ibid. (allowing complaining party to show “that other tests or selection devices, without a similarly undesirable racial effect, would also serve the employer’s legitimate interest”).
Twenty years after Griggs, the Civil Rights Act of 1991, 105 Stat. 1071, was enacted. The Act included a provision codifying the prohibition on disparate-impact discrimination. That provision is now in force along with the disparate-treatment section already noted. Under the disparate-impact statute, a plaintiff establishes a prima facie violation by showing that an employer uses “a particular employment practice that causes a disparate impact on the basis of race, colour, religion, sex, or national origin.” 42 U. S. C. §2000e–2(k)(1)(A)(i). An employer may defend against liability by demonstrating that the practice is “job related for the position in question and consistent with business necessity.” Ibid. Even if the employer meets that burden, however, a plaintiff may still succeed by showing that the employer refuses to adopt an available alternative employment practice that has less disparate impact and serves the employer’s legitimate needs. §§2000e–2(k)(1)(A)(ii) and (C).
Petitioners allege that when the CSB refused to certify the captain and lieutenant exam results based on the race of the successful candidates, it discriminated against them in violation of Title VII’s disparate-treatment provision. The City counters that its decision was permissible because the tests “appear[ed] to violate Title VII’s disparate-impact provisions.” Brief for Respondents 12.
Our analysis begins with this premise: The City’s actions would violate the disparate-treatment prohibition of Title VII absent some valid defence. All the evidence demonstrates that the City chose not to certify the examination results because of the statistical disparity based on race – i.e., how minority candidates had performed when compared to white candidates. As the District Court put it, the City rejected the test results because “too many whites and not enough minorities would be promoted were the lists to be certified.” 554 F. Supp. 2d, at 152; see also ibid. (respondents’ “own arguments .... show that the City’s reasons for advocating non-certification were related to the racial distribution of the results”). Without some other justification, this express, race-based decision-making violates Title VII’s command that employers cannot take adverse employment actions because of an individual’s race. See §2000e–2(a)(1).
The District Court did not adhere to this principle, however. It held that respondents’ “motivation to avoid making promotions based on a test with a racially disparate impact .... does not, as a matter of law, constitute discriminatory intent.” 554 F. Supp. 2d, at 160. And the Government makes a similar argument in this Court. It contends that the “structure of Title VII belies any claim that an employer’s intent to comply with Title VII’s disparate-impact provisions constitutes prohibited discrimination on the basis of race.” Brief for United States as Amicus Curiae 11. But both of those statements turn upon the City’s objective – avoiding disparate-impact liability – while ignoring the City’s conduct in the name of reaching that objective. Whatever the City’s ultimate aim – however well intentioned or benevolent it might have seemed – the City made its employment decision because of race. The City rejected the test results solely because the higher scoring candidates were white. The question is not whether that conduct was discriminatory but whether the City had a lawful justification for its race-based action.
We consider, therefore, whether the purpose to avoid disparate-impact liability excuses what otherwise would be prohibited disparate-treatment discrimination. Courts often confront cases in which statutes and principles point in different directions. Our task is to provide guidance to employers and courts for situations when these two prohibitions could be in conflict absent a rule to reconcile them. In providing this guidance our decision must be consistent with the important purpose of Title VII – that the workplace be an environment free of discrimination, where race is not a barrier to opportunity.
With these principles in mind, we turn to the parties’ proposed means of reconciling the statutory provisions. Petitioners take a strict approach, arguing that under Title VII, it cannot be permissible for an employer to take race-based adverse employment actions in order to avoid disparate-impact liability – even if the employer knows its practice violates the disparate-impact provision. See Brief for Petitioners 43. Petitioners would have us hold that, under Title VII, avoiding unintentional discrimination cannot justify intentional discrimination. That assertion, however, ignores the fact that, by codifying the disparate-impact provision in 1991, Congress has expressly prohibited both types of discrimination. We must interpret the statute to give effect to both provisions where possible. See, e.g., United States v Atlantic Research Corp., 551 U. S. 128, 137 (2007) (rejecting an interpretation that would render a statutory provision “a dead letter”). We cannot accept petitioners’ broad and inflexible formulation.
Petitioners next suggest that an employer in fact must be in violation of the disparate-impact provision before it can use compliance as a defence in a disparate-treatment suit. Again, this is overly simplistic and too restrictive of Title VII’s purpose. The rule petitioners offer would run counter to what we have recognized as Congress’s intent that “voluntary compliance” be “the preferred means of achieving the objectives of Title VII.” Fire-fighters v Cleveland, 478 U. S. 501, 515 (1986); see also Wygant v Jackson Bd. of Ed., 476 U. S. 267, 290 (1986) (O’Connor, J., concurring in part and concurring in judgment). Forbidding employers to act unless they know, with certainty, that a practice violates the disparate-impact provision would bring compliance efforts to a near standstill. Even in the limited situations when this restricted standard could be met, employers likely would hesitate before taking voluntary action for fear of later being proven wrong in the course of litigation and then held to account for disparate treatment.
At the opposite end of the spectrum, respondents and the Government assert that an employer’s good-faith belief that its actions are necessary to comply with Title VII’s disparate-impact provision should be enough to justify race-conscious conduct. But the original, foundational prohibition of Title VII bars employers from taking adverse action “because of .... race.” §2000e–2(a)(1). And when Congress codified the disparate-impact provision in 1991, it made no exception to disparate-treatment liability for actions taken in a good-faith effort to comply with the new, disparate-impact provision in subsection (k). Allowing employers to violate the disparate-treatment prohibition based on a mere good-faith fear of disparate-impact liability would encourage race-based action at the slightest hint of disparate impact. A minimal standard could cause employers to discard the results of lawful and beneficial promotional examinations even where there is little if any evidence of disparate-impact discrimination. That would amount to a de facto quota system, in which a “focus on statistics .... could put undue pressure on employers to adopt inappropriate prophylactic measures.” Watson, 487 U. S., at 992 (plurality opinion). Even worse, an employer could discard test results (or other employment practices) with the intent of obtaining the employer’s preferred racial balance. That operational principle could not be justified, for Title VII is express in disclaiming any interpretation of its requirements as calling for outright racial balancing. §2000e–2(j). The purpose of Title VII “is to promote hiring on the basis of job qualifications, rather than on the basis of race or colour.” Griggs, 401 U. S., at 434.
In searching for a standard that strikes a more appropriate balance, we note that this Court has considered cases similar to this one, albeit in the context of the Equal Protection Clause of the Fourteenth Amendment. The Court has held that certain government actions to remedy past racial discrimination – actions that are themselves based on race – are constitutional only where there is a “ ‘strong basis in evidence’ ” that the remedial actions were necessary. Richmond v J. A. Croson Co., 488 U. S. 469, 500 (1989) (quoting Wygant, supra, at 277 (plurality opinion)). This suit does not call on us to consider whether the statutory constraints under Title VII must be parallel in all respects to those under the Constitution. That does not mean the constitutional authorities are irrelevant, however. Our cases discussing constitutional principles can provide helpful guidance in this statutory context. See Watson, supra, at 993 (plurality opinion).
Writing for a plurality in Wygant and announcing the strong-basis-in-evidence standard, Justice Powell recognized the tension between eliminating segregation and discrimination on the one hand and doing away with all governmentally imposed discrimination based on race on the other. 476 U. S., at 277. The plurality stated that those “related constitutional duties are not always harmonious,” and that “reconciling them requires .... employers to act with extraordinary care.” Ibid. The plurality required a strong basis in evidence because “[e]videntiary support for the conclusion that remedial action is warranted becomes crucial when the remedial program is challenged in court by nonminority employees.” Ibid. The Court applied the same standard in Croson, observing that “an amorphous claim that there has been past discrimination .... cannot justify the use of an unyielding racial quota.” 488 U. S., at 499.
The same interests are at work in the interplay between the disparate-treatment and disparate-impact provisions of Title VII. Congress has imposed liability on employers for unintentional discrimination in order to rid the workplace of “practices that are fair in form, but discriminatory in operation.” Griggs, supra, at 431. But it has also prohibited employers from taking adverse employment actions “because of” race. §2000e–2(a)(1). Applying the strong-basis-in-evidence standard to Title VII gives effect to both the disparate-treatment and disparate-impact provisions, allowing violations of one in the name of compliance with the other only in certain, narrow circumstances. The standard leaves ample room for employers’ voluntary compliance efforts, which are essential to the statutory scheme and to Congress’s efforts to eradicate workplace discrimination. See Fire-fighters, supra, at 515. And the standard appropriately constrains employers’ discretion in making race-based decisions: It limits that discretion to cases in which there is a strong basis in evidence of disparate-impact liability, but it is not so restrictive that it allows employers to act only when there is a provable, actual violation.
Resolving the statutory conflict in this way allows the disparate-impact prohibition to work in a manner that is consistent with other provisions of Title VII, including the prohibition on adjusting employment-related test scores on the basis of race. See §2000e–2(l). Examinations like those administered by the City create legitimate expectations on the part of those who took the tests. As is the case with any promotion exam, some of the fire-fighters here invested substantial time, money, and personal commitment in preparing for the tests. Employment tests can be an important part of a neutral selection system that safeguards against the very racial animosities Title VII was intended to prevent. Here, however, the fire-fighters saw their efforts invalidated by the City in sole reliance upon race-based statistics.
If an employer cannot rescore a test based on the candidates’ race, §2000e–2(l), then it follows a fortiori that it may not take the greater step of discarding the test altogether to achieve a more desirable racial distribution of promotion-eligible candidates – absent a strong basis in evidence that the test was deficient and that discarding the results is necessary to avoid violating the disparate-impact provision. Restricting an employer’s ability to discard test results (and thereby discriminate against qualified candidates on the basis of their race) also is in keeping with Title VII’s express protection of bona fide promotional examinations. See §2000e–2(h) (“[N]or shall it be an unlawful employment practice for an employer to give and to act upon the results of any professionally developed ability test provided that such test, its administration or action upon the results is not designed, intended or used to discriminate because of race”); cf. AT&T Corp. v Hulteen, 556 U. S. 701 (2009) (slip op., at 8).
For the foregoing reasons, we adopt the strong-basis-in-evidence standard as a matter of statutory construction to resolve any conflict between the disparate-treatment and disparate-impact provisions of Title VII.
Our statutory holding does not address the constitutionality of the measures taken here in purported compliance with Title VII. We also do not hold that meeting the strong-basis-in-evidence standard would satisfy the Equal Protection Clause in a future case. As we explain below, because respondents have not met their burden under Title VII, we need not decide whether a legitimate fear of disparate impact is ever sufficient to justify discriminatory treatment under the Constitution.
Nor do we question an employer’s affirmative efforts to ensure that all groups have a fair opportunity to apply for promotions and to participate in the process by which promotions will be made. But once that process has been established and employers have made clear their selection criteria, they may not then invalidate the test results, thus upsetting an employee’s legitimate expectation not to be judged on the basis of race. Doing so, absent a strong basis in evidence of an impermissible disparate impact, amounts to the sort of racial preference that Congress has disclaimed, §2000e–2(j), and is antithetical to the notion of a workplace where individuals are guaranteed equal opportunity regardless of race.
Title VII does not prohibit an employer from considering, before administering a test or practice, how to design that test or practice in order to provide a fair opportunity for all individuals, regardless of their race. And when, during the test-design stage, an employer invites comments to ensure the test is fair, that process can provide a common ground for open discussions toward that end. We hold only that, under Title VII, before an employer can engage in intentional discrimination for the asserted purpose of avoiding or remedying an unintentional disparate impact, the employer must have a strong basis in evidence to believe it will be subject to disparate-impact liability if it fails to take the race-conscious, discriminatory action.
The City argues that, even under the strong-basis-in-evidence standard, its decision to discard the examination results was permissible under Title VII. That is incorrect. Even if respondents were motivated as a subjective matter by a desire to avoid committing disparate-impact discrimination, the record makes clear there is no support for the conclusion that respondents had an objective, strong basis in evidence to find the tests inadequate, with some consequent disparate-impact liability in violation of Title VII.
On this basis, we conclude that petitioners have met their obligation to demonstrate that there is “no genuine issue as to any material fact” and that they are “entitled to judgment as a matter of law.” Fed. Rule Civ. Proc. 56(c). On a motion for summary judgment, “facts must be viewed in the light most favourable to the non-moving party only if there is a ‘genuine’ dispute as to those facts.” Scott v Harris, 550 U. S. 372, 380 (2007). “Where the record taken as a whole could not lead a rational trier of fact to find for the non-moving party, there is no genuine issue for trial.” Matsushita Elec. Industrial Co. v Zenith Radio Corp., 475 U. S. 574, 587 (1986) (internal quotation marks omitted). In this Court, the City’s only defence is that it acted to comply with Title VII’s disparate-impact provision. To succeed on their motion, then, petitioners must demonstrate that there can be no genuine dispute that there was no strong basis in evidence for the City to conclude it would face disparate-impact liability if it certified the examination results. See Celotex Corp. v Catrett, 477 U. S. 317, 324 (1986) (where the non-moving party “will bear the burden of proof at trial on a dispositive issue,” the non-moving party bears the burden of production under Rule 56 to “designate specific facts showing that there is a genuine issue for trial” (internal quotation marks omitted)).
The racial adverse impact here was significant, and petitioners do not dispute that the City was faced with a prima facie case of disparate-impact liability. On the captain exam, the pass rate for white candidates was 64 percent but was 37.5 percent for both black and Hispanic candidates. On the lieutenant exam, the pass rate for white candidates was 58.1 percent; for black candidates, 31.6 percent; and for Hispanic candidates, 20 percent. The pass rates of minorities, which were approximately one-half the pass rates for white candidates, fall well below the 80-percent standard set by the EEOC to implement the disparate-impact provision of Title VII. See 29 CFR §1607.4(D) (2008) (selection rate that is less than 80 percent “of the rate for the group with the highest rate will generally be regarded by the Federal enforcement agencies as evidence of adverse impact”); Watson, 487 U. S., at 995–996, n. 3 (plurality opinion) (EEOC’s 80-percent standard is “a rule of thumb for the courts”). Based on how the passing candidates ranked and an application of the “rule of three,” certifying the examinations would have meant that the City could not have considered black candidates for any of the then-vacant lieutenant or captain positions.
