Texas Department of Housing and Community Affairs v. Inclusive Communities Project, Inc., 576 U.S. 519 (2015)

NOTICE: This opinion is subject to formal revision before publication in the preliminary print of the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543, of any typographical or other formal errors, in order that corrections may be made before the preliminary print goes to press.

SUPREME COURT OF THE UNITED STATES

_________________

No. 13–1371

_________________

TEXAS DEPARTMENT OF HOUSING AND COMMU-NITY AFFAIRS, et al., PETITIONERS v. THE IN-CLUSIVE COMMUNITIES PROJECT, INC., et al.

on writ of certiorari to the united states court of appeals for the fifth circuit

[June 25, 2015]

Justice Kennedy delivered the opinion of the Court.

The underlying dispute in this case concerns where housing for low-income persons should be constructed in Dallas, Texas—that is, whether the housing should be built in the inner city or in the suburbs. This dispute comes to the Court on a disparate-impact theory of liability. In contrast to a disparate-treatment case, where a “plaintiff must establish that the defendant had a discriminatory intent or motive,” a plaintiff bringing a disparate-impact claim challenges practices that have a “disproportionately adverse effect on minorities” and are otherwise unjustified by a legitimate rationale. Ricci v. DeStefano, 557 U. S. 557, 577 (2009) (internal quotation marks omitted). The question presented for the Court’s determination is whether disparate-impact claims are cognizable under the Fair Housing Act (or FHA), 82Stat. 81, as amended, 42 U. S. C. §3601 et seq.

I

A

Before turning to the question presented, it is necessary to discuss a different federal statute that gives rise to this dispute. The Federal Government provides low-income housing tax credits that are distributed to developers through designated state agencies. 26 U. S. C. §42. Congress has directed States to develop plans identifying selection criteria for distributing the credits. §42(m)(1). Those plans must include certain criteria, such as public housing waiting lists, §42(m)(1)(C), as well as certain preferences, including that low-income housing units “contribut[e] to a concerted community revitalization plan” and be built in census tracts populated predominantly by low-income residents. §§42(m)(1)(B)(ii)(III), 42(d)(5)(ii)(I). Federal law thus favors the distribution of these tax credits for the development of housing units in low-income areas.

In the State of Texas these federal credits are distrib-uted by the Texas Department of Housing and Community Affairs (Department). Under Texas law, a developer’s application for the tax credits is scored under a point system that gives priority to statutory criteria, such as the financial feasibility of the development project and the income level of tenants. Tex. Govt. Code Ann. §§2306.6710(a)–(b) (West 2008). The Texas Attorney General has interpreted state law to permit the consideration of additional criteria, such as whether the housing units will be built in a neighborhood with good schools. Those criteria cannot be awarded more points than statutorily mandated criteria. Tex. Op. Atty. Gen. No. GA–0208, pp. 2–6 (2004), 2004 WL 1434796, *4–*6.

The Inclusive Communities Project, Inc. (ICP), is a Texas-based nonprofit corporation that assists low-income families in obtaining affordable housing. In 2008, the ICP brought this suit against the Department and its officers in the United States District Court for the Northern District of Texas. As relevant here, it brought a disparate-impact claim under §§804(a) and 805(a) of the FHA. The ICP alleged the Department has caused continued segregated housing patterns by its disproportionate allocation of the tax credits, granting too many credits for housing in predominantly black inner-city areas and too few in predominantly white suburban neighborhoods. The ICP contended that the Department must modify its selection criteria in order to encourage the construction of low-income housing in suburban communities.

The District Court concluded that the ICP had established a prima facie case of disparate impact. It relied on two pieces of statistical evidence. First, it found “from 1999–2008, [the Department] approved tax credits for 49.7% of proposed non-elderly units in 0% to 9.9% Caucasian areas, but only approved 37.4% of proposed non-elderly units in 90% to 100% Caucasian areas.” 749 F. Supp. 2d 486, 499 (ND Tex. 2010) (footnote omitted). Second, it found “92.29% of [low-income housing tax credit] units in the city of Dallas were located in census tracts with less than 50% Caucasian residents.” Ibid.

The District Court then placed the burden on the Department to rebut the ICP’s prima facie showing of disparate impact. 860 F. Supp. 2d 312, 322–323 (2012). After assuming the Department’s proffered interests were legitimate, id., at 326, the District Court held that a defendant—here the Department—must prove “that there are no other less discriminatory alternatives to advancing their proffered interests,” ibid. Because, in its view, the Department “failed to meet [its] burden of proving that there are no less discriminatory alternatives,” the District Court ruled for the ICP. Id., at 331.

The District Court’s remedial order required the addition of new selection criteria for the tax credits. For instance, it awarded points for units built in neighborhoods with good schools and disqualified sites that are located adjacent to or near hazardous conditions, such as high crime areas or landfills. See 2012 WL 3201401 (Aug. 7, 2012). The remedial order contained no explicit racial targets or quotas.

While the Department’s appeal was pending, the Secretary of Housing and Urban Development (HUD) issued a regulation interpreting the FHA to encompass disparate-impact liability. See Implementation of the Fair Housing Act’s Discriminatory Effects Standard, 78 Fed. Reg. 11460 (2013). The regulation also established a burden-shifting framework for adjudicating disparate-impact claims. Under the regulation, a plaintiff first must make a prima facie showing of disparate impact. That is, the plaintiff “has the burden of proving that a challenged practice caused or predictably will cause a discriminatory effect.” 24 CFR §100.500(c)(1) (2014). If a statistical discrepancy is caused by factors other than the defendant’s policy, a plaintiff cannot establish a prima facie case, and there is no liability. After a plaintiff does establish a prima facie showing of disparate impact, the burden shifts to the defendant to “prov[e] that the challenged practice is necessary to achieve one or more substantial, legitimate, nondiscriminatory interests.” §100.500(c)(2). HUD has clarified that this step of the analysis “is analogous to the Title VII requirement that an employer’s interest in an employment practice with a disparate impact be job related.” 78 Fed. Reg. 11470. Once a defendant has satisfied its burden at step two, a plaintiff may “prevail upon proving that the substantial, legitimate, nondiscriminatory interests supporting the challenged practice could be served by another practice that has a less discriminatory effect.” §100.500(c)(3).

The Court of Appeals for the Fifth Circuit held, consistent with its precedent, that disparate-impact claims are cognizable under the FHA. 747 F. 3d 275, 280 (2014). On the merits, however, the Court of Appeals reversed and remanded. Relying on HUD’s regulation, the Court of Appeals held that it was improper for the District Court to have placed the burden on the Department to prove there were no less discriminatory alternatives for allocating low-income housing tax credits. Id., at 282–283. In a concurring opinion, Judge Jones stated that on remand the District Court should reexamine whether the ICP had made out a prima facie case of disparate impact. She suggested the District Court incorrectly relied on bare statistical evidence without engaging in any analysis about causation. She further observed that, if the fed-eral law providing for the distribution of low-income housing tax credits ties the Department’s hands to such an extent that it lacks a meaningful choice, then there is no disparate-impact liability. See id., at 283–284 (specially concurring opinion).

The Department filed a petition for a writ of certiorari on the question whether disparate-impact claims are cognizable under the FHA. The question was one of first impression, see Huntington v. Huntington Branch, NAACP, 488 U. S. 15 (1988) (per curiam), and certiorari followed, 573 U. S. ___ (2014). It is now appropriate to provide a brief history of the FHA’s enactment and its later amendment.

B

De jure residential segregation by race was declared unconstitutional almost a century ago, Buchanan v. Warley, 245 U. S. 60 (1917) , but its vestiges remain today, intertwined with the country’s economic and social life. Some segregated housing patterns can be traced to conditions that arose in the mid-20th century. Rapid urbanization, concomitant with the rise of suburban developments accessible by car, led many white families to leave the inner cities. This often left minority families concentrated in the center of the Nation’s cities. During this time, various practices were followed, sometimes with governmental support, to encourage and maintain the separation of the races: Racially restrictive covenants prevented the conveyance of property to minorities, see Shelley v. Krae-mer, 334 U. S. 1 (1948) ; steering by real-estate agents led potential buyers to consider homes in racially homogenous areas; and discriminatory lending practices, often referred to as redlining, precluded minority families from purchasing homes in affluent areas. See, e.g., M. Klarman, Unfinished Business: Racial Equality in American History 140–141 (2007); Brief for Housing Scholars as Amici Curiae 22–23. By the 1960’s, these policies, practices, and prejudices had created many predominantly black inner cities surrounded by mostly white suburbs. See K. Clark, Dark Ghetto: Dilemmas of Social Power 11, 21–26 (1965).

The mid-1960’s was a period of considerable social unrest; and, in response, President Lyndon Johnson established the National Advisory Commission on Civil Disorders, commonly known as the Kerner Commission. Exec. Order No. 11365, 3 CFR 674 (1966–1970 Comp.). After extensive factfinding the Commission identified residential segregation and unequal housing and economic conditions in the inner cities as significant, underlying causes of the social unrest. See Report of the National Advisory Commission on Civil Disorders 91 (1968) (Kerner Commission Report). The Commission found that “[n]early two-thirds of all nonwhite families living in the central cities today live in neighborhoods marked by substandard housing and general urban blight.” Id., at 13. The Commission further found that both open and covert racial discrimination prevented black families from obtaining better housing and moving to integrated communities. Ibid. The Commission concluded that “[o]ur Nation is moving toward two societies, one black, one white—separate and unequal.” Id., at 1. To reverse “[t]his deepening racial division,” ibid., it recommended enactment of “a comprehensive and enforceable open-occupancy law making it an offense to discriminate in the sale or rental of any housing . . . on the basis of race, creed, color, or national origin.” Id., at 263.

In April 1968, Dr. Martin Luther King, Jr., was assassinated in Memphis, Tennessee, and the Nation faced a new urgency to resolve the social unrest in the inner cities. Congress responded by adopting the Kerner Commission’s recommendation and passing the Fair Housing Act. The statute addressed the denial of housing opportunities on the basis of “race, color, religion, or national origin.” Civil Rights Act of 1968, §804, 82Stat. 83. Then, in 1988, Congress amended the FHA. Among other provisions, it created certain exemptions from liability and added “familial status” as a protected characteristic. See Fair Housing Amendments Act of 1988, 102Stat. 1619.

II

The issue here is whether, under a proper interpretation of the FHA, housing decisions with a disparate impact are prohibited. Before turning to the FHA, however, it is necessary to consider two other antidiscrimination statutes that preceded it.

The first relevant statute is §703(a) of Title VII of the Civil Rights Act of 1964, 78Stat. 255. The Court addressed the concept of disparate impact under this statute in Griggs v. Duke Power Co., 401 U. S. 424 (1971) . There, the employer had a policy requiring its manual laborers to possess a high school diploma and to obtain satisfactory scores on two intelligence tests. The Court of Appeals held the employer had not adopted these job requirements for a racially discriminatory purpose, and the plaintiffs did not challenge that holding in this Court. Instead, the plaintiffs argued §703(a)(2) covers the discriminatory effect of a practice as well as the motivation behind the practice. Section 703(a), as amended, provides as follows:

“It shall be an unlawful employer practice for an employer—

“(1) 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, color, religion, sex, or national origin; or

“(2) to limit, segregate, or classify his employees or applicants for employment 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, color, religion, sex, or national origin.” 42 U. S. C. §2000e–2(a).

The Court did not quote or cite the full statute, but rather relied solely on §703(a)(2). Griggs, 401 U. S., at 426, n. 1.

In interpreting §703(a)(2), the Court reasoned that disparate-impact liability furthered the purpose and design of the statute. The Court explained that, in §703(a)(2), Congress “proscribe[d] not only overt discrimination but also practices that are fair in form, but discriminatory in operation.” Id., at 431. For that reason, as the Court noted, “Congress directed the thrust of [§703(a)(2)] to the consequences of employment practices, not simply the motivation.” Id., at 432. In light of the statute’s goal of achieving “equality of employment opportunities and remov[ing] barriers that have operated in the past” to favor some races over others, the Court held §703(a)(2) of Title VII must be interpreted to allow disparate-impact claims. Id., at 429–430.

The Court put important limits on its holding: namely, not all employment practices causing a disparate impact impose liability under §703(a)(2). In this respect, the Court held that “business necessity” constitutes a defense to disparate-impact claims. Id., at 431. This rule provides, for example, that in a disparate-impact case, §703(a)(2) does not prohibit hiring criteria with a “manifest relationship” to job performance. Id., at 432; see also Ricci, 557 U. S., at 587–589 (emphasizing the importance of the business necessity defense to disparate-impact liability). On the facts before it, the Court in Griggs found a violation of Title VII because the employer could not establish that high school diplomas and general intelligence tests were related to the job performance of its manual laborers. See 401 U. S., at 431–432.

The second relevant statute that bears on the proper interpretation of the FHA is the Age Discrimination in Employment Act of 1967 (ADEA), 81Stat. 602 et seq., as amended. Section 4(a) of the ADEA provides:

“It shall be unlawful for an employer—

“(1) to fail or refuse to hire or to discharge any individual or otherwise discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s age;

“(2) 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 age; or

“(3) to reduce the wage rate of any employee in order to comply with this chapter.” 29 U. S. C. §623(a).

The Court first addressed whether this provision allows disparate-impact claims in Smith v. City of Jackson, 544 U. S. 228 (2005) . There, a group of older employees challenged their employer’s decision to give proportionately greater raises to employees with less than five years of experience.

Explaining that Griggs “represented the better reading of [Title VII’s] statutory text,” 544 U. S., at 235, a plurality of the Court concluded that the same reasoning pertained to §4(a)(2) of the ADEA. The Smith plurality emphasized that both §703(a)(2) of Title VII and §4(a)(2) of the ADEA contain language “prohibit[ing] such actions that ‘deprive any individual of employment opportunities or otherwise adversely affect his status as an employee, because of such individual’s’ race or age.” 544 U. S., at 235. As the plurality observed, the text of these provisions “focuses on the effects of the action on the employee rather than the motivation for the action of the employer” and therefore compels recognition of disparate-impact liability. Id., at 236. In a separate opinion, Justice Scalia found the ADEA’s text ambiguous and thus deferred under Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U. S. 837 (1984) , to an Equal Employment Opportunity Commission regulation interpreting the ADEA to imposedisparate-impact liability, see 544 U. S., at 243–247 (opinion concurring in part and concurring in judgment).

Together, Griggs holds and the plurality in Smith instructs that antidiscrimination laws must be construed to encompass disparate-impact claims when their text refers to the consequences of actions and not just to the mindset of actors, and where that interpretation is consistent with statutory purpose. These cases also teach that disparate-impact liability must be limited so employers and other regulated entities are able to make the practical business choices and profit-related decisions that sustain a vibrant and dynamic free-enterprise system. And before rejecting a business justification—or, in the case of a governmental entity, an analogous public interest—a court must determine that a plaintiff has shown that there is “an available alternative . . . practice that has less disparate impact and serves the [entity’s] legitimate needs.” Ricci, supra, at 578. The cases interpreting Title VII and the ADEA provide essential background and instruction in the case now before the Court.

Turning to the FHA, the ICP relies on two provisions. Section 804(a) provides that it shall be unlawful:

“To refuse to sell or rent after the making of a bona fide offer, or to refuse to negotiate for the sale orrental of, or otherwise make unavailable or deny, a dwelling to any person because of race, color, religion, sex, familial status, or national origin.” 42 U. S. C. §3604(a).

Here, the phrase “otherwise make unavailable” is of central importance to the analysis that follows.

Section 805(a), in turn, provides:

“It shall be unlawful for any person or other entity whose business includes engaging in real estate-related transactions to discriminate against any person in making available such a transaction, or in the terms or conditions of such a transaction, because of race, color, religion, sex, handicap, familial status, or national origin.” §3605(a).

Applied here, the logic of Griggs and Smith provides strong support for the conclusion that the FHA encompasses disparate-impact claims. Congress’ use of the phrase “otherwise make unavailable” refers to the consequences of an action rather than the actor’s intent. See United States v. Giles, 300 U. S. 41, 48 (1937) (explaining that the “word ‘make’ has many meanings, among them ‘[t]o cause to exist, appear or occur’ ” (quoting Webster’s New International Dictionary 1485 (2d ed. 1934))). This results-oriented language counsels in favor of recognizing disparate-impact liability. See Smith, supra, at 236. The Court has construed statutory language similar to §805(a) to include disparate-impact liability. See, e.g., Board of Ed. of City School Dist. of New York v. Harris, 444 U. S. 130 –141 (1979) (holding the term “discriminat[e]” encompassed disparate-impact liability in the context of a statute’s text, history, purpose, and structure).

A comparison to the antidiscrimination statutes examined in Griggs and Smith is useful. Title VII’s and the ADEA’s “otherwise adversely affect” language is equivalent in function and purpose to the FHA’s “otherwise make unavailable” language. In these three statutes the operative text looks to results. The relevant statutory phrases, moreover, play an identical role in the structure common to all three statutes: Located at the end of lengthy sentences that begin with prohibitions on disparate treatment, they serve as catchall phrases looking to consequences, not intent. And all three statutes use the word “otherwise” to introduce the results-oriented phrase. “Otherwise” means “in a different way or manner,” thus signaling a shift in emphasis from an actor’s intent to the consequences of his actions. Webster’s Third New International Dictionary 1598 (1971). This similarity in text and structure is all the more compelling given that Congress passed the FHA in 1968—only four years after passing Title VII and only four months after enacting the ADEA.

It is true that Congress did not reiterate Title VII’s exact language in the FHA, but that is because to do so would have made the relevant sentence awkward and unclear. A provision making it unlawful to “refuse to sell[,] . . . or otherwise [adversely affect], a dwelling to any person” because of a protected trait would be grammatically obtuse, difficult to interpret, and far more expansive in scope than Congress likely intended. Congress thus chose words that serve the same purpose and bear the same basic meaning but are consistent with the structure and objectives of the FHA.

Emphasizing that the FHA uses the phrase “because of race,” the Department argues this language forecloses disparate-impact liability since “[a]n action is not taken ‘because of race’ unless race is a reason for the action.” Brief for Petitioners 26. Griggs and Smith, however, dispose of this argument. Both Title VII and the ADEA contain identical “because of” language, see 42 U. S. C. §2000e–2(a)(2); 29 U. S. C. §623(a)(2), and the Courtnonetheless held those statutes impose disparate-impact liability.

In addition, it is of crucial importance that the existence of disparate-impact liability is supported by amendments to the FHA that Congress enacted in 1988. By that time, all nine Courts of Appeals to have addressed the question had concluded the Fair Housing Act encompassed disparate-impact claims. See Huntington Branch, NAACP v.Huntington, 844 F. 2d 926, 935–936 (CA2 1988); Resident Advisory Bd. v. Rizzo, 564 F. 2d 126, 146 (CA3 1977); Smith v. Clarkton, 682 F. 2d 1055, 1065 (CA4 1982); Hanson v. Veterans Administration, 800 F. 2d 1381, 1386 (CA5 1986); Arthur v. Toledo, 782 F. 2d 565, 574–575 (CA6 1986); Metropolitan Housing Development Corp. v. Arlington Heights, 558 F. 2d 1283, 1290 (CA7 1977); United States v. Black Jack, 508 F. 2d 1179, 1184–1185 (CA8 1974); Halet v. Wend Investment Co., 672 F. 2d 1305, 1311 (CA9 1982); United States v. Marengo Cty. Comm’n, 731 F. 2d 1546, 1559, n. 20 (CA11 1984).

When it amended the FHA, Congress was aware of this unanimous precedent. And with that understanding, it made a considered judgment to retain the relevant statutory text. See H. R. Rep. No. 100–711, p. 21, n. 52 (1988) (H. R. Rep.) (discussing suits premised on disparate-impact claims and related judicial precedent); 134 Cong. Rec. 23711 (1988) (statement of Sen. Kennedy) (noting unanimity of Federal Courts of Appeals concerning disparate impact); Fair Housing Amendments Act of 1987: Hearings on S. 558 before the Subcommittee on the Constitution of the Senate Committee on the Judiciary, 100th Cong., 1st Sess., 529 (1987) (testimony of Professor Robert Schwemm) (describing consensus judicial view that the FHA imposed disparate-impact liability). Indeed, Congress rejected a proposed amendment that would have eliminated disparate-impact liability for certain zoning decisions. See H. R. Rep., at 89–93.

Against this background understanding in the legal and regulatory system, Congress’ decision in 1988 to amend the FHA while still adhering to the operative language in §§804(a) and 805(a) is convincing support for the conclusion that Congress accepted and ratified the unanimous holdings of the Courts of Appeals finding disparate-impact liability. “If a word or phrase has been . . . given a uniform interpretation by inferior courts . . . , a later version of that act perpetuating the wording is presumed to carry forward that interpretation.” A. Scalia & B. Garner, Reading Law: The Interpretation of Legal Texts 322 (2012); see also Forest Grove School Dist. v. T. A., 557 U. S. 230, 244, n. 11 (2009) (“When Congress amended [the Act] without altering the text of [the relevant provision], it implicitly adopted [this Court’s] construction of the statute”); Manhattan Properties, Inc. v. Irving Trust Co., 291 U. S. 320, 336 (1934) (explaining, where the Courts of Appeals had reached a consensus interpretation of the Bankruptcy Act and Congress had amended the Act without changing the relevant provision, “[t]his is persua-sive that the construction adopted by the [lower federal] courts has been acceptable to the legislative arm of the government”).