Based on the degree of adverse impact reflected in the results, respondents were compelled to take a hard look at the examinations to determine whether certifying the results would have had an impermissible disparate impact. The problem for respondents is that a prima facie case of disparate-impact liability – essentially, a threshold showing of a significant statistical disparity, Connecticut v Teal, 457 U. S. 440, 446 (1982), and nothing more – is far from a strong basis in evidence that the City would have been liable under Title VII had it certified the results. That is because the City could be liable for disparate-impact discrimination only if the examinations were not job related and consistent with business necessity, or if there existed an equally valid, less-discriminatory alternative that served the City’s needs but that the City refused to adopt. §2000e–2(k)(1)(A), (C). We conclude there is no strong basis in evidence to establish that the test was deficient in either of these respects. We address each of the two points in turn, based on the record developed by the parties through discovery – a record that concentrates in substantial part on the statements various witnesses made to the CSB.
There is no genuine dispute that the examinations were job-related and consistent with business necessity. The City’s assertions to the contrary are “blatantly contradicted by the record.” Scott, supra, at 380. The CSB heard statements from Chad Legel (the IOS vice president) as well as city officials outlining the detailed steps IOS took to develop and administer the examinations. IOS devised the written examinations, which were the focus of the CSB’s inquiry, after painstaking analyses of the captain and lieutenant positions – analyses in which IOS made sure that minorities were overrepresented. And IOS drew the questions from source material approved by the Department. Of the outside witnesses who appeared before the CSB, only one, Vincent Lewis, had reviewed the examinations in any detail, and he was the only one with any fire fighting experience. Lewis stated that the “questions were relevant for both exams.” CA2 App. A1053. The only other witness who had seen any part of the examinations, Christopher Hornick (a competitor of IOS’s), criticized the fact that no one within the Department had reviewed the tests – a condition imposed by the City to protect the integrity of the exams in light of past alleged security breaches. But Hornick stated that the exams “appea[r] to be .... reasonably good” and recommended that the CSB certify the results. Id., at A1041.
Arguing that the examinations were not job-related, respondents note some candidates’ complaints that certain examination questions were contradictory or did not specifically apply to fire fighting practices in New Haven. But Legel told the CSB that IOS had addressed those concerns – that it entertained “a handful” of challenges to the validity of particular examination questions, that it “reviewed those challenges and provided feedback [to the City] as to what we thought the best course of action was,” and that he could remember at least one question IOS had thrown out (“offer[ing] credit to everybody for that particular question”). Id., at A955–A957. For his part, Hornick said he “suspect[ed] that some of the criticisms .... [leveled] by candidates” were not valid. Id., at A1035.
The City, moreover, turned a blind eye to evidence that supported the exams’ validity. Although the City’s contract with IOS contemplated that IOS would prepare a technical report consistent with EEOC guidelines for examination-validity studies, the City made no request for its report. After the January 2004 meeting between Legel and some of the city-official respondents, in which Legel defended the examinations, the City sought no further information from IOS, save its appearance at a CSB meeting to explain how it developed and administered the examinations. IOS stood ready to provide respondents with detailed information to establish the validity of the exams, but respondents did not accept that offer.
Respondents also lacked a strong basis in evidence of an equally valid, less-discriminatory testing alternative that the City, by certifying the examination results, would necessarily have refused to adopt. Respondents raise three arguments to the contrary, but each argument fails. First, respondents refer to testimony before the CSB that a different composite-score calculation – weighting the written and oral examination scores 30/70 – would have allowed the City to consider two black candidates for then-open lieutenant positions and one black candidate for then-open captain positions. (The City used a 60/40 weighting as required by its contract with the New Haven fire-fighters’ union.) But respondents have produced no evidence to show that the 60/40 weighting was indeed arbitrary. In fact, because that formula was the result of a union-negotiated collective-bargaining agreement, we presume the parties negotiated that weighting for a rational reason. Nor does the record contain any evidence that the 30/70 weighting would be an equally valid way to determine whether candidates possess the proper mix of job knowledge and situational skills to earn promotions. Changing the weighting formula, moreover, could well have violated Title VII’s prohibition of altering test scores on the basis of race. See §2000e–2(l). On this record, there is no basis to conclude that a 30/70 weighting was an equally valid alternative the City could have adopted.
Second, respondents argue that the City could have adopted a different interpretation of the “rule of three” that would have produced less discriminatory results. The rule, in the New Haven city charter, requires the City to promote only from “those applicants with the three highest scores” on a promotional examination. New Haven, Conn., Code of Ordinances, Tit. I, Art. XXX, §160 (1992). A state court has interpreted the charter to prohibit so-called “banding” – the City’s previous practice of rounding scores to the nearest whole number and considering all candidates with the same whole-number score as being of one rank. Banding allowed the City to consider three ranks of candidates (with the possibility of multiple candidates filling each rank) for purposes of the rule of three. See Kelly v New Haven, No. CV000444614, 2004 WL 114377, *3 (Conn. Super. Ct., Jan. 9, 2004). Respondents claim that employing banding here would have made four black and one Hispanic candidates eligible for then-open lieutenant and captain positions.
A state court’s prohibition of banding, as a matter of municipal law under the charter, may not eliminate banding as a valid alternative under Title VII. See 42 U. S. C. §2000e–7. We need not resolve that point, however. Here, banding was not a valid alternative for this reason: Had the City reviewed the exam results and then adopted banding to make the minority test scores appear higher, it would have violated Title VII’s prohibition of adjusting test results on the basis of race. §2000e–2(l); see also Chicago Fire-fighters Local 2 v Chicago, 249 F. 3d 649, 656 (CA7 2001) (Posner, J.) (“We have no doubt that if banding were adopted in order to make lower black scores seem higher, it would indeed be .... forbidden”). As a matter of law, banding was not an alternative available to the City when it was considering whether to certify the examination results.
Third, and finally, respondents refer to statements by Hornick in his telephone interview with the CSB regarding alternatives to the written examinations. Hornick stated his “belie[f]” that an “assessment centre process,” which would have evaluated candidates’ behaviour in typical job tasks, “would have demonstrated less adverse impact.” CA2 App. A1039. But Hornick’s brief mention of alternative testing methods, standing alone, does not raise a genuine issue of material fact that assessment centres were available to the City at the time of the examinations and that they would have produced less adverse impact. Other statements to the CSB indicated that the Department could not have used assessment centres for the 2003 examinations. Supra, at 14. And although respondents later argued to the CSB that Hornick had pushed the City to reject the test results, supra, at 15–17, the truth is that the essence of Hornick’s remarks supported its certifying the test results. See Scott, 550 U. S., at 380. Hornick stated that adverse impact in standardized testing “has been in existence since the beginning of testing,” CA2 App. A1037, and that the disparity in New Haven’s test results was “somewhat higher but generally in the range that we’ve seen professionally.” Id., at A1030–A1031. He told the CSB he was “not suggesting” that IOS “somehow created a test that had adverse impacts that it should not have had.” Id., at A1038. And he suggested that the CSB should “certify the list as it exists.” Id., at A1041.
Especially when it is noted that the strong-basis-in-evidence standard applies, respondents cannot create a genuine issue of fact based on a few stray (and contradictory) statements in the record. And there is no doubt respondents fall short of the mark by relying entirely on isolated statements by Hornick. Hornick had not “stud[ied] the test at length or in detail.” Id., at A1030. And as he told the CSB, he is a “direct competitor” of IOS’s. Id., at A1029. The remainder of his remarks showed that Hornick’s primary concern – somewhat to the frustration of CSB members – was marketing his services for the future, not commenting on the results of the tests the City had already administered. See, e.g., id., at A1026, A1027, A1032, A1036, A1040, A1041. Hornick’s hinting had its intended effect: The City has since hired him as a consultant. As for the other outside witnesses who spoke to the CSB, Vincent Lewis (the retired fire captain) thought the CSB should certify the test results. And Janet Helms (the Boston College professor) declined to review the examinations and told the CSB that, as a society, “we need to develop a new way of assessing people.” Id., at A1073. That task was beyond the reach of the CSB, which was concerned with the adequacy of the test results before it.
On the record before us, there is no genuine dispute that the City lacked a strong basis in evidence to believe it would face disparate-impact liability if it certified the examination results. In other words, there is no evidence – let alone the required strong basis in evidence – that the tests were flawed because they were not job-related or because other, equally valid and less discriminatory tests were available to the City. Fear of litigation alone cannot justify an employer’s reliance on race to the detriment of individuals who passed the examinations and qualified for promotions. The City’s discarding the test results was impermissible under Title VII, and summary judgment is appropriate for petitioners on their disparate-treatment claim.
* * *
The record in this litigation documents a process that, at the outset, had the potential to produce a testing procedure that was true to the promise of Title VII: No individual should face workplace discrimination based on race. Respondents thought about promotion qualifications and relevant experience in neutral ways. They were careful to ensure broad racial participation in the design of the test itself and its administration. As we have discussed at length, the process was open and fair.
The problem, of course, is that after the tests were completed, the raw racial results became the predominant rationale for the City’s refusal to certify the results. The injury arises in part from the high, and justified, expectations of the candidates who had participated in the testing process on the terms the City had established for the promotional process. Many of the candidates had studied for months, at considerable personal and financial expense, and thus the injury caused by the City’s reliance on raw racial statistics at the end of the process was all the more severe. Confronted with arguments both for and against certifying the test results – and threats of a lawsuit either way – the City was required to make a difficult inquiry. But its hearings produced no strong evidence of a disparate-impact violation, and the City was not entitled to disregard the tests based solely on the racial disparity in the results.
Our holding today clarifies how Title VII applies to resolve competing expectations under the disparate-treatment and disparate-impact provisions. If, after it certifies the test results, the City faces a disparate-impact suit, then in light of our holding today it should be clear that the City would avoid disparate-impact liability based on the strong basis in evidence that, had it not certified the results, it would have been subject to disparate-treatment liability.
Petitioners are entitled to summary judgment on their Title VII claim, and we therefore need not decide the underlying constitutional question. The judgment of the Court of Appeals is reversed, and the cases are remanded for further proceedings consistent with this opinion.
It is so ordered
I join the Court’s opinion in full, but write separately to observe that its resolution of this dispute merely postpones the evil day on which the Court will have to confront the question: Whether, or to what extent, are the disparate-impact provisions of Title VII of the Civil Rights Act of 1964 consistent with the Constitution’s guarantee of equal protection? The question is not an easy one. See generally Primus, Equal Protection and Disparate Impact: Round Three, 117 Harv. L. Rev. 493 (2003).
The difficulty is this: Whether or not Title VII’s disparate-treatment provisions forbid “remedial” race-based actions when a disparate-impact violation would not otherwise result – the question resolved by the Court today – it is clear that Title VII not only permits but affirmatively requires such actions when a disparate-impact violation would otherwise result. See ante, at 20–21. But if the Federal Government is prohibited from discriminating on the basis of race, Bolling v Sharpe, 347 U. S. 497, 500 (1954), then surely it is also prohibited from enacting laws mandating that third parties – e.g., employers, whether private, State, or municipal – discriminate on the basis of race. See Buchanan v Warley, 245 U. S. 60, 78–82 (1917). As the facts of these cases illustrate, Title VII’s disparate-impact provisions place a racial thumb on the scales, often requiring employers to evaluate the racial outcomes of their policies, and to make decisions based on (because of) those racial outcomes. That type of racial decision-making is, as the Court explains, discriminatory. See ante, at 19; Personnel Administrator of Mass. v Feeney, 442 U. S. 256, 279 (1979).
To be sure, the disparate-impact laws do not mandate imposition of quotas, but it is not clear why that should provide a safe harbour. Would a private employer not be guilty of unlawful discrimination if he refrained from establishing a racial hiring quota but intentionally designed his hiring practices to achieve the same end? Surely he would. Intentional discrimination is still occurring, just one step up the chain. Government compulsion of such design would therefore seemingly violate equal protection principles. Nor would it matter that Title VII requires consideration of race on a wholesale, rather than retail, level. “[T]he Government must treat citizens as individuals, not as simply components of a racial, religious, sexual or national class.” Miller v Johnson, 515 U. S. 900, 911 (1995) (internal quotation marks omitted). And of course the purportedly benign motive for the disparate-impact provisions cannot save the statute. See Adarand Constructors, Inc. v Peńa, 515 U. S. 200, 227 (1995).
It might be possible to defend the law by framing it as simply an evidentiary tool used to identify genuine, intentional discrimination – to “smoke out,” as it were, disparate treatment. See Primus, supra, at 498–499, 520–521. Disparate impact is sometimes (though not always, see Watson v Fort Worth Bank & Trust, 487 U. S. 977, 992 (1988) (plurality opinion)) a signal of something illicit, so a regulator might allow statistical disparities to play some role in the evidentiary process. Cf. McDonnell Douglas Corp. v Green, 411 U. S. 792, 802–803 (1973). But arguably the disparate-impact provisions sweep too broadly to be fairly characterized in such a fashion – since they fail to provide an affirmative defence for good-faith (i.e., non-racially motivated) conduct, or perhaps even for good faith plus hiring standards that are entirely reasonable. See post, at 15–16, and n. 1 (Ginsburg, J., dissenting) (describing the demanding nature of the “business necessity” defence). This is a question that this Court will have to consider in due course. It is one thing to free plaintiffs from proving an employer’s illicit intent, but quite another to preclude the employer from proving that its motives were pure and its actions reasonable.
The Court’s resolution of these cases makes it unnecessary to resolve these matters today. But the war between disparate impact and equal protection will be waged sooner or later, and it behoves us to begin thinking about how – and on what terms – to make peace between them.