Further and convincing confirmation of Congress’ understanding that disparate-impact liability exists under the FHA is revealed by the substance of the 1988 amendments. The amendments included three exemptions from liability that assume the existence of disparate-impact claims. The most logical conclusion is that the three amendments were deemed necessary because Congress presupposed disparate impact under the FHA as it had been enacted in 1968.

The relevant 1988 amendments were as follows. First, Congress added a clarifying provision: “Nothing in [the FHA] prohibits a person engaged in the business of furnishing appraisals of real property to take into consideration factors other than race, color, religion, national origin, sex, handicap, or familial status.” 42 U. S. C. §3605(c). Second, Congress provided: “Nothing in [the FHA] prohibits conduct against a person because such person has been convicted by any court of competent jurisdiction of the illegal manufacture or distribution of a controlled substance.” §3607(b)(4). And finally, Congress specified: “Nothing in [the FHA] limits the applicability of any reasonable . . . restrictions regarding the maximum number of occupants permitted to occupy a dwelling.” §3607(b)(1).

The exemptions embodied in these amendments would be superfluous if Congress had assumed that disparate-impact liability did not exist under the FHA. See Gustafson v. Alloyd Co., 513 U. S. 561, 574 (1995) (“[T]he Court will avoid a reading which renders some words altogether redundant”). Indeed, none of these amendments would make sense if the FHA encompassed only disparate-treatment claims. If that were the sole ground for liability, the amendments merely restate black-letter law. If an actor makes a decision based on reasons other than a protected category, there is no disparate-treatment liability. See, e.g., Texas Dept. of Community Affairs v. Burdine, 450 U. S. 248, 254 (1981) . But the amendments do constrain disparate-impact liability. For instance, certain criminal convictions are correlated with sex and race. See, e.g., Kimbrough v. United States, 552 U. S. 85, 98 (2007) (discussing the racial disparity in convictions for crack cocaine offenses). By adding an exemption from liability for exclusionary practices aimed at individuals with drug convictions, Congress ensured disparate-impact liability would not lie if a landlord excluded tenants with such convictions. The same is true of the provision allowing for reasonable restrictions on occupancy. And the exemption from liability for real-estate appraisers is in the same section as §805(a)’s prohibition of discriminatory practices in real-estate transactions, thus indicating Congress’ recognition that disparate-impact liability arose under §805(a). In short, the 1988 amendments signal that Congress ratified disparate-impact liability.

A comparison to Smith’s discussion of the ADEA further demonstrates why the Department’s interpretation would render the 1988 amendments superfluous. Under the ADEA’s reasonable-factor-other-than-age (RFOA) provision, an employer is permitted to take an otherwise prohibited action where “the differentiation is based on reasonable factors other than age.” 29 U. S. C. §623(f )(1). In other words, if an employer makes a decision based on a reasonable factor other than age, it cannot be said to have made a decision on the basis of an employee’s age. According to the Smith plurality, the RFOA provision “plays its principal role” “in cases involving disparate-impact claims” “by precluding liability if the adverse impact was attributable to a nonage factor that was ‘reasonable.’ ” 544 U. S., at 239. The plurality thus reasoned that the RFOA provision would be “simply unnecessary to avoid liability under the ADEA” if liability were limited to disparate-treatment claims. Id., at 238.

A similar logic applies here. If a real-estate appraiser took into account a neighborhood’s schools, one could not say the appraiser acted because of race. And by embedding 42 U. S. C. §3605(c)’s exemption in the statutory text, Congress ensured that disparate-impact liability would not be allowed either. Indeed, the inference of disparate-impact liability is even stronger here than it was in Smith. As originally enacted, the ADEA included the RFOA provision, see §4(f)(1), 81Stat. 603, whereas here Congress added the relevant exemptions in the 1988 amendments against the backdrop of the uniform view of the Courts of Appeals that the FHA imposed disparate-impact liability.

Recognition of disparate-impact claims is consistent with the FHA’s central purpose. See Smith, supra, at 235 (plurality opinion); Griggs, 401 U. S., at 432. The FHA, like Title VII and the ADEA, was enacted to eradicate discriminatory practices within a sector of our Nation’s economy. See 42 U. S. C. §3601 (“It is the policy of the United States to provide, within constitutional limitations, for fair housing throughout the United States”); H. R. Rep., at 15 (explaining the FHA “provides a clear national policy against discrimination in housing”).

These unlawful practices include zoning laws and other housing restrictions that function unfairly to exclude minorities from certain neighborhoods without any sufficient justification. Suits targeting such practices reside at the heartland of disparate-impact liability. See, e.g., Huntington, 488 U. S., at 16–18 (invalidating zoning law preventing construction of multifamily rental units); Black Jack, 508 F. 2d, at 1182–1188 (invalidating ordinance prohibiting construction of new multifamily dwellings); Greater New Orleans Fair Housing Action Center v. St. Bernard Parish, 641 F. Supp. 2d 563, 569, 577–578 (ED La. 2009) (invalidating post-Hurricane Katrina ordinance restricting the rental of housing units to only “ ‘blood relative[s]’ ” in an area of the city that was 88.3% white and 7.6% black); see also Tr. of Oral Arg. 52–53 (discussing these cases). The availability of disparate-impact liability, furthermore, has allowed private developers to vindicate the FHA’s objectives and to protect their prop-erty rights by stopping municipalities from enforcing arbi-trary and, in practice, discriminatory ordinances barring the construction of certain types of housing units. See, e.g., Huntington, supra, at 18. Recognition of disparate-impact liability under the FHA also plays a role in uncovering discriminatory intent: It permits plaintiffs to counteract unconscious prejudices and disguised animus that escape easy classification as disparate treatment. In this way disparate-impact liability may prevent segregated housing patterns that might otherwise result from covert and illicit stereotyping.

But disparate-impact liability has always been properly limited in key respects that avoid the serious constitutional questions that might arise under the FHA, for instance,if such liability were imposed based solely on a showing of a statistical disparity. Disparate-impact liability mandates the “removal of artificial, arbitrary, and unnecessary barriers,” not the displacement of valid governmental policies. Griggs, supra, at 431. The FHA is not an instrument to force housing authorities to reorder their priorities. Rather, the FHA aims to ensure that those priorities can be achieved without arbitrarily creating discriminatory effects or perpetuating segregation.

Unlike the heartland of disparate-impact suits targeting artificial barriers to housing, the underlying dispute in this case involves a novel theory of liability. See Seicsh-naydre, Is Disparate Impact Having Any Impact? An Appellate Analysis of Forty Years of Disparate Impact Claims Under the Fair Housing Act, 63 Am. U. L. Rev. 357, 360–363 (2013) (noting the rarity of this type of claim). This case, on remand, may be seen simply as an attempt to second-guess which of two reasonable approaches a housing authority should follow in the sound exercise of its discretion in allocating tax credits for low-income housing.

An important and appropriate means of ensuring that disparate-impact liability is properly limited is to give housing authorities and private developers leeway to state and explain the valid interest served by their policies. This step of the analysis is analogous to the business necessity standard under Title VII and provides a defense against disparate-impact liability. See 78 Fed. Reg. 11470 (explaining that HUD did not use the phrase “business necessity” because that “phrase may not be easily understood to cover the full scope of practices covered by the Fair Housing Act, which applies to individuals, busi-nesses, nonprofit organizations, and public entities”). As the Court explained in Ricci, an entity “could be liable for disparate-impact discrimination only if the [challenged practices] were not job related and consistent with business necessity.” 557 U. S., at 587. Just as an employer may maintain a workplace requirement that causes a disparate impact if that requirement is a “reasonable measure[ment] of job performance,” Griggs, supra, at 436, so too must housing authorities and private developers be allowed to maintain a policy if they can prove it is necessary to achieve a valid interest. To be sure, the Title VII framework may not transfer exactly to the fair-housing context, but the comparison suffices for present purposes.

It would be paradoxical to construe the FHA to impose onerous costs on actors who encourage revitalizing dilapidated housing in our Nation’s cities merely because some other priority might seem preferable. Entrepreneurs must be given latitude to consider market factors. Zoning officials, moreover, must often make decisions based on a mix of factors, both objective (such as cost and traffic patterns) and, at least to some extent, subjective (such as preserving historic architecture). These factors contribute to a community’s quality of life and are legitimate concerns for housing authorities. The FHA does not decree a particular vision of urban development; and it does not put housing authorities and private developers in a double bind of liability, subject to suit whether they choose to rejuvenate a city core or to promote new low-income housing in suburban communities. As HUD itself recognized in its recent rulemaking, disparate-impact liability “does not mandate that affordable housing be located in neighborhoods with any particular characteristic.” 78 Fed. Reg. 11476.

In a similar vein, a disparate-impact claim that relies on a statistical disparity must fail if the plaintiff cannot point to a defendant’s policy or policies causing that disparity. A robust causality requirement ensures that “[r]acial imbalance . . . does not, without more, establish a prima facie case of disparate impact” and thus protects defendants from being held liable for racial disparities they did not create. Wards Cove Packing Co. v. Atonio, 490 U. S. 642, 653 (1989) , superseded by statute on other grounds, 42 U. S. C. §2000e–2(k). Without adequate safeguards at the prima facie stage, disparate-impact liability might cause race to be used and considered in a pervasive way and “would almost inexorably lead” governmental or private entities to use “numerical quotas,” and serious constitutional questions then could arise. 490 U. S., at 653.

The litigation at issue here provides an example. From the standpoint of determining advantage or disadvantage to racial minorities, it seems difficult to say as a general matter that a decision to build low-income housing in a blighted inner-city neighborhood instead of a suburb is discriminatory, or vice versa. If those sorts of judgments are subject to challenge without adequate safeguards, then there is a danger that potential defendants may adopt racial quotas—a circumstance that itself raises serious constitutional concerns.

Courts must therefore examine with care whether a plaintiff has made out a prima facie case of disparate impact and prompt resolution of these cases is important. A plaintiff who fails to allege facts at the pleading stage or produce statistical evidence demonstrating a causal connection cannot make out a prima facie case of disparate impact. For instance, a plaintiff challenging the decision of a private developer to construct a new building in one location rather than another will not easily be able to show this is a policy causing a disparate impact because such a one-time decision may not be a policy at all. It may also be difficult to establish causation because of the multiple factors that go into investment decisions about where to construct or renovate housing units. And as Judge Jones observed below, if the ICP cannot show a causal connection between the Department’s policy and a disparate impact—for instance, because federal law substantially limits the Department’s discretion—that should resultin dismissal of this case. 747 F. 3d, at 283–284 (specially concurring opinion).