(with whom Justice Scalia and Justice Thomas join, concurring)
I join the Court’s opinion in full. I write separately only because the dissent, while claiming that “[t]he Court’s recitation of the facts leaves out important parts of the story,” post, at 2 (opinion of Ginsburg, J.), provides an incomplete description of the events that led to New Haven’s decision to reject the results of its exam. The dissent’s omissions are important because, when all of the evidence in the record is taken into account, it is clear that, even if the legal analysis in Parts II and III–A of the dissent were accepted, affirmance of the decision below is untenable.
When an employer in a disparate-treatment case under Title VII of the Civil Rights Act of 1964 claims that an employment decision, such as the refusal to promote, was based on a legitimate reason, two questions – one objective and one subjective – must be decided. The first, objective question is whether the reason given by the employer is one that is legitimate under Title VII. See St. Mary’s Honour Centre v Hicks, 509 U. S. 502, 506–507 (1993). If the reason provided by the employer is not legitimate on its face, the employer is liable. Id., at 509. The second, subjective question concerns the employer’s intent. If an employer offers a facially legitimate reason for its decision but it turns out that this explanation was just a pretext for discrimination, the employer is again liable. See id., at 510–512.
The question on which the opinion of the Court and the dissenting opinion disagree concerns the objective component of the determination that must be made when an employer justifies an employment decision, like the one made in this litigation, on the ground that a contrary decision would have created a risk of disparate-impact liability. The Court holds – and I entirely agree – that concern about disparate-impact liability is a legitimate reason for a decision of the type involved here only if there was a “substantial basis in evidence to find the tests inadequate.” Ante, at 26. The Court ably demonstrates that in this litigation no reasonable jury could find that the city of New Haven (City) possessed such evidence and therefore summary judgment for petitioners is required. Because the Court correctly holds that respondents cannot satisfy this objective component, the Court has no need to discuss the question of the respondents’ actual intent. As the Court puts it, “[e]ven if respondents were motivated as a subjective matter by a desire to avoid committing disparate-impact discrimination, the record makes clear there is no support for the conclusion that respondents had an objective, substantial basis in evidence to find the tests inadequate.” Ibid.
The dissent advocates a different objective component of the governing standard. According to the dissent, the objective component should be whether the evidence provided “good cause” for the decision, post, at 19, and the dissent argues – incorrectly, in my view – that no reasonable juror could fail to find that such evidence was present here. But even if the dissent were correct on this point, I assume that the dissent would not countenance summary judgment for respondents if respondents’ professed concern about disparate-impact litigation was simply a pretext. Therefore, the decision below, which sustained the entry of summary judgment for respondents, cannot be affirmed unless no reasonable jury could find that the City’s asserted reason for scrapping its test – concern about disparate-impact liability – was a pretext and that the City’s real reason was illegitimate, namely, the desire to placate a politically important racial constituency.
As initially described by the dissent, see post, at 2–12, the process by which the City reached the decision not to accept the test results was open, honest, serious, and deliberative. But even the District Court admitted that “a jury could rationally infer that city officials worked behind the scenes to sabotage the promotional examinations because they knew that, were the exams certified, the Mayor would incur the wrath of [Rev. Boise] Kimber and other influential leaders of New Haven’s African-American community.” 554 F. Supp. 2d 142, 162 (Conn. 2006), summarily aff’d, 530 F. 3d 87 (CA2 2008) (per curiam).
This admission finds ample support in the record. Reverend Boise Kimber, to whom the District Court referred, is a politically powerful New Haven pastor and a self-professed “ ‘kingmaker.’ ” App. to Pet. for Cert. in No. 07–1428, p. 906a; see also id., at 909a. On one occasion, “[i]n front of TV cameras, he threatened a race riot during the murder trial of the black man arrested for killing white Yalie Christian Prince. He continues to call whites racist if they question his actions.” Id., at 931a.
Reverend Kimber’s personal ties with seven-term New Haven Mayor John DeStefano (Mayor) stretch back more than a decade. In 1996, for example, Mayor DeStefano testified for Rev. Kimber as a character witness when Rev. Kimber – then the manager of a funeral home – was prosecuted and convicted for stealing prepaid funeral expenses from an elderly woman and then lying about the matter under oath. See id., at 126a, 907a. “Reverend Kimber has played a leadership role in all of Mayor DeStefano’s political campaigns, [and] is considered a valuable political supporter and vote-getter.” Id., at 126a. According to the Mayor’s former campaign manager (who is currently his executive assistant), Rev. Kimber is an invaluable political asset because “[h]e’s very good at organizing people and putting together field operations, as a result of his ties to labour, his prominence in the religious community and his long-standing commitment to roots.” Id., at 908a (internal quotation marks and alteration omitted).
In 2002, the Mayor picked Rev. Kimber to serve as the Chairman of the New Haven Board of Fire Commissioners (BFC), “despite the fact that he had no experience in the profession, fire administration, [or] municipal management.” Id., at 127a; see also id., at 928a–929a. In that capacity, Rev. Kimber told fire-fighters that certain new recruits would not be hired because “ ‘they just have too many vowels in their name[s].’ ” Thanawala, New Haven Fire Panel Chairman Steps Down Over Racial Slur, Hartford Courant, June 13, 2002, p. B2. After protests about this comment, Rev. Kimber stepped down as chairman of the BFC, ibid.; see also App. to Pet. for Cert. in No. 07–1428, at 929a, but he remained on the BFC and retained “a direct line to the mayor,” id., at 816a.
Almost immediately after the test results were revealed in “early January” 2004, Rev. Kimber called the City’s Chief Administrative Officer, Karen Dubois-Walton, who “acts ‘on behalf of the Mayor.’ ” Id., at 221a, 812a. Dubois-Walton and Rev. Kimber met privately in her office because he wanted “to express his opinion” about the test results and “to have some influence” over the City’s response. Id., at 815a–816a. As discussed in further detail below, Rev. Kimber adamantly opposed certification of the test results – a fact that he or someone in the Mayor’s office eventually conveyed to the Mayor. Id., at 229a.
On January 12, 2004, Tina Burgett (the director of the City’s Department of Human Resources) sent an e-mail to Dubois-Walton to coordinate the City’s response to the test results. Burgett wanted to clarify that the City’s executive officials would meet “sans the Chief, and that once we had a better fix on the next steps we would meet with the Mayor (possibly) and then the two Chiefs.” Id., at 446a. The “two Chiefs” are Fire Chief William Grant (who is white) and Assistant Fire Chief Ronald Dumas (who is African-American). Both chiefs believed that the test results should be certified. Id., at 228a, 817a. Petitioners allege, and the record suggests, that the Mayor and his staff colluded “sans the Chief[s]” because “the defendants did not want Grant’s or Dumas’ views to be publicly known; accordingly both men were prevented by the Mayor and his staff from making any statements regarding the matter.” Id., at 228a.
The next day, on January 13, 2004, Chad Legel, who had designed the tests, flew from Chicago to New Haven to meet with Dubois-Walton, Burgett, and Thomas Ude, the City’s corporate counsel. Id., at 179a. “Legel outlined the merits of the examination and why city officials should be confident in the validity of the results.” Ibid. But according to Legel, Dubois-Walton was “argumentative” and apparently had already made up her mind that the tests were “ ‘discriminatory.’ ” Id., at 179a–180a. Again according to Legel, “[a] theme” of the meeting was “the political and racial overtones of what was going on in the City.” Id., at 181a. “Legel came away from the January 13, 2004 meeting with the impression that defendants were already leaning toward discarding the examination results.” Id., at 180a.
On January 22, 2004, the Civil Service Board (CSB or Board) convened its first public meeting. Almost immediately, Rev. Kimber began to exert political pressure on the CSB. He began a loud, minutes-long outburst that required the CSB Chairman to shout him down and hold him out of order three times. See id., at 187a, 467a–468a; see also App. in No. 06–4996–cv (CA2), pp. A703–A705. Reverend Kimber protested the public meeting, arguing that he and the other fire commissioners should first be allowed to meet with the CSB in private. App. to Pet. for Cert. in No. 07–1428, at 188a.
Four days after the CSB’s first meeting, Mayor DeStefano’s executive aide sent an e-mail to Dubois-Walton, Burgett, and Ude. Id., at 190a. The message clearly indicated that the Mayor had made up his mind to oppose certification of the test results (but nevertheless wanted to conceal that fact from the public), Ibid.:
I wanted to make sure we are all on the same page for this meeting tomorrow .... [L]et’s remember, that these folks are not against certification yet. So we can’t go in and tell them that is our position; we have to deliberate and arrive there as the fairest and most cogent outcome.
On February 5, 2004, the CSB convened its second public meeting. Reverend Kimber again testified and threatened the CSB with political recriminations if they voted to certify the test results, Id., at 492a:
I look at this [Board] tonight. I look at three whites and one Hispanic and no blacks .... I would hope that you would not put yourself in this type of position, a political ramification that may come back upon you as you sit on this [Board] and decide the future of a department and the future of those who are being promoted.
.... .... ....
One of the CSB members “t[ook] great offense” because he believed that Rev. Kimber “consider[ed] [him] a bigot because [his] face is white.” Id., at 496a. The offended CSB member eventually voted not to certify the test results. Id., at 586a–587a.
One of Rev. Kimber’s “friends and allies,” Lieutenant Gary Tinney, also exacerbated racial tensions before the CSB. Id., at 129a. After some fire-fighters applauded in support of certifying the test results, “Lt. Tinney exclaimed, ‘Listen to the Klansmen behind us.’ ” Id., at 225a.
Tinney also has strong ties to the Mayor’s office. See, e.g., id., at 129a–130a, 816a–817a. After learning that he had not scored well enough on the captain’s exam to earn a promotion, Tinney called Dubois-Walton and arranged a meeting in her office. Id., at 830a–831a, 836a. Tinney alleged that the white fire-fighters had cheated on their exams – an accusation that Dubois-Walton conveyed to the Board without first conducting an investigation into its veracity. Id., at 837a–838a; see also App. 164 (statement of CSB Chairman, noting the allegations of cheating). The allegation turned out to be baseless. App. to Pet. for Cert. in No. 07–1428, at 836a.
Dubois-Walton never retracted the cheating allegation, but she and other executive officials testified several times before the CSB. In accordance with directions from the Mayor’s office to make the CSB meetings appear deliberative, see id., at 190a, executive officials remained publicly uncommitted about certification – while simultaneously “work[ing] as a team” behind closed doors with the secretary of the CSB to devise a political message that would convince the CSB to vote against certification, see id., at 447a. At the public CSB meeting on March 11, 2004, for example, Corporation Counsel Ude bristled at one board member’s suggestion that City officials were recommending against certifying the test results. See id., at 215a (“Attorney Ude took offense, stating, ‘Frankly, because I would never make a recommendation – I would not have made a recommendation like that’”). But within days of making that public statement, Ude privately told other members of the Mayor’s team “the ONLY way we get to a decision not to certify is” to focus on something other than “a big discussion re: adverse impact” law. Id., at 458a–459a.
As part of its effort to deflect attention from the specifics of the test, the City relied heavily on the testimony of Dr. Christopher Hornick, who is one of Chad Legel’s competitors in the test-development business. Hornick never “stud[ied] the test [that Legel developed] at length or in detail,” id., at 549a; see also id., at 203a, 553a, but Hornick did review and rely upon literature sent to him by Burgett to criticize Legel’s test. For example, Hornick “noted in the literature that [Burgett] sent that the test was not customized to the New Haven Fire Department.” Id., at 551a. The Chairman of the CSB immediately corrected Hornick. Id., at 552a (“Actually, it was, Dr. Hornick”). Hornick also relied on newspaper accounts – again, sent to him by Burgett – pertaining to the controversy surrounding the certification decision. See id., at 204a, 557a. Although Hornick again admitted that he had no knowledge about the actual test that Legel had developed and that the City had administered, see id., at 560a–561a, the City repeatedly relied upon Hornick as a testing “guru” and, in the CSB Chairman’s words, “the City ke[pt] quoting him as a person that we should rely upon more than anybody else [to conclude that there] is a better way – a better mousetrap.” App. in No. 06–4996–cv (CA2), at A1128. Dubois-Walton later admitted that the City rewarded Hornick for his testimony by hiring him to develop and administer an alternative test. App. to Pet. for Cert. in No. 07–1428, at 854a; see also id., at 562a–563a (Hornick’s plea for future business from the City on the basis of his criticisms of Legel’s tests).
At some point prior to the CSB’s public meeting on March 18, 2004, the Mayor decided to use his executive authority to disregard the test results – even if the CSB ultimately voted to certify them. Id., at 819a–820a. Accordingly, on the evening of March 17th, Dubois-Walton sent an e-mail to the Mayor, the Mayor’s executive assistant, Burgett, and attorney Ude, attaching two alternative press releases. Id., at 457a. The first would be issued if the CSB voted not to certify the test results; the second would be issued (and would explain the Mayor’s invocation of his executive authority) if the CSB voted to certify the test results. Id., at 217a–218a, 590a–591a, 819a–820a. Half an hour after Dubois-Walton circulated the alternative drafts, Burgett replied: “[W]ell, that seems to say it all. Let’s hope draft #2 hits the shredder tomorrow nite.” Id., at 457a.
Soon after the CSB voted against certification, Mayor DeStefano appeared at a dinner event and “took credit for the scu[tt]ling of the examination results.” Id., at 230a.