The FHA imposes a command with respect to disparate-impact liability. Here, that command goes to a state entity. In other cases, the command will go to a private person or entity. Governmental or private policies are not contrary to the disparate-impact requirement unless they are “artificial, arbitrary, and unnecessary barriers.” Griggs, 401 U. S., at 431. Difficult questions might arise if disparate-impact liability under the FHA caused race to be used and considered in a pervasive and explicit manner to justify governmental or private actions that, in fact, tend to perpetuate race-based considerations rather than move beyond them. Courts should avoid interpreting disparate-impact liability to be so expansive as to inject racial considerations into every housing decision.

The limitations on disparate-impact liability discussed here are also necessary to protect potential defendants against abusive disparate-impact claims. If the specter of disparate-impact litigation causes private developers to no longer construct or renovate housing units for low-income individuals, then the FHA would have undermined its own purpose as well as the free-market system. And as to governmental entities, they must not be prevented from achieving legitimate objectives, such as ensuring compliance with health and safety codes. The Department’s amici, in addition to the well-stated principal dissenting opinion in this case, see post, at 1–2, 29–30 (opinion of Alito, J.), call attention to the decision by the Court of Appeals for the Eighth Circuit in Gallagher v. Magner, 619 F. 3d 823 (2010). Although the Court is reluctant to approve or disapprove a case that is not pending, it should be noted that Magner was decided without the cautionary standards announced in this opinion and, in all events, the case was settled by the parties before an ultimate determination of disparate-impact liability.

Were standards for proceeding with disparate-impact suits not to incorporate at least the safeguards discussed here, then disparate-impact liability might displace valid governmental and private priorities, rather than solely “remov[ing] . . . artificial, arbitrary, and unnecessary barriers.” Griggs, 401 U. S., at 431. And that, in turn, would set our Nation back in its quest to reduce the salience of race in our social and economic system.

It must be noted further that, even when courts do find liability under a disparate-impact theory, their remedial orders must be consistent with the Constitution. Remedial orders in disparate-impact cases should concentrate onthe elimination of the offending practice that “arbitrar[ily] . . . operate[s] invidiously to discriminate on the basis of rac[e].” Ibid. If additional measures are adopted, courts should strive to design them to eliminate racial disparities through race-neutral means. See Richmond v. J. A. Croson Co., 488 U. S. 469, 510 (1989) (plurality opinion) (“[T]he city has at its disposal a whole array of race-neutral devices to increase the accessibility of city contracting opportunities to small entrepreneurs of all races”). Remedial orders that impose racial targets or quotas might raise more difficult constitutional questions.

While the automatic or pervasive injection of race into public and private transactions covered by the FHA has special dangers, it is also true that race may be considered in certain circumstances and in a proper fashion. Cf. Parents Involved in Community Schools v. Seattle School Dist. No. 1, 551 U. S. 701, 789 (2007) (Kennedy, J., concurring in part and concurring in judgment) (“School boards may pursue the goal of bringing together students of diverse backgrounds and races through other means, including strategic site selection of new schools; [and] drawing attendance zones with general recognition of the demographics of neighborhoods”). Just as this Court has not “question[ed] an employer’s affirmative efforts to ensure that all groups have a fair opportunity to apply for promotions and to participate in the [promotion] process,” Ricci, 557 U. S., at 585, it likewise does not impugn housing authorities’ race-neutral efforts to encourage revitalization of communities that have long suffered the harsh consequences of segregated housing patterns. When setting their larger goals, local housing authorities may choose to foster diversity and combat racial isolation with race-neutral tools, and mere awareness of race in attempting to solve the problems facing inner cities does not doom that endeavor at the outset.

The Court holds that disparate-impact claims are cognizable under the Fair Housing Act upon considering its results-oriented language, the Court’s interpretation of similar language in Title VII and the ADEA, Congress’ ratification of disparate-impact claims in 1988 against the backdrop of the unanimous view of nine Courts of Appeals, and the statutory purpose.

III

In light of the longstanding judicial interpretation of the FHA to encompass disparate-impact claims and congressional reaffirmation of that result, residents and policymakers have come to rely on the availability of disparate-impact claims. See Brief for Massachusetts et al. as Amici Curiae 2 (“Without disparate impact claims, States and others will be left with fewer crucial tools to combat the kinds of systemic discrimination that the FHA was intended to address”). Indeed, many of our Nation’s largest cities—entities that are potential defendants in disparate-impact suits—have submitted an amicus brief in this case supporting disparate-impact liability under the FHA. See Brief for City of San Francisco et al. as Amici Curiae 3–6. The existence of disparate-impact liability in the substantial majority of the Courts of Appeals for the last several decades “has not given rise to . . . dire consequences.” Hosanna-Tabor Evangelical Lutheran Church and School v. EEOC, 565 U. S. ___, ___ (2012) (slip op., at 21).

Much progress remains to be made in our Nation’s continuing struggle against racial isolation. In striving to achieve our “historic commitment to creating an integrated society,” Parents Involved, supra, at 797 (Kennedy, J., concurring in part and concurring in judgment), we must remain wary of policies that reduce homeowners to nothing more than their race. But since the passage of the Fair Housing Act in 1968 and against the backdrop of disparate-impact liability in nearly every jurisdiction, many cities have become more diverse. The FHA must play an important part in avoiding the Kerner Commission’s grim prophecy that “[o]ur Nation is moving toward two societies, one black, one white—separate and un-equal.” Kerner Commission Report 1. The Court acknowl-edges the Fair Housing Act’s continuing role in moving the Nation toward a more integrated society.

The judgment of the Court of Appeals for the Fifth Circuit is affirmed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

SUPREME COURT OF THE UNITED STATES

_________________

No. 13–1371

_________________

TEXAS DEPARTMENT OF HOUSING AND COMMU-NITY AFFAIRS, et al., PETITIONERS v. THE IN-CLUSIVE COMMUNITIES PROJECT, INC., et al.

on writ of certiorari to the united states court of appeals for the fifth circuit

[June 25, 2015]

Justice Thomas, dissenting.

I join Justice Alito’s dissent in full. I write separately to point out that the foundation on which the Court builds its latest disparate-impact regime—Griggs v. Duke Power Co., 401 U. S. 424 (1971) —is made of sand. That decision, which concluded that Title VII of the Civil Rights Act of 1964 authorizes plaintiffs to bring disparate-impact claims, id., at 429–431, represents the triumph of an agency’s preferences over Congress’ enactment and of assumption over fact. Whatever respect Griggs merits as a matter of stare decisis, I would not amplify its error by importing its disparate-impact scheme into yet another statute.

I

A

We should drop the pretense that Griggs’ interpretation of Title VII was legitimate. “The Civil Rights Act of 1964 did not include an express prohibition on policies or practices that produce a disparate impact.” Ricci v. DeStefano, 557 U. S. 557, 577 (2009) . It did not include an implicit one either. Instead, Title VII’s operative provision, 42 U. S. C. §2000e–2(a) (1964 ed.), addressed only employer decisions motivated by a protected characteristic. That provision made it “an unlawful employment practice for an employer—

“(1) 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, color, religion, sex, or national origin; or

“(2) 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, color, religion, sex, or national origin.” §703, 78Stat. 255 (emphasis added).[1]

Each paragraph in §2000e–2(a) is limited to actions taken “because of” a protected trait, and “the ordinary meaning of ‘because of’ is ‘by reason of’ or ‘on account of,’ ” University of Tex. Southwestern Medical Center v. Nassar, 570 U. S. ___, ___ (2013) (slip op., at 9) (some internal quotation marks omitted). Section 2000e–2(a) thus applies only when a protected characteristic “was the ‘reason’ that the employer decided to act.” Id., at ___ (slip op., at 10) (some internal quotation marks omitted).[2] In other words, “to take action against an individual because of” a protected trait “plainly requires discriminatory intent.” See Smith v. City of Jackson, 544 U. S. 228, 249 (2005) (O’Connor, J., joined by Kennedy and Thomas, JJ., concurring in judgment) (internal quotation marks omitted); accord, e.g., Gross v. FBL Financial Services, Inc., 557 U. S. 167, 176 (2009) .

No one disputes that understanding of §2000e–2(a)(1). We have repeatedly explained that a plaintiff bringing an action under this provision “must establish ‘that the defendant had a discriminatory intent or motive’ for taking a job-related action.” Ricci, supra, at 577 (quoting Watson v. Fort Worth Bank & Trust, 487 U. S. 977, 986 (1988) ). The only dispute is whether the same language—“because of”—means something different in §2000e–2(a)(2) than it does in §2000e–2(a)(1).

The answer to that question should be obvious. We ordinarily presume that “identical words used in different parts of the same act are intended to have the same meaning,” Desert Palace, Inc. v. Costa, 539 U. S. 90, 101 (2003) (internal quotation marks omitted), and §2000e–2(a)(2) contains nothing to warrant a departure from that presumption. That paragraph “uses the phrase ‘because of . . . [a protected characteristic]’ in precisely the same manner as does the preceding paragraph—to make plain that an employer is liable only if its adverse action against an individual is motivated by the individual’s [protected characteristic].” Smith, supra, at 249 (opinion of O’Connor, J.) (interpreting nearly identical provision of the Age Discrimination in Employment Act of 1967 (ADEA)).

The only difference between §2000e–2(a)(1) and §2000e–2(a)(2) is the type of employment decisions they address. See Smith, supra, at 249 (opinion of O’Connor, J.). Section 2000e–2(a)(1) addresses hiring, firing, and setting the terms of employment, whereas §2000e–2(a)(2) generally addresses limiting, segregating, or classifying employees. But no decision is an unlawful employment practice under these paragraphs unless it occurs “because of such individual’s race, color, religion, sex, or national origin.” §§2000e–2(a)(1), (2) (emphasis added).

Contrary to the majority’s assumption, see ante, at 10–13, the fact that §2000e–2(a)(2) uses the phrase “otherwise adversely affect” in defining the employment decisions targeted by that paragraph does not eliminate its mandate that the prohibited decision be made “because of” a protected characteristic. Section 2000e–2(a)(2) does not make unlawful all employment decisions that “limit, segregate, or classify . . . employees . . . in any way which would . . . otherwise adversely affect [an individual’s] status as an employee,” but those that “otherwise adversely affect [an individual’s] status as an employee, because of such individual’s race, color, religion, sex, or national origin.” (Emphasis added); accord, 78Stat. 255. Reading §2000e–2(a)(2) to sanction employers solely on the basis of the effects of their decisions would delete an entire clause of this provision, a result we generally try to avoid. Under any fair reading of the text, there can be no doubt that the Title VII enacted by Congress did not permit disparate-impact claims.[3]

B

The author of disparate-impact liability under Title VII was not Congress, but the Equal Employment Opportunity Commission (EEOC). EEOC’s “own official history of these early years records with unusual candor the commission’s fundamental disagreement with its founding charter, especially Title VII’s literal requirement that the discrimination be intentional.” H. Graham, The Civil Rights Era: Origins and Development of National Policy 1960–1972, p. 248 (1990). The Commissioners and their legal staff thought that “discrimination” had become “less often an individual act of disparate treatment flowing from an evil state of mind” and “more institutionalized.” Jackson, EEOC vs. Discrimination, Inc., 75 The Crisis 16 (1968). They consequently decided they should target employment practices “which prove to have a demonstrable racial effect without a clear and convincing business motive.” Id., at 16–17 (emphasis deleted). EEOC’s “legal staff was aware from the beginning that a normal, traditional, and literal interpretation of Title VII could blunt their efforts” to penalize employers for practices that had a disparate impact, yet chose “to defy Title VII’s restrictions and attempt to build a body of case law that would justify [their] focus on effects and [their] disregard of intent.” Graham, supra, at 248, 250.

The lack of legal authority for their agenda apparently did not trouble them much. For example, Alfred Blumrosen, one of the principal creators of disparate-impact liability at EEOC, rejected what he described as a “defeatist view of Title VII” that saw the statute as a “compromise” with a limited scope. A. Blumrosen, Black Employment and the Law 57–58 (1971). Blumrosen “felt that most of the problems confronting the EEOC could be solved by creative interpretation of Title VII which would be upheld by the courts, partly out of deference to the administrators.” Id., at 59.

EEOC’s guidelines from those years are a case study in Blumrosen’s “creative interpretation.” Although EEOC lacked substantive rulemaking authority, see Faragher v. Boca Raton, 524 U. S. 775 , n. 1 (1998) (Thomas, J., dissenting), it repeatedly issued guidelines on the subject of disparate impact. In 1966, for example, EEOC issued guidelines suggesting that the use of employment tests in hiring decisions could violate Title VII based on disparate impact, notwithstanding the statute’s express statement that “it shall not be an unlawful employment practice . . . to give and to act upon the results of any professionally developed ability test provided that such test . . . is not designed, intended, or used to discriminate because of race, color, religion, sex, or national origin,” §2000e–2(h) (emphasis added). See EEOC, Guidelines on Employment Testing Procedures 2–4 (Aug. 24, 1966). EEOC followed this up with a 1970 guideline that was even more explicit, declaring that, unless certain criteria were met, “[t]he use of any test which adversely affects hiring, promotion, transfer or any other employment or membership opportunity of classes protected by title VII constitutes discrimination.” 35 Fed. Reg. 12334 (1970).

EEOC was initially hesitant to take its approach to this Court, but the Griggs plaintiffs forced its hand. After they lost on their disparate-impact argument in the Court of Appeals, EEOC’s deputy general counsel urged the plaintiffs not to seek review because he believed “ ‘that the record in the case present[ed] a most unappealing situation for finding tests unlawful,’ ” even though he found the lower court’s adherence to an intent requirement to be “ ‘tragic.’ ” Graham, supra, at 385. The plaintiffs ignored his advice. Perhaps realizing that a ruling on its disparate-impact theory was inevitable, EEOC filed an amicusbrief in this Court seeking deference for its position.[4]

EEOC’s strategy paid off. The Court embraced EEOC’s theory of disparate impact, concluding that the agency’s position was “entitled to great deference.” See Griggs, 401 U. S., at 433–434. With only a brief nod to the text of §2000e–2(a)(2) in a footnote, id., at 426, n. 1, the Court tied this novel theory of discrimination to “the statute’s perceived purpose” and EEOC’s view of the best way of effectuating it, Smith, 544 U. S., at 262 (opinion of O’Connor, J.); see id., at 235 (plurality opinion). But statutory provisions—not purposes—go through the process of bicameralism and presentment mandated by our Constitution. We should not replace the former with the latter, see Wyeth v. Levine, 555 U. S. 555, 586 (2009) (Thomas, J., concurring in judgment), nor should we transfer our responsibility for interpreting those provisions to administrative agencies, let alone ones lacking substantive rulemaking authority, see Perez v. Mortgage Bankers Assn., 575 U. S. ___, ___–___ (2015) (Thomas, J., concurring in judgment) (slip op., at 8–13).

II

Griggs’ disparate-impact doctrine defies not only the statutory text, but reality itself. In their quest to eradicate what they view as institutionalized discrimination, disparate-impact proponents doggedly assume that a given racial disparity at an institution is a product of that institution rather than a reflection of disparities that exist outside of it. See T. Sowell, Intellectuals and Race 132 (2013) (Sowell). That might be true, or it might not. Standing alone, the fact that a practice has a disparate impact is not conclusive evidence, as the Griggs Court appeared to believe, that a practice is “discriminatory,” 401 U. S., at 431. “Although presently observed racial imbalance might result from past [discrimination], racial imbalance can also result from any number of innocent private decisions.” Parents Involved in Community Schools v. Seattle School Dist. No. 1, 551 U. S. 701, 750 (2007) (Thomas, J., concurring) (emphasis added).[5] We should not automatically presume that any institution with a neutral practice that happens to produce a racial disparity is guilty of discrimination until proved innocent.

As best I can tell, the reason for this wholesale inversion of our law’s usual approach is the unstated—and unsubstantiated—assumption that, in the absence of discrimination, an institution’s racial makeup would mirror that of society. But the absence of racial disparities in multi-ethnic societies has been the exception, not the rule. When it comes to “proportiona[l] represent[ation]” of ethnic groups, “few, if any, societies have ever approximated this description.” D. Horowitz, Ethnic Groups in Conflict 677 (1985). “All multi-ethnic societies exhibit a tendency for ethnic groups to engage in different occupations, have different levels (and, often, types) of education, receive different incomes, and occupy a different place in the social hierarchy.” Weiner, The Pursuit of Ethnic Equality Through Preferential Policies: A Comparative Public Policy Perspective, in From Independence to Statehood 64 (R. Goldmann & A. Wilson eds. 1984).

Racial imbalances do not always disfavor minorities. At various times in history, “racial or ethnic minorities . . . have owned or directed more than half of whole industries in particular nations.” Sowell 8. These minorities “have included the Chinese in Malaysia, the Lebanese in West Africa, Greeks in the Ottoman Empire, Britons in Argentina, Belgians in Russia, Jews in Poland, and Spaniards in Chile—among many others.” Ibid. (footnotes omitted). “In the seventeenth century Ottoman Empire,” this phenomenon was seen in the palace itself, where the “medical staff consisted of 41 Jews and 21 Muslims.” Ibid. And in our own country, for roughly a quarter-century now, over 70 percent of National Basketball Association players have been black. R. Lapchick, D. Donovan, E. Loomer, & L. Martinez, Institute for Diversity and Ethics in Sport, U. of Central Fla., The 2014 Racial and Gender Report Card: National Basketball Association 21 (June 24, 2014). To presume that these and all other measurable disparities are products of racial discrimination is to ignore the complexities of human existence.

Yet, if disparate-impact liability is not based on this assumption and is instead simply a way to correct for imbalances that do not result from any unlawful conduct, it is even less justifiable. This Court has repeatedly reaffirmed that “ ‘racial balancing’ ” by state actors is “ ‘pat-ently unconstitutional,’ ” even when it supposedly springs from good intentions. Fisher v. University of Tex. at Austin, 570 U. S. ___, ___ (2013) (slip op., at 9). And if that “racial balancing” is achieved through disparate-impact claims limited to only some groups—if, for instance, white basketball players cannot bring disparate-impact suits—then we as a Court have constructed a scheme that parcels out legal privileges to individuals on the basis of skin color. A problem with doing so should be obvious: “Government action that classifies individuals on the basis of race is inherently suspect.” Schuette v. BAMN, 572 U. S. ___, ___ (2014) (plurality opinion) (slip op., at 12); accord, id., at ___ (Scalia, J., concurring in judgment) (slip op., at 9). That is no less true when judges are the ones doing the classifying. See id., at ___ (plurality opinion) (slip op., at 12); id., at ___ (Scalia, J., concurring in judgment) (slip op., at 9). Disparate-impact liability is thus a rule without a reason, or at least without a legitimate one.

III

The decision in Griggs was bad enough, but this Court’s subsequent decisions have allowed it to move to other areas of the law. In Smith, for example, a plurality of this Court relied on Griggs to include disparate-impact liability in the ADEA. See 544 U. S., at 236. As both I and the author of today’s majority opinion recognized at the time, that decision was as incorrect as it was regrettable. See id., at 248–249 (O’Connor, J., joined by Kennedy and Thomas, JJ., concurring in judgment). Because we knew that Congress did not create disparate-impact liability under Title VII, we explained that “there [wa]s no reason to suppose that Congress in 1967”—four years before Griggs—“could have foreseen the interpretation of Title VII that was to come.” Smith, supra, at 260 (opinion of O’Connor, J.). It made little sense to repeat Griggs’ error in a new context.

My position remains the same. Whatever deference is due Griggs as a matter of stare decisis, we should at the very least confine it to Title VII. We should not incorporate it into statutes such as the Fair Housing Act and the ADEA, which were passed years before Congress had any reason to suppose that this Court would take the position it did in Griggs. See Smith, supra, at 260 (opinion of O’Connor, J.). And we should certainly not allow it to spread to statutes like the Fair Housing Act, whose operative text, unlike that of the ADEA’s, does not even mirror Title VII’s.

Today, however, the majority inexplicably declares that “the logic of Smith and Griggs” leads to the conclusion that “the FHA encompasses disparate-impact claims.” Ante, at 11. Justice Alito ably dismantles this argument. Post, at 21–28 (dissenting opinion). But, even if the majority were correct, I would not join it in following that “logic” here. “[E]rroneous precedents need not be extended to their logical end, even when dealing with related provisions that normally would be interpreted in lockstep. Otherwise, stare decisis, designed to be a principle of stability and repose, would become a vehicle of change . . . distorting the law.” CBOCS West, Inc. v. Humphries, 553 U. S. 442 –470 (2008) (Thomas, J., dissenting) (footnote omitted). Making the same mistake in different areas of the law furthers neither certainty nor judicial economy. It furthers error.