Taking into account all the evidence in the summary judgment record, a reasonable jury could find the following. Almost as soon as the City disclosed the racial makeup of the list of fire-fighters who scored the highest on the exam, the City administration was lobbied by an influential community leader to scrap the test results, and the City administration decided on that course of action before making any real assessment of the possibility of a disparate-impact violation. To achieve that end, the City administration concealed its internal decision but worked – as things turned out, successfully – to persuade the CSB that acceptance of the test results would be illegal and would expose the City to disparate-impact liability. But in the event that the CSB was not persuaded, the Mayor, wielding ultimate decision-making authority, was prepared to overrule the CSB immediately. Taking this view of the evidence, a reasonable jury could easily find that the City’s real reason for scrapping the test results was not a concern about violating the disparate-impact provision of Title VII but a simple desire to please a politically important racial constituency. It is noteworthy that the Solicitor General – whose position on the principal legal issue in this case is largely aligned with the dissent – concludes that “[n]either the district court nor the court of appeals .... adequately considered whether, viewing the evidence in the light most favourable to petitioners, a genuine issue of material fact remained whether respondents’ claimed purpose to comply with Title VII was a pretext for intentional racial discrimination ....” Brief for United States as Amicus Curiae 6; see also id., at32–33.
I will not comment at length on the dissent’s criticism of my analysis, but two points require a response.
The first concerns the dissent’s statement that I “equat[e] political considerations with unlawful discrimination.” Post, at 36. The dissent misrepresents my position: I draw no such equation. Of course “there are many ways in which a politician can attempt to win over a constituency – including a racial constituency – without engaging in unlawful discrimination.” Post, at 36–37. But – as I assume the dissent would agree – there are some things that a public official cannot do, and one of those is engaging in intentional racial discrimination when making employment decisions.
The second point concerns the dissent’s main argument – that efforts by the Mayor and his staff to scuttle the test results are irrelevant because the ultimate decision was made by the CSB. According to the dissent, “[t]he relevant decision was made by the CSB,” post, at 34, and there is “scant cause to suspect” that anything done by the opponents of certification, including the Mayor and his staff, “prevented the CSB from even-handedly assessing the reliability of the exams and rendering an independent, good-faith decision on certification,” post, at 36.
Adoption of the dissent’s argument would implicitly decide an important question of Title VII law that this Court has never resolved – the circumstances in which an employer may be held liable based on the discriminatory intent of subordinate employees who influence but do not make the ultimate employment decision. There is a large body of court of appeals case law on this issue, and these cases disagree about the proper standard. See EEOC v BCI Coca-Cola Bottling Co. of Los Angeles, 450 F. 3d 476, 484–488 (CA10 2006) (citing cases and describing the approaches taken in different Circuits). One standard is whether the subordinate “exerted influenc[e] over the titular decision-maker.” Russell v McKinney Hosp. Venture, 235 F. 3d 219, 227 (CA5 2000); see also Poland v Chertoff, 494 F. 3d 1174, 1182 (CA9 2007) (A subordinate’s bias is imputed to the employer where the subordinate “influenced or was involved in the decision or decision-making process”). Another is whether the discriminatory input “caused the adverse employment action.” See BCI Coca-Cola Bottling Co. of Los Angeles, supra, at 487.
In the present cases, a reasonable jury could certainly find that these standards were met. The dissent makes much of the fact that members of the CSB swore under oath that their votes were based on the good-faith belief that certification of the results would have violated federal law. See post, at 34. But the good faith of the CSB members would not preclude a finding that the presentations engineered by the Mayor and his staff influenced or caused the CSB decision.
The least employee-friendly standard asks only whether “the actual decision-maker” acted with discriminatory intent, see Hill v Lockheed Martin Logistics Management, Inc., 354 F. 3d 277, 291 (CA4 2004) (en banc), and it is telling that, even under this standard, summary judgment for respondents would not be proper. This is so because a reasonable jury could certainly find that in New Haven, the Mayor – not the CSB – wielded the final decision-making power. After all, the Mayor claimed that authority and was poised to use it in the event that the CSB decided to accept the test results. See supra, at 9. If the Mayor had the authority to overrule a CSB decision accepting the test results, the Mayor also presumably had the authority to overrule the CSB’s decision rejecting the test results. In light of the Mayor’s conduct, it would be quite wrong to throw out petitioners’ case on the ground that the CSB was the ultimate decision-maker.
* * *
Petitioners are fire-fighters who seek only a fair chance to move up the ranks in their chosen profession. In order to qualify for promotion, they made personal sacrifices. Petitioner Frank Ricci, who is dyslexic, found it necessary to “hir[e] someone, at considerable expense, to read onto audiotape the content of the books and study materials.” App. to Pet. for Cert. in No. 07–1428, at 169a. He “studied an average of eight to thirteen hours a day ...., even listening to audio tapes while driving his car.” Ibid. Petitioner Benjamin Vargas, who is Hispanic, had to “give up a part-time job,” and his wife had to “take leave from her own job in order to take care of their three young children while Vargas studied.” Id., at 176a. “Vargas devoted countless hours to study ...., missed two of his children’s birthdays and over two weeks of vacation time,” and “incurred significant financial expense” during the three-month study period. Id., at 176a–177a.
Petitioners were denied promotions for which they qualified because of the race and ethnicity of the fire-fighters who achieved the highest scores on the City’s exam. The District Court threw out their case on summary judgment, even though that court all but conceded that a jury could find that the City’s asserted justification was pre-textual. The Court of Appeals then summarily affirmed that decision.
The dissent grants that petitioners’ situation is “unfortunate” and that they “understandably attract this Court’s sympathy.” Post, at 1, 39. But “sympathy” is not what petitioners have a right to demand. What they have a right to demand is even-handed enforcement of the law – of Title VII’s prohibition against discrimination based on race. And that is what, until today’s decision, has been denied them.
(with whom Justice Stevens, Justice Souter, and Justice Breyer join, dissenting)
In assessing claims of race discrimination, “[c]ontext matters.” Grutter v Bollinger, 539 U. S. 306, 327 (2003). In 1972, Congress extended Title VII of the Civil Rights Act of 1964 to cover public employment. At that time, municipal fire departments across the country, including New Haven’s, pervasively discriminated against minorities. The extension of Title VII to cover jobs in fire fighting effected no overnight change. It took decades of persistent effort, advanced by Title VII litigation, to open fire fighting posts to members of racial minorities.
The white fire-fighters who scored high on New Haven’s promotional exams understandably attract this Court’s sympathy. But they had no vested right to promotion. Nor have other persons received promotions in preference to them. New Haven maintains that it refused to certify the test results because it believed, for good cause, that it would be vulnerable to a Title VII disparate-impact suit if it relied on those results. The Court today holds that New Haven has not demonstrated “a strong basis in evidence” for its plea. Ante, at 2. In so holding, the Court pretends that “[t]he City rejected the test results solely because the higher scoring candidates were white.” Ante, at 20. That pretension, essential to the Court’s disposition, ignores substantial evidence of multiple flaws in the tests New Haven used. The Court similarly fails to acknowledge the better tests used in other cities, which have yielded less racially skewed outcomes.
By order of this Court, New Haven, a city in which African-Americans and Hispanics account for nearly 60 percent of the population, must today be served – as it was in the days of undisguised discrimination – by a fire department in which members of racial and ethnic minorities are rarely seen in command positions. In arriving at its order, the Court barely acknowledges the path marking decision in Griggs v Duke Power Co., 401 U. S. 424 (1971), which explained the centrality of the disparate-impact concept to effective enforcement of Title VII. The Court’s order and opinion, I anticipate, will not have staying power.
The Court’s recitation of the facts leaves out important parts of the story. Fire fighting is a profession in which the legacy of racial discrimination casts an especially long shadow. In extending Title VII to state and local government employers in 1972, Congress took note of a U. S. Commission on Civil Rights (USCCR) report finding racial discrimination in municipal employment even “more pervasive than in the private sector.” H. R. Rep. No. 92–238, p. 17 (1971). According to the report, overt racism was partly to blame, but so too was a failure on the part of municipal employers to apply merit-based employment principles. In making hiring and promotion decisions, public employers often “rel[ied] on criteria unrelated to job performance,” including nepotism or political patronage. 118 Cong. Rec. 1817 (1972). Such flawed selection methods served to entrench pre-existing racial hierarchies. The USCCR report singled out police and fire departments for having “[b]arriers to equal employment .... greater .... than in any other area of State or local government,” with African-Americans “hold[ing] almost no positions in the officer ranks.” Ibid. See also National Commission on Fire Prevention and Control, America Burning 5 (1973) (“Racial minorities are under-represented in the fire departments in nearly every community in which they live.”).
The city of New Haven (City) was no exception. In the early 1970’s, African-Americans and Hispanics composed 30 percent of New Haven’s population, but only 3.6 percent of the City’s 502 fire-fighters. The racial disparity in the officer ranks was even more pronounced: “[O]f the 107 officers in the Department only one was black, and he held the lowest rank above private.” Firebird Soc. of New Haven, Inc. v New Haven Bd. of Fire Comm’rs, 66 F. R. D. 457, 460 (Conn. 1975).
Following a lawsuit and settlement agreement, see ibid., the City initiated efforts to increase minority representation in the New Haven Fire Department (Department). Those litigation-induced efforts produced some positive change. New Haven’s population includes a greater proportion of minorities today than it did in the 1970’s: Nearly 40 percent of the City’s residents are African-American and more than 20 percent are Hispanic. Among entry-level fire-fighters, minorities are still underrepresented, but not starkly so. As of 2003, African-Americans and Hispanics constituted 30 percent and 16 percent of the City’s fire-fighters, respectively. In supervisory positions, however, significant disparities remain. Overall, the senior officer ranks (captain and higher) are nine percent African-American and nine percent Hispanic. Only one of the Department’s 21 fire captains is African-American. See App. in No. 06–4996–cv (CA2), p. A1588 (hereinafter CA2 App.). It is against this backdrop of entrenched inequality that the promotion process at issue in this litigation should be assessed.
By order of its charter, New Haven must use competitive examinations to fill vacancies in fire officer and other civil-service positions. Such examinations, the City’s civil service rules specify, “shall be practical in nature, shall relate to matters which fairly measure the relative fitness and capacity of the applicants to discharge the duties of the position which they seek, and shall take into account character, training, experience, physical and mental fitness.” Id., at A331. The City may choose among a variety of testing methods, including written and oral exams and “[p]erformance tests to demonstrate skill and ability in performing actual work.” Id., at A332.
New Haven, the record indicates, did not closely consider what sort of “practical” examination would “fairly measure the relative fitness and capacity of the applicants to discharge the duties” of a fire officer. Instead, the City simply adhered to the testing regime outlined in its two-decades-old contract with the local fire-fighters’ union: a written exam, which would account for 60 percent of an applicant’s total score, and an oral exam, which would account for the remaining 40 percent. Id., at A1045. In soliciting bids from exam development companies, New Haven made clear that it would entertain only “proposals that include a written component that will be weighted at 60%, and an oral component that will be weighted at 40%.” Id., at A342. Chad Legel, a representative of the winning bidder, Industrial/Organizational Solutions, Inc. (IOS), testified during his deposition that the City never asked whether alternative methods might better measure the qualities of a successful fire officer, including leadership skills and command presence. See id., at A522 (“I was under contract and had responsibility only to create the oral interview and the written exam.”).
Pursuant to New Haven’s specifications, IOS developed and administered the oral and written exams. The results showed significant racial disparities. On the lieutenant exam, the pass rate for African-American candidates was about one-half the rate for Caucasian candidates; the pass rate for Hispanic candidates was even lower. On the captain exam, both African-American and Hispanic candidates passed at about half the rate of their Caucasian counterparts. See App. 225–226. More striking still, although nearly half of the 77 lieutenant candidates were African-American or Hispanic, none would have been eligible for promotion to the eight positions then vacant. The highest scoring African-American candidate ranked 13th; the top Hispanic candidate was 26th. As for the seven then-vacant captain positions, two Hispanic candidates would have been eligible, but no African-Americans. The highest scoring African-American candidate ranked 15th. See id., at 218–219.
These stark disparities, the Court acknowledges, sufficed to state a prima facie case under Title VII’s disparate-impact provision. See ante, at 27 (“The pass rates of minorities .... f[e]ll well below the 80-percent standard set by the [Equal Employment Opportunity Commission (EEOC)] to implement the disparate-impact provision of Title VII.”). New Haven thus had cause for concern about the prospect of Title VII litigation and liability. City officials referred the matter to the New Haven Civil Service Board (CSB), the entity responsible for certifying the results of employment exams.
Between January and March 2004, the CSB held five public meetings to consider the proper course. At the first meeting, New Haven’s Corporation Counsel, Thomas Ude, described the legal standard governing Title VII disparate-impact claims. Statistical imbalances alone, Ude correctly recognized, do not give rise to liability. Instead, presented with a disparity, an employer “has the opportunity and the burden of proving that the test is job-related and consistent with business necessity.” CA2 App. A724. A Title VII plaintiff may attempt to rebut an employer’s showing of job-relatedness and necessity by identifying alternative selection methods that would have been at least as valid but with “less of an adverse or disparate or discriminatory effect.” Ibid. See also id., at A738. Accordingly, the CSB Commissioners understood, their principal task was to decide whether they were confident about the reliability of the exams: Had the exams fairly measured the qualities of a successful fire officer despite their disparate results? Might an alternative examination process have identified the most qualified candidates without creating such significant racial imbalances?
Seeking a range of input on these questions, the CSB heard from test takers, the test designer, subject-matter experts, City officials, union leaders, and community members. Several candidates for promotion, who did not yet know their exam results, spoke at the CSB’s first two meetings. Some candidates favoured certification. The exams, they emphasized, had closely tracked the assigned study materials. Having invested substantial time and money to prepare themselves for the test, they felt it would be unfair to scrap the results. See, e.g., id., at A772–A773, A785–A789.
Other fire-fighters had a different view. A number of the exam questions, they pointed out, were not germane to New Haven’s practices and procedures. See, e.g., id., at A774–A784. At least two candidates opposed to certification noted unequal access to study materials. Some individuals, they asserted, had the necessary books even before the syllabus was issued. Others had to invest substantial sums to purchase the materials and “wait a month and a half for some of the books because they were on back-order.” Id., at A858. These disparities, it was suggested, fell at least in part along racial lines. While many Caucasian applicants could obtain materials and assistance from relatives in the fire service, the overwhelming majority of minority applicants were “first-generation fire-fighters” without such support networks. See id., at A857–A861, A886–A887.