That error will take its toll. The recent experience of the Houston Housing Authority (HHA) illustrates some of the many costs of disparate-impact liability. HHA, which provides affordable housing developments to low-income residents of Houston, has over 43,000 families on its waiting lists. The overwhelming majority of those families are black. Because Houston is a majority-minority city with minority concentrations in all but the more affluent areas, any HHA developments built outside of those areas will increase the concentration of racial minorities. Unsurprisingly, the threat of disparate-impact suits based on those concentrations has hindered HHA’s efforts to provide affordable housing. State and federal housing agencies have refused to approve all but two of HHA’s eight proposed development projects over the past two years out of fears of disparate-impact liability. Brief for Houston Housing Authority as Amicus Curiae 8–12. That the majority believes that these are not “ ‘dire consequences,’ ” see ante, at 24, is cold comfort for those who actually need a home.

*  *  *

I agree with the majority that Griggs “provide[s] essential background” in this case, ante, at 10: It shows that our disparate-impact jurisprudence was erroneous from its inception. Divorced from text and reality, driven by an agency with its own policy preferences, Griggs bears little relationship to the statutory interpretation we should expect from a court of law. Today, the majority repeats that error.

I respectfully dissent.

SUPREME COURT OF THE UNITED STATES

_________________

No. 13–1371

_________________

TEXAS DEPARTMENT OF HOUSING AND COMMU-NITY AFFAIRS, et al., PETITIONERS v. THE IN-CLUSIVE COMMUNITIES PROJECT, INC., et al.

on writ of certiorari to the united states court of appeals for the fifth circuit

[June 25, 2015]

Justice Alito, with whom The Chief Justice, Justice Scalia, and Justice Thomas join, dissenting.

No one wants to live in a rat’s nest. Yet in Gallagher v. Magner, 619 F. 3d 823 (2010), a case that we agreed to review several Terms ago, the Eighth Circuit held that the Fair Housing Act (or FHA), 42 U. S. C. §3601 et seq., could be used to attack St. Paul, Minnesota’s efforts to combat “rodent infestation” and other violations of the city’s housing code. 619 F. 3d, at 830. The court agreed that there was no basis to “infer discriminatory intent” on the part of St. Paul. Id., at 833. Even so, it concluded that the city’s “aggressive enforcement of the Housing Code” was actionable because making landlords respond to “rodent infestation, missing dead-bolt locks, inadequate sanitation facilities, inadequate heat, inoperable smoke detectors, broken or missing doors,” and the like increased the price of rent. Id., at 830, 835. Since minorities were statistically more likely to fall into “the bottom bracket for household ad-justed median family income,” they were disproportionately affected by those rent increases, i.e., there was a “dis-parate impact.” Id., at 834. The upshot was that even St. Paul’s good-faith attempt to ensure minimally acceptable housing for its poorest residents could not ward off a disparate-impact lawsuit.

Today, the Court embraces the same theory that drove the decision in Magner.[1] This is a serious mistake. The Fair Housing Act does not create disparate-impact liability, nor do this Court’s precedents. And today’s decision will have unfortunate consequences for local government, private enterprise, and those living in poverty. Something has gone badly awry when a city can’t even make slumlords kill rats without fear of a lawsuit. Because Congress did not authorize any of this, I respectfully dissent.

I

Everyone agrees that the FHA punishes intentional discrimination. Treating someone “less favorably than others because of a protected trait” is “ ‘the most easily understood type of discrimination.’ ” Ricci v. DeStefano, 557 U. S. 557, 577 (2009) (quoting Teamsters v. United States, 431 U. S. 324 , n. 15 (1977); some internal quotation marks omitted). Indeed, this classic form of discrimination—called disparate treatment—is the only one prohibited by the Constitution itself. See, e.g., Arlington Heights v. Metropolitan Housing Development Corp., 429 U. S. 252 –265 (1977). It is obvious that Congress intended the FHA to cover disparate treatment.

The question presented here, however, is whether the FHA also punishes “practices that are not intended to discriminate but in fact have a disproportionately adverse effect on minorities.” Ricci, supra, at 577. The answer is equally clear. The FHA does not authorize disparate-impact claims. No such liability was created when the law was enacted in 1968. And nothing has happened since then to change the law’s meaning.

A

I begin with the text. Section 804(a) of the FHA makes it unlawful “[t]o refuse to sell or rent after the making of a bona fide offer, or to refuse to negotiate for the sale or rental of, or otherwise make unavailable or deny, a dwelling to any person because of race, color, religion, sex, familial status, or national origin.” 42 U. S. C. §3604(a) (emphasis added). Similarly, §805(a) prohibits any party “whose business includes engaging in residential real estate-related transactions” from “discriminat[ing] against any person in making available such a transaction, or in the terms or conditions of such a transaction, because of race, color, religion, sex, handicap, familial status, or national origin.” §3605(a) (emphasis added).

In both sections, the key phrase is “because of.” These provisions list covered actions (“refus[ing] to sell or rent . . . a dwelling,” “refus[ing] to negotiate for the sale or rental of . . . a dwelling,” “discriminat[ing]” in a residential real estate transaction, etc.) and protected characteristics (“race,” “religion,” etc.). The link between the actions and the protected characteristics is “because of.”

What “because of” means is no mystery. Two Terms ago, we held that “the ordinary meaning of ‘because of’ is ‘by reason of’ or ‘on account of.’ ” University of Tex. Southwestern Medical Center v. Nassar, 570 U. S. ___, ___ (2013) (slip op., at 9) (quoting Gross v. FBL Financial Services, Inc., 557 U. S. 167, 176 (2009) ; some internal quotation marks omitted). A person acts “because of” something else, we explained, if that something else “ ‘was the “reason” that the [person] decided to act.’ ” 570 U. S., at ___ (slip op., at 10).

Indeed, just weeks ago, the Court made this same point in interpreting a provision of Title VII of the Civil Rights Act of 1964, 42 U. S. C. §2000e–2(m), that makes it unlawful for an employer to take a variety of adverse employment actions (such as failing or refusing to hire a job applicant or discharging an employee) “because of” religion. See EEOC v. Abercrombie & Fitch Stores, Inc., 575 U. S. ___, ___ (2015) (slip op., at 4). The Court wrote: “ ‘Because of’ in §2000e–2(a)(1) links the forbidden consideration to each of the verbs preceding it.” Ibid.

Nor is this understanding of “because of” an arcane feature of legal usage. When English speakers say that someone did something “because of” a factor, what they mean is that the factor was a reason for what was done. For example, on the day this case was argued, January 21, 2015, Westlaw and Lexis searches reveal that the phrase “because of” appeared in 14 Washington Post print articles. In every single one, the phrase linked an action and a reason for the action.[2]

Without torturing the English language, the meaning of these provisions of the FHA cannot be denied. They make it unlawful to engage in any of the covered actions “because of”—meaning “by reason of” or “on account of,” Nassar, supra, at ___ (slip op., at 9)—race, religion, etc. Put another way, “the terms [after] the ‘because of’ clauses in the FHA supply the prohibited motivations for the intentional acts . . . that the Act makes unlawful.” American Ins. Assn. v. Department of Housing and Urban Development, ___ F. Supp. 3d ___, ___, n. 20, 2014 WL 5802283, at *8, n. 20 (DC 2014). Congress accordingly outlawed the covered actions only when they are motivated by race or one of the other protected characteristics.

It follows that the FHA does not authorize disparate-impact suits. Under a statute like the FHA that prohibits actions taken “because of” protected characteristics, intent makes all the difference. Disparate impact, however, does not turn on “ ‘subjective intent.’ ” Raytheon Co. v. Hernandez, 540 U. S. 44, 53 (2003) . Instead, “ ‘treat[ing] [a] particular person less favorably than others because of’ a protected trait” is “ ‘disparate treatment,’ ” not disparate impact. Ricci, 557 U. S., at 577 (emphasis added). See also, e.g., Personnel Administrator of Mass. v. Feeney, 442 U. S. 256, 279 (1979) (explaining the difference between “because of” and “in spite of”); Hernandez v. New York, 500 U. S. 352 –360 (1991) (plurality opinion) (same); Alexander v. Sandoval, 532 U. S. 275, 278, 280 (2001) (holding that it is “beyond dispute” that banning discrimination “ ‘on the ground of race’ ” “prohibits only intentional discrimination”).

This is precisely how Congress used the phrase “because of” elsewhere in the FHA. The FHA makes it a crime to willfully “interfere with . . . any person because of his race” (or other protected characteristic) who is engaging in a variety of real-estate-related activities, such as “selling, purchasing, [or] renting” a dwelling. 42 U. S. C. §3631(a). No one thinks a defendant could be convicted of this crime without proof that he acted “because of,” i.e., on account of or by reason of, one of the protected characteristics. But the critical language in this section—“because of”—is identical to the critical language in the sections at issue in this case. “One ordinarily assumes” Congress means the same words in the same statute to mean the same thing. Utility Air Regulatory Group v. EPA, 573 U. S. ___, ___ (2014) (slip op., at 15). There is no reason to doubt that ordinary assumption here.

Like the FHA, many other federal statutes use the phrase “because of” to signify what that phrase means in ordinary speech. For instance, the federal hate crime statute, 18 U. S. C. §249, authorizes enhanced sentences for defendants convicted of committing certain crimes “because of” race, color, religion, or other listed characteristics. Hate crimes require bad intent—indeed, that is the whole point of these laws. See, e.g., Wisconsin v. Mitchell, 508 U. S. 476 –485 (1993) (“[T]he same criminal conduct may be more heavily punished if the victim is selected because of his race or other protected status”). All of this confirms that “because of” in the FHA should be read to mean what it says.

B

In an effort to find at least a sliver of support for disparate-impact liability in the text of the FHA, the principalrespondent, the Solicitor General, and the Court pounce on the phrase “make unavailable.” Under §804(a), it is unlawful “[t]o . . . make unavailable . . . a dwelling to any person because of race, color, religion, sex, familial status, or national origin.” 42 U. S. C. §3604(a). See also §3605(a) (barring “discriminat[ion] against any person in making available such a [housing] transaction . . . because of race, color, religion, sex, handicap, familial status, or national origin”). The Solicitor General argues that “[t]he plain meaning of the phrase ‘make unavailable’ includes actions that have the result of making housing or transactions unavailable, regardless of whether the actions were intended to have that result.” Brief for United States as Amicus Curiae 18 (emphasis added). This argument is not consistent with ordinary English usage.