A representative of the Northeast Region of the International Association of Black Professional Fire-fighters, Donald Day, also spoke at the second meeting. Statistical disparities, he told the CSB, had been present in the Department’s previous promotional exams. On earlier tests, however, a few minority candidates had fared well enough to earn promotions. Id., at A828. See also App. 218–219. Day contrasted New Haven’s experience with that of nearby Bridgeport, where minority fire-fighters held one-third of lieutenant and captain positions. Bridgeport, Day observed, had once used a testing process similar to New Haven’s, with a written exam accounting for 70 percent of an applicant’s score, an oral exam for 25 percent, and seniority for the remaining five percent. CA2 App. A830. Bridgeport recognized, however, that the oral component, more so than the written component, addressed the sort of “real-life scenarios” fire officers encounter on the job. Id., at A832. Accordingly, that city “changed the relative weights” to give primacy to the oral exam. Ibid. Since that time, Day reported, Bridgeport had seen minorities “fairly represented” in its exam results. Ibid.
The CSB’s third meeting featured IOS representative Legel, the leader of the team that had designed and administered the exams for New Haven. Several City officials also participated in the discussion. Legel described the exam development process in detail. The City, he recounted, had set the “parameters” for the exams, specifically, the requirement of written and oral components with a 60/40 weighting. Id., at A923, A974. For security reasons, Department officials had not been permitted to check the content of the questions prior to their administration. Instead, IOS retained a senior fire officer from Georgia to review the exams “for content and fidelity to the source material.” Id., at A936. Legel defended the exams as “facially neutral,” and stated that he “would stand by the[ir] validity.” Id., at A962. City officials did not dispute the neutrality of IOS’s work. But, they cautioned, even if individual exam questions had no intrinsic bias, the selection process as a whole may nevertheless have been deficient. The officials urged the CSB to consult with experts about the “larger picture.” Id., at A1012.
At its fourth meeting, CSB solicited the views of three individuals with testing-related expertise. Dr. Christopher Hornick, an industrial/organizational psychology consultant with 25 years’ experience with police and firefighter testing, described the exam results as having “relatively high adverse impact.” Id., at A1028. Most of the tests he had developed, Hornick stated, exhibited “significantly and dramatically less adverse impact.” Id., at A1029. Hornick downplayed the notion of “facial neutrality.” It was more important, he advised the CSB, to consider “the broader issue of how your procedures and your rules and the types of tests that you are using are contributing to the adverse impact.” Id., at A1038.
Specifically, Hornick questioned New Haven’s union-prompted 60/40 written/oral examination structure, noting the availability of “different types of testing procedures that are much more valid in terms of identifying the best potential supervisors in [the] fire department.” Id., at A1032. He suggested, for example, “an assessment centre process, which is essentially an opportunity for candidates .... to demonstrate how they would address a particular problem as opposed to just verbally saying it or identifying the correct option on a written test.” Id., at A1039–A1040. Such selection processes, Hornick said, better “identif[y] the best possible people” and “demonstrate dramatically less adverse impacts.” Ibid. Hornick added:
“I’ve spoken to at least 10,000, maybe 15,000 fire-fighters in group settings in my consulting practice and I have never one time ever had anyone in the fire service say to me, ‘Well, the person who answers – gets the highest score on a written job knowledge, multiple-guess test makes the best company officer.’ We know that it’s not as valid as other procedures that exist.” Id., at A1033.
See also id., at A1042–A1043 (“I think a person’s leadership skills, their command presence, their interpersonal skills, their management skills, their tactical skills could have been identified and evaluated in a much more appropriate way.”).
Hornick described the written test itself as “reasonably good,” id., at A1041, but he criticized the decision not to allow Department officials to check the content. According to Hornick, this “inevitably” led to “test[ing] for processes and procedures that don’t necessarily match up into the department.” Id., at A1034–A1035. He preferred “experts from within the department who have signed confidentiality agreements .... to make sure that the terminology and equipment that’s being identified from standardized reading sources apply to the department.” Id., at A1035.
Asked whether he thought the City should certify the results, Hornick hedged: “There is adverse impact in the test. That will be identified in any proceeding that you have. You will have industrial psychology experts, if it goes to court, on both sides. And it will not be a pretty or comfortable position for anyone to be in.” Id., at A1040–A1041. Perhaps, he suggested, New Haven might certify the results but immediately begin exploring “alternative ways to deal with these issues” in the future. Id., at A1041.
The two other witnesses made relatively brief appearances. Vincent Lewis, a specialist with the Department of Homeland Security and former fire officer in Michigan, believed the exams had generally tested relevant material, although he noted a relatively heavy emphasis on questions pertaining to being an “apparatus driver.” He suggested that this may have disadvantaged test takers “who had not had the training or had not had an opportunity to drive the apparatus.” Id., at A1051. He also urged the CSB to consider whether candidates had, in fact, enjoyed equal access to the study materials. Ibid. Cf. supra, at 7.
Janet Helms, a professor of counselling psychology at Boston College, observed that two-thirds of the incumbent fire officers who submitted job analyses to IOS during the exam design phase were Caucasian. Members of different racial groups, Helms told the CSB, sometimes do their jobs in different ways, “often because the experiences that are open to white male fire-fighters are not open to members of these other under-represented groups.” CA2 App. A1063–A1064. The heavy reliance on job analyses from white fire-fighters, she suggested, may thus have introduced an element of bias. Id., at A1063.
The CSB’s fifth and final meeting began with statements from City officials recommending against certification. Ude, New Haven’s counsel, repeated the applicable disparate-impact standard:
“[A] finding of adverse impact is the beginning, not the end, of a review of testing procedures. Where a procedure demonstrates adverse impact, you look to how closely it is related to the job that you’re looking to fill and you also look at whether there are other ways to test for those qualities, those traits, those positions that are equally valid with less adverse impact.” Id., at A1100–A1101.
New Haven, Ude and other officials asserted, would be vulnerable to Title VII liability under this standard. Even if the exams were “facially neutral,” significant doubts had been raised about whether they properly assessed the key attributes of a successful fire officer. Id., at A1103. See also id., at A1125 (“Upon close reading of the exams, the questions themselves would appear to test a candidate’s ability to memorize textbooks but not necessarily to identify solutions to real problems on the fire ground.”). Moreover, City officials reminded the CSB, Hornick and others had identified better, less discriminatory selection methods–such as assessment centres or exams with a more heavily weighted oral component. Id., at A1108–A1109, A1129–A1130.
After giving members of the public a final chance to weigh in, the CSB voted on certification, dividing 2 to 2. By rule, the result was non-certification. Voting no, Commissioner Webber stated, “I originally was going to vote to certify. .... But I’ve heard enough testimony here to give me great doubts about the test itself and .... some of the procedures. And I believe we can do better.” Id., at A1157. Commissioner Tirado likewise concluded that the “flawed” testing process counselled against certification. Id., at A1158. Chairman Segaloff and Commissioner Caplan voted to certify. According to Segaloff, the testimony had not “compelled [him] to say this exam was not job-related,” and he was unconvinced that alternative selection processes would be “less discriminatory.” Id., at A1159–A1160. Both Segalhoff and Caplan, however, urged the City to undertake civil service reform. Id., at A1150–A1154.
Following the CSB’s vote, petitioners – 17 white fire-fighters and one Hispanic firefighter, all of whom had high marks on the exams – filed suit in the United States District Court for the District of Connecticut. They named as defendants – respondents here – the City, several City officials, a local political activist, and the two CSB members who voted against certifying the results. By opposing certification, petitioners alleged, respondents had discriminated against them in violation of Title VII’s disparate-treatment provision and the Fourteenth Amendment ’s Equal Protection Clause. The decision not to certify, respondents answered, was a lawful effort to comply with Title VII’s disparate-impact provision and thus could not have run afoul of Title VII’s prohibition of disparate treatment. Characterizing respondents’ stated rationale as a mere pretext, petitioners insisted that New Haven would have had a solid defence to any disparate-impact suit.
In a decision summarily affirmed by the Court of Appeals, the District Court granted summary judgment for respondents. 554 F. Supp. 2d 142 (Conn. 2006), aff’d, 530 F. 3d 87 (CA2 2008) (per curiam). Under Second Circuit precedent, the District Court explained, “the intent to remedy the disparate impact” of a promotional exam “is not equivalent to an intent to discriminate against non-minority applicants.” 554 F. Supp. 2d, at 157 (quoting Hayden v County of Nassau, 180 F. 3d 42, 51 (CA2 1999)). Rejecting petitioners’ pretext argument, the court observed that the exam results were sufficiently skewed “to make out a prima facie case of discrimination” under Title VII’s disparate-impact provision. 554 F. Supp. 2d, at 158. Had New Haven gone forward with certification and been sued by aggrieved minority test takers, the City would have been forced to defend tests that were presumptively invalid. And, as the CSB testimony of Hornick and others indicated, overcoming that presumption would have been no easy task. Id., at 153–156. Given Title VII’s preference for voluntary compliance, the court held, New Haven could lawfully discard the disputed exams even if the City had not definitively “pinpoint[ed]” the source of the disparity and “ha[d] not yet formulated a better selection method.” Id., at 156.
Respondents were no doubt conscious of race during their decision-making process, the court acknowledged, but this did not mean they had engaged in racially disparate treatment. The conclusion they had reached and the action thereupon taken were race-neutral in this sense: “[A]ll the test results were discarded, no one was promoted, and fire-fighters of every race will have to participate in another selection process to be considered for promotion.” Id., at 158. New Haven’s action, which gave no individual a preference, “was ‘simply not analogous to a quota system or a minority set-aside where candidates, on the basis of their race, are not treated uniformly.’ ” Id., at 157 (quoting Hayden, 180 F. 3d, at 50). For these and other reasons, the court also rejected petitioners’ equal protection claim.
Title VII became effective in July 1965. Employers responded to the law by eliminating rules and practices that explicitly barred racial minorities from “white” jobs. But removing overtly race-based job classifications did not usher in genuinely equal opportunity. More subtle – and sometimes unconscious – forms of discrimination replaced once undisguised restrictions.
In Griggs v Duke Power Co., 401 U. S. 424 (1971), this Court responded to that reality and supplied important guidance on Title VII’s mission and scope. Congress, the landmark decision recognized, aimed beyond “disparate treatment”; it targeted “disparate impact” as well. Title VII’s original text, it was plain to the Court, “proscribe[d] not only overt discrimination but also practices that are fair in form, but discriminatory in operation.” Id., at 431. Only by ignoring Griggs could one maintain that intentionally disparate treatment alone was Title VII’s “original, foundational prohibition,” and disparate impact a mere afterthought. Cf. ante, at 21.
Griggs addressed Duke Power Company’s policy that applicants for positions, save in the company’s labour department, be high school graduates and score satisfactorily on two professionally prepared aptitude tests. “[T]here was no showing of a discriminatory purpose in the adoption of the diploma and test requirements.” 401 U. S., at 428. The policy, however, “operated to render ineligible a markedly disproportionate number of [African-Americans].” Id., at 429. At the time of the litigation, in North Carolina, where the Duke Power plant was located, 34 percent of white males, but only 12 percent of African-American males, had high school diplomas. Id., at 430, n. 6. African-Americans also failed the aptitude tests at a significantly higher rate than whites. Ibid. Neither requirement had been “shown to bear a demonstrable relationship to successful performance of the jobs for which it was used.” Id., at 431.
The Court unanimously held that the company’s diploma and test requirements violated Title VII. “[T]o achieve equality of employment opportunities,” the Court comprehended, Congress “directed the thrust of the Act to the consequences of employment practices, not simply the motivation.” Id., at 429, 432. That meant “unnecessary barriers to employment” must fall, even if “neutral on their face” and “neutral in terms of intent.” Id., at 430, 431. “The touchstone” for determining whether a test or qualification meets Title VII’s measure, the Court said, is not “good intent or the absence of discriminatory intent”; it is “business necessity.” Id., at 431, 432. Matching procedure to substance, the Griggs Court observed, Congress “placed on the employer the burden of showing that any given requirement .... ha[s] a manifest relationship to the employment in question.” Id., at 432.
In Albemarle Paper Co. v Moody, 422 U. S. 405 (1975), the Court, again without dissent, elaborated on Griggs. When an employment test “select[s] applicants for hire or promotion in a racial pattern significantly different from the pool of applicants,” the Court reiterated, the employer must demonstrate a “manifest relationship” between test and job. 422 U. S., at 425. Such a showing, the Court cautioned, does not necessarily mean the employer prevails: “[I]t remains open to the complaining party to show that other tests or selection devices, without a similarly undesirable racial effect, would also serve the employer’s legitimate interest in ‘efficient and trustworthy workmanship.’ ” Ibid.
Federal trial and appellate courts applied Griggs and Albemarle to disallow a host of hiring and promotion practices that “operate[d] as ‘built in headwinds’ for minority groups.” Griggs, 401 U. S., at 432. Practices discriminatory in effect, courts repeatedly emphasized, could be maintained only upon an employer’s showing of “an overriding and compelling business purpose.” Chrisner v Complete Auto Transit, Inc., 645 F. 2d 1251, 1261, n. 9 (CA6 1981). That a practice served “legitimate management functions” did not, it was generally understood, suffice to establish business necessity. Williams v Colorado Springs, Colo., School Dist., 641 F. 2d 835, 840–841 (CA10 1981) (internal quotation marks omitted). Among selection methods cast aside for lack of a “manifest relationship” to job performance were a number of written hiring and promotional examinations for fire-fighters.
Moving in a different direction, in Wards Cove Packing Co. v Atonio, 490 U. S. 642 (1989), a bare majority of this Court significantly modified the Griggs-Albemarle delineation of Title VII’s disparate-impact proscription. As to business necessity for a practice that disproportionately excludes members of minority groups, Wards Cove held, the employer bears only the burden of production, not the burden of persuasion. 490 U. S., at 659–660. And in place of the instruction that the challenged practice “must have a manifest relationship to the employment in question,” Griggs, 401 U. S., at 432, Wards Cove said that the practice would be permissible as long as it “serve[d], in a significant way, the legitimate employment goals of the employer.” 490 U. S., at 659.