It is doubtful that the Solicitor General’s argument accurately captures the “plain meaning” of the phrase “make unavailable” even when that phrase is not linked to the phrase “because of.” “[M]ake unavailable” must be viewed together with the rest of the actions covered by §804(a), which applies when a party “refuse[s] to sell or rent” a dwelling, “refuse[s] to negotiate for the sale or rental” of a dwelling, “den[ies] a dwelling to any person,” “or otherwise make[s] unavailable” a dwelling. §3604(a) (emphasis added). When a statute contains a list like this, we “avoid ascribing to one word a meaning so broad that it is inconsistent with its accompanying words, thus giving ‘unintended breadth to the Acts of Congress.’ ” Gustafson v. Alloyd Co., 513 U. S. 561, 575 (1995) (quoting Jarecki v. G. D. Searle & Co., 367 U. S. 303, 307 (1961) ). See also, e.g., Yates v. United States, 574 U. S. ___, ___ (2015) (plurality opinion) (slip op., at 14); id., at ___ (Alito, J., concurring in judgment) (slip op., at 1). Here, the phrases that precede “make unavailable” unmistakably describe intentional deprivations of equal treatment, not merely actions that happen to have a disparate effect. See American Ins. Assn., ___ F. Supp. 3d, at ___, 2014 WL 5802283, at *8 (citing Webster’s Third New International Dictionary 603, 848, 1363, 1910 (1966)). Section 804(a), moreover, prefaces “make unavailable” with “or otherwise,” thus creating a catchall. Catchalls must be read “restrictively” to be “like” the listed terms. Washington State Dept. of Social and Health Servs. v. Guardianship Estate of Keffeler, 537 U. S. 371 –385 (2003). The result of these ordinary rules of interpretation is that even without “because of,” the phrase “make unavailable” likely would require intentionality.

The FHA’s inclusion of “because of,” however, removes any doubt. Sections 804(a) and 805(a) apply only when a party makes a dwelling or transaction unavailable “because of” race or another protected characteristic. In ordinary English usage, when a person makes something unavailable “because of” some factor, that factor must be a reason for the act.

Here is an example. Suppose that Congress increases the minimum wage. Some economists believe that such legislation reduces the number of jobs available for “unskilled workers,” Fuller & Geide-Stevenson, Consensus Among Economists: Revisited, 34 J. Econ. Educ. 369, 378 (2003), and minorities tend to be disproportionately represented in this group, see, e.g., Dept. of Commerce, Bureau of Census, Detailed Years of School Completed by People 25 Years and Over by Sex, Age Groups, Race and Hispanic Origin: 2014, online at http://www.census.gov/hhes/socdemo/education/data/cps/2014/tables.html (all Inter-net materials as visited June 23, 2015, and available in Clerk of Court’s case file). Assuming for the sake of argument that these economists are correct, would it be fair to say that Congress made jobs unavailable to African-Americans or Latinos “because of” their race or ethnicity?

A second example. Of the 32 college players selected by National Football League (NFL) teams in the first round of the 2015 draft, it appears that the overwhelming majority were members of racial minorities. See Draft 2015, http://www.nfl.com/draft/2015. See also Miller, Powerful Sports Agents Representing Color, Los Angeles Sentinel, Feb. 6, 2014, p. B3 (noting “there are 96 players (76 of whom are African-American) chosen in the first rounds of the 2009, 2010, and 2011 NFL drafts”). Teams presumably chose the players they think are most likely to help them win games. Would anyone say the NFL teams made draft slots unavailable to white players “because of” their race?

A third example. During the present Court Term, of the 21 attorneys from the Solicitor General’s Office who argued cases in this Court, it appears that all but 5 (76%) were under the age of 45. Would the Solicitor General say he made argument opportunities unavailable to older attorneys “because of” their age?

The text of the FHA simply cannot be twisted to authorize disparate-impact claims. It is hard to imagine how Congress could have more clearly stated that the FHA prohibits only intentional discrimination than by forbidding acts done “because of race, color, religion, sex, familial status, or national origin.”

II

The circumstances in which the FHA was enacted only confirm what the text says. In 1968, “the predominant focus of antidiscrimination law was on intentional discrimination.” Smith v. City of Jackson, 544 U. S. 228, 258 (2005) (O’Connor, J., concurring in judgment). The very “concept of disparate impact liability, by contrast, was quite novel.” Ibid. (collecting citations). See also Tr. of Oral Arg. 15 (“Justice Ginsburg: . . . If we’re going to be realistic about this, . . . in 1968, when the Fair Housing Act passed, nobody knew anything about disparate impact”). It is anachronistic to think that Congress authorized disparate-impact claims in 1968 but packaged that striking innovation so imperceptibly in the FHA’s text.

Eradicating intentional discrimination was and is the FHA’s strategy for providing fair housing opportunities for all. The Court recalls the country’s shameful history of segregation and de jure housing discrimination and then jumps to the conclusion that the FHA authorized disparate-impact claims as a method of combatting that evil.Ante, at 5–7. But the fact that the 1968 Congress sought to end housing discrimination says nothing about the means it devised to achieve that end. The FHA’s text plainly identifies the weapon Congress chose—outlawing disparate treatment “because of race” or another protected characteristic. 42 U. S. C. §§3604(a), 3605(a). Accordingly, in any FHA claim, “[p]roof of discriminatory motive is critical.” Teamsters, 431 U. S., at 335, n. 15.

III

Congress has done nothing since 1968 to change the meaning of the FHA prohibitions at issue in this case. In 1968, those prohibitions forbade certain housing practices if they were done “because of” protected characteristics. Today, they still forbid certain housing practices if done “because of” protected characteristics. The meaning of the unaltered language adopted in 1968 has not evolved.

Rather than confronting the plain text of §§804(a) and 805(a), the Solicitor General and the Court place heavy reliance on certain amendments enacted in 1988, but those amendments did not modify the meaning of the provisions now before us. In the Fair Housing Amendments Act of 1988, 102Stat. 1619, Congress expanded the list of protected characteristics. See 42 U. S. C. §§3604(a), (f )(1). Congress also gave the Department of Housing and Urban Development (HUD) rulemaking authority and the power to adjudicate certain housing claims. See §§3612, 3614a. And, what is most relevant for present purposes, Congress added three safe-harbor provisions, specifying that “[n]othing in [the FHA]” prohibits (a) certain actions taken by real property appraisers, (b) certain occupancy requirements, and (c) the treatment of persons convicted of manufacturing or distributing illegal drugs.[3]

According to the Solicitor General and the Court, these amendments show that the FHA authorizes disparate-impact claims. Indeed, the Court says that they are “of crucial importance.” Ante, at 13. This “crucial” argument, however, cannot stand.

A

The Solicitor General and the Court contend that the 1988 Congress implicitly authorized disparate-impact liability by adopting the amendments just noted while leaving the operative provisions of the FHA untouched. Congress knew at that time, they maintain, that the Courts of Appeals had held that the FHA sanctionsdisparate-impact claims, but Congress failed to enact bills that would have rejected that theory of liability. Based on this, they submit that Congress silently ratified those decisions. See ante, at 13–14; Brief for United States as Amicus Curiae 23–24. This argument is deeply flawed.

Not the greatest of its defects is its assessment of what Congress must have known about the judiciary’s interpretation of the FHA. The Court writes that by 1988, “all nine Courts of Appeals to have addressed the question had concluded the Fair Housing Act encompassed disparate-impact claims.” Ante, at 13 (emphasis added). See also Brief for United States as Amicus Curiae 12. But this Court had not addressed that question. While we always give respectful consideration to interpretations of statutes that garner wide acceptance in other courts, this Court has “no warrant to ignore clear statutory language on the ground that other courts have done so,” even if they have “ ‘consistently’ ” done so for “ ‘30 years.’ ” Milner v. Department of Navy, 562 U. S. 562 –576 (2011). See also, e.g., CSX Transp., Inc. v. McBride, 564 U. S. ___, ___ (2011) (Roberts, C. J., dissenting) (slip op., at 11) (explaining that this Court does not interpret statutes by asking for “a show of hands” (citing Buckhannon Board & Care Home, Inc. v. West Virginia Dept. of Health and Human Resources, 532 U. S. 598 (2001) ; McNally v. United States, 483 U. S. 350 (1987) )).

In any event, there is no need to ponder whether it would have been reasonable for the 1988 Congress, without considering the clear meaning of §§804(a) and 805(a), to assume that the decisions of the lower courts effectively settled the matter. While the Court highlights the decisions of the Courts of Appeals, it fails to mention something that is of at least equal importance: The official view of the United States in 1988.

Shortly before the 1988 amendments were adopted, the United States formally argued in this Court that the FHA prohibits only intentional discrimination. See Brief for United States as Amicus Curiae in Huntington v. Huntington Branch, NAACP, O. T. 1988, No. 87–1961, p. 15 (“An action taken because of some factor other than race, i.e., financial means, even if it causes a discriminatory effect, is not an example of the intentional discrimination outlawed by the statute”); id., at 14 (“The words ‘because of’ plainly connote a causal connection between the housing-related action and the person’s race or color”).[4] Thiswas the same position that the United States had taken in lower courts for years. See, e.g., United States v. Birmingham, 538 F. Supp. 819, 827, n. 9 (ED Mich. 1982) (noting positional change), aff’d, 727 F. 2d 560, 565–566 (CA6 1984) (adopting United States’ “concession” that there must be a “ ‘discriminatory motive’ ”). It is implausible that the 1988 Congress was aware of certain lower court decisions but oblivious to the United States’ considered and public view that those decisions were wrong.

This fact is fatal to any notion that Congress implicitly ratified disparate impact in 1988. The canon of interpretation on which the Court and the Solicitor General purport to rely—the so-called “prior-construction canon”—does not apply where lawyers cannot “justifiably regard the point as settled” or when “other sound rules of interpretation” are implicated. A. Scalia & B. Garner, Reading Law: The Interpretation of Legal Texts 324, 325 (2012). That was the case here. Especially after the United States began repudiating disparate impact, no one could have reasonably thought that the question was settled.

Nor can such a faulty argument be salvaged by pointing to Congress’ failure in 1988 to enact language that would have made it clear that the FHA does not authorizedisparate-impact suits based on zoning decisions. See ante, at 13–14.[5] To change the meaning of language in an already enacted law, Congress must pass a new law amending that language. See, e.g., West Virginia Univ. Hospitals, Inc. v. Casey, 499 U. S. 83 , and n. 7 (1991). Intent that finds no expression in a statute is irrelevant. See, e.g., New York Telephone Co. v. New York State Dept. of Labor, 440 U. S. 519 –545 (1979); Easterbrook, Statutes’ Domains, 50 U. Chi. L. Rev. 533, 538–540 (1983). Hence, “we walk on quicksand when we try to find in the absence of corrective legislation a controlling legal principle.” Helvering v. Hallock, 309 U. S. 106, 121 (1940) .