In response to Wards Cove and “a number of [other] recent decisions by the United States Supreme Court that sharply cut back on the scope and effectiveness of [civil rights] laws,” Congress enacted the Civil Rights Act of 1991. H. R. Rep. No. 102–40, pt. 2, p. 2 (1991). Among the 1991 alterations, Congress formally codified the disparate-impact component of Title VII. In so amending the statute, Congress made plain its intention to restore “the concepts of ‘business necessity’ and ‘job related’ enunciated by the Supreme Court in Griggs v Duke Power Co. .... and in other Supreme Court decisions prior to Wards Cove Packing Co. v Atonio.” §3(2), 105 Stat. 1071. Once a complaining party demonstrates that an employment practice causes a disparate impact, amended Title VII states, the burden is on the employer “to demonstrate that the challenged practice is job related for the position in question and consistent with business necessity.” 42 U. S. C. §2000e–2(k)(1)(A)(i). If the employer carries that substantial burden, the complainant may respond by identifying “an alternative employment practice” which the employer “refuses to adopt.” §2000e–2(k)(1)(A)(ii), (C).
Neither Congress’ enactments nor this Court’s Title VII precedents (including the now-discredited decision in Wards Cove) offer even a hint of “conflict” between an employer’s obligations under the statute’s disparate-treatment and disparate-impact provisions. Cf. ante, at 20. Standing on an equal footing, these twin pillars of Title VII advance the same objectives: ending workplace discrimination and promoting genuinely equal opportunity. See McDonnell Douglas Corp. v Green, 411 U. S. 792, 800 (1973).
Yet the Court today sets at odds the statute’s core directives. When an employer changes an employment practice in an effort to comply with Title VII’s disparate-impact provision, the Court reasons, it acts “because of race” – something Title VII’s disparate-treatment provision, see §2000e–2(a)(1), generally forbids. Ante, at 20. This characterization of an employer’s compliance-directed action shows little attention to Congress’ design or to the Griggs line of cases Congress recognized as path marking.
“[O]ur task in interpreting separate provisions of a single Act is to give the Act the most harmonious, comprehensive meaning possible in light of the legislative policy and purpose.” Weinberger v Hynson, Westcott & Dunning, Inc., 412 U. S. 609, 631–632 (1973) (internal quotation marks omitted). A particular phrase need not “extend to the outer limits of its definitional possibilities” if an incongruity would result. Dolan v Postal Service, 546 U. S. 481, 486 (2006). Here, Title VII’s disparate-treatment and disparate-impact proscriptions must be read as complementary.
In codifying the Griggs and Albemarle instructions, Congress declared unambiguously that selection criteria operating to the disadvantage of minority group members can be retained only if justified by business necessity. In keeping with Congress’ design, employers who reject such criteria due to reasonable doubts about their reliability can hardly be held to have engaged in discrimination “because of” race. A reasonable endeavour to comply with the law and to ensure that qualified candidates of all races have a fair opportunity to compete is simply not what Congress meant to interdict. I would therefore hold that an employer who jettisons a selection device when its disproportionate racial impact becomes apparent does not violate Title VII’s disparate-treatment bar automatically or at all, subject to this key condition: The employer must have good cause to believe the device would not withstand examination for business necessity. Cf. Faragher v Boca Raton, 524 U. S. 775, 806 (1998) (observing that it accords with “clear statutory policy” for employers “to prevent violations” and “make reasonable efforts to discharge their duty” under Title VII).
EEOC’s interpretative guidelines are corroborative. “[B]y the enactment of title VII,” the guidelines state, “Congress did not intend to expose those who comply with the Act to charges that they are violating the very statute they are seeking to implement.” 29 CFR §1608.1(a) (2008). Recognizing EEOC’s “enforcement responsibility” under Title VII, we have previously accorded the Commission’s position respectful consideration. See, e.g., Albemarle, 422 U. S., at 431; Griggs, 401 U. S., at 434. Yet the Court today does not so much as mention EEOC’s counsel.
Our precedents defining the contours of Title VII’s disparate-treatment prohibition further confirm the absence of any intra-statutory discord. In Johnson v Transportation Agency, Santa Clara Cty., 480 U. S. 616 (1987), we upheld a municipal employer’s voluntary affirmative-action plan against a disparate-treatment challenge. Pursuant to the plan, the employer selected a woman for a road-dispatcher position, a job category traditionally regarded as “male.” A male applicant who had a slightly higher interview score brought suit under Title VII. This Court rejected his claim and approved the plan, which allowed consideration of gender as “one of numerous factors.” Id., at 638. Such consideration, we said, is “fully consistent with Title VII” because plans of that order can aid “in eliminating the vestiges of discrimination in the workplace.” Id., at 642.
This litigation does not involve affirmative action. But if the voluntary affirmative action at issue in Johnson does not discriminate within the meaning of Title VII, neither does an employer’s reasonable effort to comply with Title VII’s disparate-impact provision by refraining from action of doubtful consistency with business necessity.
To “reconcile” the supposed “conflict” between disparate treatment and disparate impact, the Court offers an enigmatic standard. Ante, at 20. Employers may attempt to comply with Title VII’s disparate-impact provision, the Court declares, only where there is a “strong basis in evidence” documenting the necessity of their action. Ante, at 22. The Court’s standard, drawn from inapposite equal protection precedents, is not elaborated. One is left to wonder what cases would meet the standard and why the Court is so sure this case does not.
In construing Title VII, I note preliminarily, equal protection doctrine is of limited utility. The Equal Protection Clause, this Court has held, prohibits only intentional discrimination; it does not have a disparate-impact component. See Personnel Administrator of Mass. v Feeney, 442 U. S. 256, 272 (1979); Washington v Davis, 426 U. S. 229, 239 (1976). Title VII, in contrast, aims to eliminate all forms of employment discrimination, unintentional as well as deliberate. Until today, cf. ante, at 25; ante, p. 1 (Scalia, J., concurring), this Court has never questioned the constitutionality of the disparate-impact component of Title VII, and for good reason. By instructing employers to avoid needlessly exclusionary selection processes, Title VII’s disparate-impact provision calls for a “race-neutral means to increase minority .... participation” – something this Court’s equal protection precedents also encourage. See Adarand Constructors, Inc. v Peńa, 515 U. S. 200, 238 (1995) (quoting Richmond v J. A. Croson Co., 488 U. S. 469, 507 (1989)). “The very radicalism of holding disparate impact doctrine unconstitutional as a matter of equal protection,” moreover, “suggests that only a very uncompromising court would issue such a decision.” Primus, Equal Protection and Disparate Impact: Round Three, 117 Harv. L. Rev. 493, 585 (2003).
The cases from which the Court draws its strong-basis-in-evidence standard are particularly inapt; they concern the constitutionality of absolute racial preferences. See Wygant v Jackson Bd. of Ed., 476 U. S. 267, 277 (1986) (plurality opinion) (invalidating a school district’s plan to lay off nonminority teachers while retaining minority teachers with less seniority); Croson, 488 U. S., at 499–500 (rejecting a set-aside program for minority contractors that operated as “an unyielding racial quota”). An employer’s effort to avoid Title VII liability by repudiating a suspect selection method scarcely resembles those cases. Race was not merely a relevant consideration in Wygant and Croson; it was the decisive factor. Observance of Title VII’s disparate-impact provision, in contrast, calls for no racial preference, absolute or otherwise. The very purpose of the provision is to ensure that individuals are hired and promoted based on qualifications manifestly necessary to successful performance of the job in question, qualifications that do not screen out members of any race.
The Court’s decision in this litigation underplays a dominant Title VII theme. This Court has repeatedly emphasized that the statute “should not be read to thwart” efforts at voluntary compliance. Johnson, 480 U. S., at 630. Such compliance, we have explained, is “the preferred means of achieving [Title VII’s] objectives.” Fire-fighters v Cleveland, 478 U. S. 501, 515 (1986). See also Kolstad v American Dental Assn., 527 U. S. 526, 545 (1999) (“Dissuading employers from [taking voluntary action] to prevent discrimination in the workplace is directly contrary to the purposes underlying Title VII.”); 29 CFR §1608.1(c). The strong-basis-in-evidence standard, however, as barely described in general, and cavalierly applied in this case, makes voluntary compliance a hazardous venture.
As a result of today’s decision, an employer who discards a dubious selection process can anticipate costly disparate-treatment litigation in which its chances for success – even for surviving a summary-judgment motion – are highly problematic. Concern about exposure to disparate-impact liability, however well grounded, is insufficient to insulate an employer from attack. Instead, the employer must make a “strong” showing that
its selection method was “not job related and consistent with business necessity,” or
that it refused to adopt “an equally valid, less-discriminatory alternative.” Ante, at 28.
It is hard to see how these requirements differ from demanding that an employer establish “a provable, actual violation” against itself. Cf. ante, at 24. There is indeed a sharp conflict here, but it is not the false one the Court describes between Title VII’s core provisions. It is, instead, the discordance of the Court’s opinion with the voluntary compliance ideal. Cf. Wygant, 476 U. S., at 290 (O’Connor, J., concurring in part and concurring in judgment) (“The imposition of a requirement that public employers make findings that they have engaged in illegal discrimination before they [act] would severely undermine public employers’ incentive to meet voluntarily their civil rights obligations.”).
The Court’s additional justifications for announcing a strong-basis-in-evidence standard are unimpressive. First, discarding the results of tests, the Court suggests, calls for a heightened standard because it “upset[s] an employee’s legitimate expectation.” Ante, at 25. This rationale puts the cart before the horse. The legitimacy of an employee’s expectation depends on the legitimacy of the selection method. If an employer reasonably concludes that an exam fails to identify the most qualified individuals and needlessly shuts out a segment of the applicant pool, Title VII surely does not compel the employer to hire or promote based on the test, however unreliable it may be. Indeed, the statute’s prime objective is to prevent exclusionary practices from “operat[ing] to ‘freeze’ the status quo.” Griggs, 401 U. S., at 430.
Second, the Court suggests, anything less than a strong-basis-in-evidence standard risks creating “a de facto quota system, in which .... an employer could discard test results .... with the intent of obtaining the employer’s preferred racial balance.” Ante, at 22. Under a reasonableness standard, however, an employer could not cast aside a selection method based on a statistical disparity alone. The employer must have good cause to believe that the method screens out qualified applicants and would be difficult to justify as grounded in business necessity. Should an employer repeatedly reject test results, it would be fair, I agree, to infer that the employer is simply seeking a racially balanced outcome and is not genuinely endeavouring to comply with Title VII.
The Court stacks the deck further by denying respondents any chance to satisfy the newly announced strong-basis-in-evidence standard. When this Court formulates a new legal rule, the ordinary course is to remand and allow the lower courts to apply the rule in the first instance. See, e.g., Johnson v California, 543 U. S. 499, 515 (2005); Pullman-Standard v Swint, 456 U. S. 273, 291 (1982). I see no good reason why the Court fails to follow that course in this case. Indeed, the sole basis for the Court’s peremptory ruling is the demonstrably false pretension that respondents showed “nothing more” than “a significant statistical disparity.” Ante, at 27–28; see supra, at 24, n. 8.
Applying what I view as the proper standard to the record thus far made, I would hold that New Haven had ample cause to believe its selection process was flawed and not justified by business necessity. Judged by that standard, petitioners have not shown that New Haven’s failure to certify the exam results violated Title VII’s disparate-treatment provision.
The City, all agree, “was faced with a prima facie case of disparate-impact liability,” ante, at 27: The pass rate for minority candidates was half the rate for nonminority candidates, and virtually no minority candidates would have been eligible for promotion had the exam results been certified. Alerted to this stark disparity, the CSB heard expert and lay testimony, presented at public hearings, in an endeavour to ascertain whether the exams were fair and consistent with business necessity. Its investigation revealed grave cause for concern about the exam process itself and the City’s failure to consider alternative selection devices.
Chief among the City’s problems was the very nature of the tests for promotion. In choosing to use written and oral exams with a 60/40 weighting, the City simply adhered to the union’s preference and apparently gave no consideration to whether the weighting was likely to identify the most qualified fire-officer candidates. There is strong reason to think it was not.
Relying heavily on written tests to select fire officers is a questionable practice, to say the least. Successful fire officers, the City’s description of the position makes clear, must have the “[a]bility to lead personnel effectively, maintain discipline, promote harmony, exercise sound judgment, and cooperate with other officials.” CA2 App. A432. These qualities are not well measured by written tests. Testifying before the CSB, Christopher Hornick, an exam-design expert with more than two decades of relevant experience, was emphatic on this point: Leadership skills, command presence, and the like “could have been identified and evaluated in a much more appropriate way.” Id., at A1042–A1043.
Hornick’s commonsense observation is mirrored in case law and in Title VII’s administrative guidelines. Courts have long criticized written firefighter promotion exams for being “more probative of the test-taker’s ability to recall what a particular text stated on a given topic than of his fire fighting or supervisory knowledge and abilities.” Vulcan Pioneers, Inc. v New Jersey Dept. of Civil Serv., 625 F. Supp. 527, 539 (NJ 1985). A fire officer’s job, courts have observed, “involves complex behaviours, good interpersonal skills, the ability to make decisions under tremendous pressure, and a host of other abilities – none of which is easily measured by a written, multiple choice test.” Fire-fighters Inst. for Racial Equality v St. Louis, 616 F. 2d 350, 359 (CA8 1980). Interpreting the Uniform Guidelines, EEOC and other federal agencies responsible for enforcing equal opportunity employment laws have similarly recognized that, as measures of “interpersonal relations” or “ability to function under danger (e.g., fire-fighters),” “[p]encil-and-paper tests .... generally are not close enough approximations of work behaviours to show content validity.” 44 Fed. Reg. 12007 (1979). See also 29 CFR §1607.15(C)(4).
Given these unfavourable appraisals, it is unsurprising that most municipal employers do not evaluate their fire-officer candidates as New Haven does. Although comprehensive statistics are scarce, a 1996 study found that nearly two-thirds of surveyed municipalities used assessment centres (“simulations of the real world of work”) as part of their promotion processes. P. Lowry, A Survey of the Assessment Centre Process in the Public Sector, 25 Public Personnel Management 307, 315 (1996). That figure represented a marked increase over the previous decade, see ibid., so the percentage today may well be even higher. Among municipalities still relying in part on written exams, the median weight assigned to them was 30 percent – half the weight given to New Haven’s written exam. Id., at 309.
Testimony before the CSB indicated that these alternative methods were both more reliable and notably less discriminatory in operation. According to Donald Day of the International Association of Black Professional Fire-fighters, nearby Bridgeport saw less skewed results after switching to a selection process that placed primary weight on an oral exam. CA2 App. A830–A832; see supra, at 7–8. And Hornick described assessment centres as “demonstrat[ing] dramatically less adverse impacts” than written exams. CA2 App. A1040. Considering the prevalence of these proven alternatives, New Haven was poorly positioned to argue that promotions based on its outmoded and exclusionary selection process qualified as a business necessity. Cf. Robinson v Lorillard Corp., 444 F. 2d 791, 798, n. 7 (CA4 1971) (“It should go without saying that a practice is hardly ‘necessary’ if an alternative practice better effectuates its intended purpose or is equally effective but less discriminatory.”).
Ignoring the conceptual and other defects in New Haven’s selection process, the Court describes the exams as “painstaking[ly]” developed to test “relevant” material and on that basis finds no substantial risk of disparate-impact liability. See ante, at 28. Perhaps such reasoning would have sufficed under Wards Cove, which permitted exclusionary practices as long as they advanced an employer’s “legitimate” goals. 490 U. S., at 659. But Congress repudiated Wards Cove and reinstated the “business necessity” rule attended by a “manifest relationship” requirement. See Griggs, 401 U. S., at 431–432. See also supra, at 17. Like the chess player who tries to win by sweeping the opponent’s pieces off the table, the Court simply shuts from its sight the formidable obstacles New Haven would have faced in defending against a disparate-impact suit. See Lanning v Southeastern Pa. Transp. Auth., 181 F. 3d 478, 489 (CA3 1999) (“Judicial application of a standard focusing solely on whether the qualities measured by an .... exam bear some relationship to the job in question would impermissibly write out the business necessity prong of the Act’s chosen standard.”).
That IOS representative Chad Legel and his team may have been diligent in designing the exams says little about the exams’ suitability for selecting fire officers. IOS worked within the City’s constraints. Legel never discussed with the City the propriety of the 60/40 weighting and “was not asked to consider the possibility of an assessment centre.” CA2 App. A522. See also id., at A467. The IOS exams, Legel admitted, had not even attempted to assess “command presence”: “[Y]ou would probably be better off with an assessment centre if you cared to measure that.” Id., at A521. Cf. Boston Chapter, NAACP v Beecher, 504 F. 2d 1017, 1021–1022 (CA1 1974) (“A test fashioned from materials pertaining to the job .... superficially may seem job-related. But what is at issue is whether it demonstrably selects people who will perform better the required on-the-job behaviours.”).
In addition to the highly questionable character of the exams and the neglect of available alternatives, the City had other reasons to worry about its vulnerability to disparate-impact liability. Under the City’s ground rules, IOS was not allowed to show the exams to anyone in the New Haven Fire Department prior to their administration. This “precluded [IOS] from being able to engage in [its] normal subject matter expert review process” – something Legel described as “very critical.” CA2 App. A477, A506. As a result, some of the exam questions were confusing or irrelevant, and the exams may have over-tested some subject-matter areas while missing others. See, e.g., id., at A1034–A1035, A1051. Testimony before the CSB also raised questions concerning unequal access to study materials, see id., at A857–A861, and the potential bias introduced by relying principally on job analyses from nonminority fire officers to develop the exams, see id., at A1063–A1064. See also supra, at 7, 10.
The Court criticizes New Haven for failing to obtain a “technical report” from IOS, which, the Court maintains, would have provided “detailed information to establish the validity of the exams.” Ante, at 29. The record does not substantiate this assertion. As Legel testified during his deposition, the technical report merely summarized “the steps that [IOS] took methodologically speaking,” and would not have established the exams’ reliability. CA2 App. A461. See also id., at A462 (the report “doesn’t say anything that other documents that already existed wouldn’t say”).
In sum, the record solidly establishes that the City had good cause to fear disparate-impact liability. Moreover, the Court supplies no tenable explanation why the evidence of the tests’ multiple deficiencies does not create at least a triable issue under a strong-basis-in-evidence standard.
Concurring in the Court’s opinion, Justice Alito asserts that summary judgment for respondents would be improper even if the City had good cause for its non-certification decision. A reasonable jury, he maintains, could have found that respondents were not actually motivated by concern about disparate-impact litigation, but instead sought only “to placate a politically important [African-American] constituency.” Ante, at 3. As earlier noted, I would not oppose a remand for further proceedings fair to both sides. See supra, at 26, n. 10. It is the Court that has chosen to short-circuit this litigation based on its pretension that the City has shown, and can show, nothing more than a statistical disparity. See supra, at 24, n. 8, 25. Justice Alito compounds the Court’s error.
Offering a truncated synopsis of the many hours of deliberations undertaken by the CSB, Justice Alito finds evidence suggesting that respondents’ stated desire to comply with Title VII was insincere, a mere “pretext” for discrimination against white fire-fighters. Ante, at 2–3. In support of his assertion, Justice Alito recounts at length the alleged machinations of Rev. Boise Kimber (a local political activist), Mayor John DeStefano, and certain members of the mayor’s staff. See ante, at 3–10.
Most of the allegations Justice Alito repeats are drawn from petitioners’ statement of facts they deem undisputed, a statement displaying an adversarial zeal not uncommonly found in such presentations. What cannot credibly be denied, however, is that the decision against certification of the exams was made neither by Kimber nor by the mayor and his staff. The relevant decision was made by the CSB, an unelected, politically insulated body. It is striking that Justice Alito’s concurrence says hardly a word about the CSB itself, perhaps because there is scant evidence that its motivation was anything other than to comply with Title VII’s disparate-impact provision. Notably, petitioners did not even seek to take depositions of the two commissioners who voted against certification. Both submitted uncontested affidavits declaring unequivocally that their votes were “based solely on [their] good faith belief that certification” would have discriminated against minority candidates in violation of federal law. CA2 App. A1605, A1611.
Justice Alito discounts these sworn statements, suggesting that the CSB’s deliberations were tainted by the preferences of Kimber and City officials, whether or not the CSB itself was aware of the taint. Kimber and City officials, Justice Alito speculates, decided early on to oppose certification and then “engineered” a skewed presentation to the CSB to achieve their preferred outcome. Ante, at 12.
As an initial matter, Justice Alito exaggerates the influence of these actors. The CSB, the record reveals, designed and conducted an inclusive decision-making process, in which it heard from numerous individuals on both sides of the certification question. See, e.g., CA2 App. A1090. Kimber and others no doubt used strong words to urge the CSB not to certify the exam results, but the CSB received “pressure” from supporters of certification as well as opponents. Cf. ante, at 6. Petitioners, for example, engaged counsel to speak on their behalf before the CSB. Their counsel did not mince words: “[I]f you discard these results,” she warned, “you will get sued. You will force the taxpayers of the city of New Haven into protracted litigation.” CA2 App. A816. See also id., at A788.
The local fire-fighters union – an organization required by law to represent all the City’s fire-fighters – was similarly outspoken in favour of certification. Discarding the test results, the union’s president told the CSB, would be “totally ridiculous.” Id., at A806. He insisted, inaccurately, that the City was not at risk of disparate-impact liability because the exams were administered pursuant to “a collective bargaining agreement.” Id., at A1137. Cf. supra, at 26–27, n. 11. Never mentioned by Justice Alito in his attempt to show testing expert Christopher Hornick’s alliance with the City, ante, at 8–9, the CSB solicited Hornick’s testimony at the union’s suggestion, not the City’s. CA2 App. A1128. Hornick’s cogent testimony raised substantial doubts about the exams’ reliability. See supra, at 8–10.
There is scant cause to suspect that manoeuvring or overheated rhetoric, from either side, prevented the CSB from even-handedly assessing the reliability of the exams and rendering an independent, good-faith decision on certification. Justice Alito acknowledges that the CSB had little patience for Kimber’s antics. Ante, at 6–7. As to petitioners, Chairman Segaloff – who voted to certify the exam results – dismissed the threats made by their counsel as unhelpful and needlessly “inflammatory.” CA2 App. A821. Regarding the views expressed by City officials, the CSB made clear that they were entitled to no special weight. Id., at A1080.
In any event, Justice Alito’s analysis contains a more fundamental flaw: It equates political considerations with unlawful discrimination. As Justice Alito sees it, if the mayor and his staff were motivated by their desire “to placate a .... racial constituency,” ante, at 3, then they engaged in unlawful discrimination against petitioners. But Justice Alito fails to ask a vital question: “[P]lacate” how? That political officials would have politics in mind is hardly extraordinary, and there are many ways in which a politician can attempt to win over a constituency – including a racial constituency – without engaging in unlawful discrimination. As courts have recognized, “[p]oliticians routinely respond to bad press ...., but it is not a violation of Title VII to take advantage of a situation to gain political favour.” Henry v Jones, 507 F. 3d 558, 567 (CA7 2007).
The real issue, then, is not whether the mayor and his staff were politically motivated; it is whether their attempt to score political points was legitimate (i.e., non-discriminatory). Were they seeking to exclude white fire-fighters from promotion (unlikely, as a fair test would undoubtedly result in the addition of white fire-fighters to the officer ranks), or did they realize, at least belatedly, that their tests could be toppled in a disparate-impact suit? In the latter case, there is no disparate-treatment violation. Justice Alito, I recognize, would disagree. In his view, an employer’s action to avoid Title VII disparate-impact liability qualifies as a presumptively improper race-based employment decision. See ante, at 2. I reject that construction of Title VII. See supra, at 18–20. As I see it, when employers endeavour to avoid exposure to disparate-impact liability, they do not thereby encounter liability for disparate treatment.
Applying this understanding of Title VII, supported by Griggs and the long line of decisions following Griggs, see supra, at 16–17, and nn. 3–4, the District Court found no genuine dispute of material fact. That court noted, particularly, the guidance furnished by Second Circuit precedent. See supra, at 12. Petitioners’ allegations that City officials took account of politics, the District Court determined, simply “d[id] not suffice” to create an inference of unlawful discrimination. 554 F. Supp. 2d, at 160, n. 12. The non-certification decision, even if undertaken “in a political context,” reflected a legitimate “intent not to implement a promotional process based on testing results that had an adverse impact.” Id., at 158, 160. Indeed, the District Court perceived “a total absence of any evidence of discriminatory animus towards [petitioners].” Id., at 158. See also id., at 162 (“Nothing in the record in this case suggests that the City defendants or CSB acted ‘because of’ discriminatory animus toward [petitioners] or other non-minority applicants for promotion.”). Perhaps the District Court could have been more expansive in its discussion of these issues, but its conclusions appear entirely consistent with the record before it.
It is indeed regrettable that the City’s non-certification decision would have required all candidates to go through another selection process. But it would have been more regrettable to rely on flawed exams to shut out candidates who may well have the command presence and other qualities needed to excel as fire officers. Yet that is the choice the Court makes today. It is a choice that breaks the promise of Griggs that groups long denied equal opportunity would not be held back by tests “fair in form, but discriminatory in operation.” 401 U. S., at 431.
* * *
This case presents an unfortunate situation, one New Haven might well have avoided had it utilized a better selection process in the first place. But what this case does not present is race-based discrimination in violation of Title VII. I dissent from the Court’s judgment, which rests on the false premise that respondents showed “a significant statistical disparity,” but “nothing more.” See ante, at 27–28.
 Although the dissent disputes it, see post, at 33–34, n. 17, the record certainly permits the inference that petitioners’ allegation is true. See App. to Pet. for Cert. in No. 07–1428, pp. 846a–851a (deposition of Dubois-Walton).
 The City’s heavy reliance on Hornick’s testimony makes the two chiefs’ silence all the more striking. See supra, at 5. While Hornick knew little or nothing about the tests he criticized, the two chiefs were involved “during the lengthy process that led to the devising of the administration of these exams,” App. to Pet. for Cert. in No. 07–1428, at 847a, including “collaborating with City officials on the extensive job analyses that were done,” “selection of the oral panellists,” and selection of “the proper content and subject matter of the exams,” id., at 847a–848a.
 Never mind the flawed tests New Haven used and the better selection methods used elsewhere, Justice Alito’s concurring opinion urges. Overriding all else, racial politics, fired up by a strident African-American pastor, were at work in New Haven. See ante, at 4–9. Even a detached and disinterested observer, however, would have every reason to ask: Why did such racially skewed results occur in New Haven, when better tests likely would have produced less disproportionate results?
 The Court’s disparate-impact analysis rested on two provisions of Title VII: §703(a)(2), which made it unlawful for an employer “to limit, segregate, or classify his employees in any way which would deprive or tend to deprive any individual of employment opportunities or otherwise adversely affect his status as an employee, because of such individual’s race, colour, religion, sex, or national origin”; and §703(h), which permitted employers “to act upon the results of any professionally developed ability test provided that such test, its administration or action upon the results is not designed, intended or used to discriminate because of race, colour, religion, sex or national origin.” Griggs v Duke Power Co., 401 U. S. 424, 426, n. 1 (1971) (quoting 78 Stat. 255, 42 U. S. C. §2000e–2(a)(2), (h) (1964 ed.)). See also 401 U. S., at 433–436 (explaining that §703(h) authorizes only tests that are “demonstrably a reasonable measure of job performance”).
 See also Dothard v Rawlinson, 433 U. S. 321, 332, n. 14 (1977) (“a discriminatory employment practice must be shown to be necessary to safe and efficient job performance to survive a Title VII challenge”); Williams v Colorado Springs, Colo., School Dist., 641 F. 2d 835, 840–841 (CA10 1981) (“The term ‘necessity’ connotes that the exclusionary practice must be shown to be of great importance to job performance.”); Kirby v Colony Furniture Co., 613 F. 2d 696, 705, n. 6 (CA8 1980) (“the proper standard for determining whether ‘business necessity’ justifies a practice which has a racially discriminatory result is not whether it is justified by routine business considerations but whether there is a compelling need for the employer to maintain that practice and whether the employer can prove there is no alternative to the challenged practice”); Pettway v American Cast Iron Pipe Co., 494 F. 2d 211, 244, n. 87 (CA5 1974) (“this doctrine of business necessity .... connotes an irresistible demand” (internal quotation marks omitted)); United States v Bethlehem Steel Corp., 446 F. 2d 652, 662 (CA2 1971) (an exclusionary practice “must not only directly foster safety and efficiency of a plant, but also be essential to those goals”); Robinson v Lorillard Corp., 444 F. 2d 791, 798 (CA4 1971) (“The test is whether there exists an overriding legitimate business purpose such that the practice is necessary to the safe and efficient operation of the business.”).
 See, e.g., Nash v Jacksonville, 837 F. 2d 1534 (CA11 1988), vacated, 490 U. S. 1103 (1989), opinion reinstated, 905 F. 2d 355 (CA11 1990); Vulcan Pioneers, Inc. v New Jersey Dept. of Civil Serv., 832 F. 2d 811 (CA3 (1987); Guardians Assn. of N. Y. City Police Dept. v Civil Serv. Comm’n, 630 F. 2d 79 (CA2 1980); Ensley Branch of NAACP v Seibels, 616 F. 2d 812 (CA5 1980); Fire-fighters Inst. for Racial Equality v St. Louis, 616 F. 2d 350 (CA8 1980); Boston Chapter, NAACP v Beecher, 504 F. 2d 1017 (CA1 1974).
 What was the “business necessity” for the tests New Haven used? How could one justify, e.g., the 60/40 written/oral ratio, see supra, at 4–5, 7–8, under that standard? Neither the Court nor the concurring opinions attempt to defend the ratio.
 Even in Title VII cases involving race-conscious (or gender-conscious) affirmative-action plans, the Court has never proposed a strong-basis-in-evidence standard. In Johnson v Transportation Agency, Santa Clara Cty., 480 U. S. 616 (1987), the Court simply examined the municipal employer’s action for reasonableness: “Given the obvious imbalance in the Skilled Craft category, and given the Agency’s commitment to eliminating such imbalances, it was plainly not unreasonable for the Agency .... to consider as one factor the sex of [applicants] in making its decision.” Id., at 637. See also Fire-fighters v Cleveland, 478 U. S. 501, 516 (1986) (“Title VII permits employers and unions voluntarily to make use of reasonable race-conscious affirmative action.”).
 Notably, prior decisions applying a strong-basis-in-evidence standard have not imposed a burden as heavy as the one the Court imposes today. In Croson, the Court found no strong basis in evidence because the City had offered “nothing approaching a prima facie case.” Richmond v J. A. Croson Co., 488 U. S. 469, 500 (1989). The Court did not suggest that anything beyond a prima facie case would have been required. In the context of race-based electoral districting, the Court has indicated that a “strong basis” exists when the “threshold conditions” for liability are present. Bush v Vera, 517 U. S. 952, 978 (1996) (plurality opinion).
 Infecting the Court’s entire analysis is its insistence that the City rejected the test results “in sole reliance upon race-based statistics.” Ante, at 24. See also ante, at 20, 27–28. But as the part of the story the Court leaves out, see supra, at 2–12, so plainly shows – the long history of rank discrimination against African-Americans in the fire fighting profession, the multiple flaws in New Haven’s test for promotions – “sole reliance” on statistics certainly is not descriptive of the CSB’s decision.
 The Court’s refusal to remand for further proceedings also deprives respondents of an opportunity to invoke 42 U. S. C. §2000e–12(b) as a shield to liability. Section 2000e–12(b) provides: “In any action or proceeding based on any alleged unlawful employment practice, no person shall be subject to any liability or punishment for or on account of (1) the commission by such person of an unlawful employment practice if he pleads and proves that the act or omission complained of was in good faith, in conformity with, and in reliance on any written interpretation or opinion of the [EEOC] .... Such a defence, if established, shall be a bar to the action or proceeding, notwithstanding that (A) after such act or omission, such interpretation or opinion is modified or rescinded or is determined by judicial authority to be invalid or of no legal effect ....” Specifically, given the chance, respondents might have called attention to the EEOC guidelines set out in 29 CFR §§1608.3 and 1608.4 (2008). The guidelines recognize that employers may “take affirmative action based on an analysis which reveals facts constituting actual or potential adverse impact.” §1608.3(a). If “affirmative action” is in order, so is the lesser step of discarding a dubious selection device.
 The lower courts focused on respondents’ “intent” rather than on whether respondents in fact had good cause to act. See 554 F. Supp. 2d 142, 157 (Conn. 2006). Ordinarily, a remand for fresh consideration would be in order. But the Court has seen fit to preclude further proceedings. I therefore explain why, if final adjudication by this Court is indeed appropriate, New Haven should be the prevailing party.
 This alone would have posed a substantial problem for New Haven in a disparate-impact suit, particularly in light of the disparate results the City’s scheme had produced in the past. See supra, at 7. Under the Uniform Guidelines on Employee Selection Procedures (Uniform Guidelines), employers must conduct “an investigation of suitable alternative selection procedures.” 29 CFR §1607.3(B). See also Officers for Justice v Civil Serv. Comm’n, 979 F. 2d 721, 728 (CA9 1992) (“before utilizing a procedure that has an adverse impact on minorities, the City has an obligation pursuant to the Uniform Guidelines to explore alternative procedures and to implement them if they have less adverse impact and are substantially equally valid”). It is no answer to “presume” that the two-decades-old 60/40 formula was adopted for a “rational reason” because it “was the result of a union-negotiated collective bargaining agreement.” Cf. ante, at 30. That the parties may have been “rational” says nothing about whether their agreed-upon selection process was consistent with business necessity. It is not at all unusual for agreements negotiated between employers and unions to run afoul of Title VII. See, e.g., Peters v Missouri-Pacific R. Co., 483 F. 2d 490, 497 (CA5 1973) (an employment practice “is not shielded [from the requirements of Title VII] by the facts that it is the product of collective bargaining and meets the standards of fair representation”).
 See also Nash, 837 F. 2d, at 1538 (“the examination did not test the one aspect of job performance that differentiated the job of firefighter engineer from fire lieutenant (combat): supervisory skills”); Fire-fighters Inst. for Racial Equality v St. Louis, 549 F. 2d 506, 512 (CA8 1977) (“there is no good pen and paper test for evaluating supervisory skills”); Boston Chapter, NAACP, 504 F. 2d, at 1023 (“[T]here is a difference between memorizing .... fire fighting terminology and being a good fire fighter. If the Boston Red Sox recruited players on the basis of their knowledge of baseball history and vocabulary, the team might acquire [players] who could not bat, pitch or catch.”).
 Cf. Gillespie v Wisconsin, 771 F. 2d 1035, 1043 (CA7 1985) (courts must evaluate “the degree to which the nature of the examination procedure approximates the job conditions”). In addition to “content validity,” the Uniform Guidelines discuss “construct validity” and “criterion validity” as means by which an employer might establish the reliability of a selection method. See 29 CFR §1607.14(B)–(D). Content validity, however, is the only type of validity addressed by the parties and “the only feasible type of validation in these circumstances.” Brief for Industrial-Organizational Psychologists as Amicus Curiae 7, n. 2 (hereinafter I-O Psychologists Brief).
 See also G. Thornton & D. Rupp, Assessment Centres in Human Resource Management 15 (2006) (“Assessment centres predict future success, do not cause adverse impact, and are seen as fair by participants.”); W. Cascio & H. Aguinis, Applied Psychology in Human Resource Management 372 (6th ed. 2005) (“research has demonstrated that adverse impact is less of a problem in an [assessment centre] as compared to an aptitude test”). Cf. Fire-fighters Inst. for Racial Equality, 549 F. 2d, at 513 (recommending assessment centres as an alternative to written exams).
 Finding the evidence concerning these alternatives insufficiently developed to “create a genuine issue of fact,” ante, at 32, the Court effectively confirms that an employer cannot prevail under its strong-basis-in-evidence standard unless the employer decisively proves a disparate-impact violation against itself. The Court’s specific arguments are unavailing. First, the Court suggests, changing the oral/written weighting may have violated Title VII’s prohibition on altering test scores. Ante, at 31. No one is arguing, however, that the results of the exams given should have been altered. Rather, the argument is that the City could have availed itself of a better option when it initially decided what selection process to use. Second, with respect to assessment centres, the Court identifies “statements to the CSB indicat[ing] that the Department could not have used [them] for the 2003 examinations.” Ante, at 31–32. The Court comes up with only a single statement on this subject – an offhand remark made by petitioner Ricci, who hardly qualifies as an expert in testing methods. See ante, at 14. Given the large number of municipalities that regularly use assessment centres, it is impossible to fathom why the City, with proper planning, could not have done so as well.
 The I-O Psychologists Brief identifies still other, more technical flaws in the exams that may well have precluded the City from prevailing in a disparate-impact suit. Notably, the exams were never shown to be suitably precise to allow strict rank ordering of candidates. A difference of one or two points on a multiple-choice exam should not be decisive of an applicant’s promotion chances if that difference bears little relationship to the applicant’s qualifications for the job. Relatedly, it appears that the line between a passing and failing score did not accurately differentiate between qualified and unqualified candidates. A number of fire-officer promotional exams have been invalidated on these bases. See, e.g., Guardians Assn., 630 F. 2d, at 105 (“When a cut-off score unrelated to job performance produces disparate racial results, Title VII is violated.”); Vulcan Pioneers, Inc. v New Jersey Dept. of Civil Serv., 625 F. Supp. 527, 538 (NJ 1985) (“[T]he tests here at issue are not appropriate for ranking candidates.”).
 Some of petitioners’ so-called facts find little support in the record, and many others can scarcely be deemed material. Petitioners allege, for example, that City officials prevented New Haven’s fire chief and assistant chief from sharing their views about the exams with the CSB. App. to Pet. for Cert. in No. 07–1428, p. 228a. None of the materials petitioners cite, however, “suggests” that this proposition is accurate. Cf. ante, at 5. In her deposition testimony, City official Karen Dubois-Walton specifically denied that she or her colleagues directed the chief and assistant chief not to appear. App. to Pet. for Cert. in No. 07–1428, p. 850a. Moreover, contrary to the insinuations of petitioners and Justice Alito, the statements made by City officials before the CSB did not emphasize allegations of cheating by test takers. Cf. ante, at 7–8. In her deposition, Dubois-Walton acknowledged sharing the cheating allegations not with the CSB, but with a different City commission. App. to Pet. for Cert. in No. 07–1428, p. 837a. Justice Alito also reports that the City’s attorney advised the mayor’s team that the way to convince the CSB not to certify was “to focus on something other than ‘a big discussion re: adverse impact’ law.” Ante, at 8 (quoting App. to Pet. for Cert. in No. 07–1428, p. 458a). This is a misleading abbreviation of the attorney’s advice. Focusing on the exams’ defects and on disparate-impact law is precisely what he recommended. See id., at 458a–459a.
 City officials, Justice Alito reports, sent Hornick newspaper accounts and other material about the exams prior to his testimony. Ante, at 8. Some of these materials, Justice Alito intimates, may have given Hornick an inaccurate portrait of the exams. But Hornick’s testimony before the CSB, viewed in full, indicates that Hornick had an accurate understanding of the exam process. Much of Hornick’s analysis focused on the 60/40 weighting of the written and oral exams, something that neither the Court nor the concurrences even attempt to defend. It is, moreover, entirely misleading to say that the City later hired union-proposed Hornick as a “rewar[d]” for his testimony. Cf. Ante, at 9.
 To be clear, the Board of Fire Commissioners on which Kimber served is an entity separate from the CSB. Kimber was not a member of the CSB. Kimber, Justice Alito states, requested a private meeting with the CSB. Ante, at 6. There is not a shred of evidence that a private meeting with Kimber or anyone else took place.
 Justice Alito points to evidence that the mayor had decided not to make promotions based on the exams even if the CSB voted to certify the results, going so far as to prepare a press release to that effect. Ante, at 9. If anything, this evidence reinforces the conclusion that the CSB – which made the non-certification decision – remained independent and above the political fray. The mayor and his staff needed a contingency plan precisely because they did not control the CSB.
 The District Court, Justice Alito writes, “all but conceded that a jury could find that the City’s asserted justification was pre-textual” by “admitt[ing] that ‘a jury could rationally infer that city officials worked behind the scenes to sabotage the promotional examinations because they knew that, were the exams certified, the Mayor would incur the wrath of [Rev. Boise] Kimber and other influential leaders of New Haven’s African-American community.’ ” Ante, at 3, 13 (quoting 554 F. Supp. 2d, at 162). The District Court drew the quoted passage from petitioners’ lower court brief, and used it in reference to a First Amendment claim not before this Court. In any event, it is not apparent why these alleged political manoeuvres suggest an intent to discriminate against petitioners. That City officials may have wanted to please political supporters is entirely consistent with their stated desire to avoid a disparate-impact violation. Cf. Ashcroft v Iqbal, 556 U. S. ___, ___ (2009) (slip op., at 18) (allegations that senior Government officials condoned the arrest and detention of thousands of Arab Muslim men following the September 11 attacks failed to establish even a “plausible inference” of unlawful discrimination sufficient to survive a motion to dismiss).
all rights reserved