Unsurprisingly, we have rejected identical arguments about implicit ratification in other cases. For example, in Central Bank of Denver, N. A. v. First Interstate Bank of Denver, N. A., 511 U. S. 164 (1994) , a party argued that §10(b) of the Securities Exchange Act of 1934 imposes liability on aiders and abettors because “Congress ha[d] amended the securities laws on various occasions since 1966, when courts first began to interpret §10(b) to cover aiding and abetting, but ha[d] done so without providing that aiding and abetting liability is not available under §10(b).” Id., at 186. “From that,” a party asked the Court to “infer that these Congresses, by silence, ha[d] acquiesced in the judicial interpretation of §10(b).” Ibid. The Court dismissed this argument in words that apply almost verbatim here:

“ ‘It does not follow that Congress’ failure to overturn a statutory precedent is reason for this Court to adhere to it. It is “impossible to assert with any degree of assurance that congressional failure to act represents” affirmative congressional approval of the courts’ statutory interpretation. Congress may legislate, moreover, only through the passage of a bill which is approved by both Houses and signed by the President. See U. S. Const., Art. I, §7, cl. 2. Congressional inaction cannot amend a duly enacted statute.’ Patterson v. McLean Credit Union, 491 U. S. 164 , n. 1 (1989) (quoting Johnson v. Transportation Agency, Santa Clara Cty., 480 U. S. 616, 672 (1987) (Scalia, J., dissenting)).” Ibid. (alterations omitted).

We made the same point again in Sandoval, 532 U. S. 275 . There it was argued that amendments to Title VI of the Civil Rights Act of 1964 implicitly ratified lower court decisions upholding a private right of action. We rejected that argument out of hand. See id., at 292–293.

Without explanation, the Court ignores these cases.

B

The Court contends that the 1988 amendments provide “convincing confirmation of Congress’ understanding that disparate-impact liability exists under the FHA” because the three safe-harbor provisions included in those amendments “would be superfluous if Congress had assumed that disparate-impact liability did not exist under the FHA.” Ante, at 14, 15. As just explained, however, what matters is what Congress did, not what it might have “assumed.” And although the Court characterizes these provisions as “exemptions,” that characterization is inaccurate. They make no reference to §804(a) or §805(a) or any other provision of the FHA; nor do they state that they apply to conduct that would otherwise be prohibited. Instead, they simply make clear that certain conduct is not forbidden by the Act. E.g., 42 U. S. C. §3607(b)(4) (“Nothing in this subchapter prohibits . . .”). The Court should read these amendments to mean what they say.

In 1988, policymakers were not of one mind about disparate-impact housing suits. Some favored the theory and presumably would have been happy to have it enshrined in the FHA. See ante, at 13–14; 134 Cong. Rec. 23711 (1988) (statement of Sen. Kennedy). Others worried about disparate-impact liability and recognized that this Court had not decided whether disparate-impact claims were authorized under the 1968 Act. See H. R. Rep. No. 100–711, pp. 89–93 (1988). Still others disapproved of disparate-impact liability and believed that the 1968 Act did not authorize it. That was the view of President Reagan when he signed the amendments. See Remarks on Signing the Fair Housing Amendments Act of 1988, 24 Weekly Comp. of Pres. Doc. 1140, 1141 (1988) (explaining that the amendments did “not represent any congressional or executive branch endorsement of the notion, expressed in some judicial opinions, that [FHA] violations may be established by a showing of disparate impact” because the FHA “speaks only to intentional discrimination”).[6]

The 1988 safe-harbor provisions have all the hallmarks of a compromise among these factions. These provisions neither authorize nor bar disparate-impact claims, but they do provide additional protection for persons and entities engaging in certain practices that Congress especially wished to shield. We “must respect and give effect to these sorts of compromises.” Ragsdale v. Wolverine World Wide, Inc., 535 U. S. 81 –94 (2002).

It is not hard to see why such a compromise was attractive. For Members of Congress who supported disparate impact, the safe harbors left the favorable lower court decisions in place. And for those who hoped that this Court would ultimately agree with the position being urged by the United States, those provisions were not surplusage. In the Circuits in which disparate-impact FHA liability had been accepted, the safe-harbor provisions furnished a measure of interim protection until the question was resolved by this Court. They also provided partial protection in the event that this Court ultimately rejected the United States’ argument. Neither the Court, the principal respondent, nor the Solicitor General has cited any case in which the canon against surplusage has been applied in circumstances like these.[7]

On the contrary, we have previously refused to interpret enactments like the 1988 safe-harbor provisions in such a way. Our decision in O’Gilvie v. United States, 519 U. S. 79 (1996) —also ignored by the Court today—is instructive. In that case, the question was whether a provision of the Internal Revenue Code excluding a recovery for personal injury from gross income applied to punitive damages. Well after the critical provision was enacted, Congress adopted an amendment providing that punitive damages for nonphysical injuries were not excluded. Pointing to this amendment, a taxpayer argued: “Why . . . would Congress have enacted this amendment removing punitive damages (in nonphysical injury cases) unless Congress believed that, in the amendment’s absence, punitive damages did fall within the provision’s coverage?” Id., at 89. This argument, of course, is precisely the same as the argument made in this case. To paraphrase O’Gilvie, the Court today asks: Why would Congress have enacted the 1988 amendments, providing safe harbors from three types of disparate-impact claims, unless Congress believed that, in the amendments’ absence, disparate-impact claims did fall within the FHA’s coverage?

The Court rejected the argument in O’Gilvie. “The short answer,” the Court wrote, is that Congress might have simply wanted to “clarify the matter in respect to nonphysical injuries” while otherwise “leav[ing] the law where it found it.” Ibid. Although other aspects of O’Gilvie triggered a dissent, see id., at 94–101 (opinion of Scalia, J.), no one quarreled with this self-evident piece of the Court’s analysis. Nor was the O’Gilvie Court troubled that Congress’ amendment regarding nonphysical injuries turned out to have been unnecessary because punitive damages for any injuries were not excluded all along.

The Court saw the flaw in the argument in O’Gilvie, and the same argument is no better here. It is true that O’Gilvie involved a dry question of tax law while this case involves a controversial civil rights issue. But how we read statutes should not turn on such distinctions.

In sum, as the principal respondent’s attorney candidly admitted, the 1988 amendments did not create disparate-impact liability. See Tr. of Oral Arg. 36 (“[D]id the things that [Congress] actually did in 1988 expand the coverage of the Act? MR. DANIEL: No, Justice”).

C

The principal respondent and the Solicitor General—but not the Court—have one final argument regarding the text of the FHA. They maintain that even if the FHA does not unequivocally authorize disparate-impact suits, it is at least ambiguous enough to permit HUD to adopt that interpretation. Even if the FHA were ambiguous, how-ever, we do not defer “when there is reason to suspect that the agency’s interpretation ‘does not reflect the agency’s fair and considered judgment on the matter in question.’ ” Christopher v. SmithKline Beecham Corp., 567 U. S. ___, ___ (2012) (slip op., at 10).

Here, 43 years after the FHA was enacted and nine days after the Court granted certiorari in Magner (the “rodent infestation” case), HUD proposed “to prohibit housing practices with a discriminatory effect, even where there has been no intent to discriminate.” Implementation of the Fair Housing Act’s Discriminatory Effects Standard, 76 Fed. Reg. 70921 (2011). After Magner settled, the Court called for the views of the Solicitor General in Township of Mount Holly v. Mt. Holly Gardens Citizens in Action, Inc., 568 U. S. ___ (2012), another case raising the same question. Before the Solicitor General filed his brief, however, HUD adopted disparate-impact regulations. See Implementation of the Fair Housing Act’s Discriminatory Effects Standard, 78 Fed. Reg. 11460 (2013). The Solicitor General then urged HUD’s rule as a reason to deny certiorari. We granted certiorari anyway, 570 U. S. ___ (2013), and shortly thereafter Mount Holly also unexpectedly settled. Given this unusual pattern, there is an argument that deference may be unwarranted. Cf. Young v. United Parcel Service, Inc., 575 U. S. ___, ___ (2015) (slip op., at 16–17) (refusing to defer where “[t]he EEOC promulgated its 2014 guidelines only recently, after this Court had granted certiorari” (discussing Skidmore v. Swift & Co., 323 U. S. 134, 140 (1944) )).[8]

There is no need to dwell on these circumstances, however, because deference is inapt for a more familiar reason: The FHA is not ambiguous. The FHA prohibits only disparate treatment, not disparate impact. It is a bedrock rule that an agency can never “rewrite clear statutory terms to suit its own sense of how the statute should operate.” Utility Air Regulatory Group, 573 U. S., at ___ (slip op., at 23). This rule makes even more sense where the agency’s view would open up a deeply disruptive avenue of liability that Congress never contemplated.

IV

Not only does disparate-impact liability run headlong into the text of the FHA, it also is irreconcilable with our precedents. The Court’s decision today reads far too much into Griggs v. Duke Power Co., 401 U. S. 424 (1971) , and far too little into Smith v. City of Jackson, 544 U. S. 228 (2005) . In Smith, the Court explained that the statutory justification for the decision in Griggs depends on language that has no parallel in the FHA. And when the Smith Court addressed a provision that does have such a parallel in the FHA, the Court concluded—unanimously—that it does not authorize disparate-impact liability. The same result should apply here.

A

Rather than focusing on the text of the FHA, much of the Court’s reasoning today turns on Griggs. In Griggs, the Court held that black employees who sued their employer under §703(a)(2) of Title VII of the Civil Rights Act of 1964, 42 U. S. C. §2000e–2(a)(2), could recover without proving that the employer’s conduct—requiring a high school diploma or a qualifying grade on a standardized test as a condition for certain jobs—was motivated by a discriminatory intent. Instead, the Court held that, unless it was proved that the requirements were “job re-lated,” the plaintiffs could recover by showing that the re-quirements “operated to render ineligible a markedly disproportionate number of Negroes.” 401 U. S., at 429.

Griggs was a case in which an intent to discriminate might well have been inferred. The company had “openly discriminated on the basis of race” prior to the date on which the 1964 Civil Rights Act took effect. Id., at 427. Once that date arrived, the company imposed new educational requirements for those wishing to transfer into jobs that were then being performed by white workers who did not meet those requirements. Id., at 427–428. These new hurdles disproportionately burdened African-Americans, who had “long received inferior education in segregated schools.” Id., at 430. Despite all this, the lower courts found that the company lacked discriminatory intent. See id., at 428. By convention, we do not overturn a finding of fact accepted by two lower courts, see, e.g., Rogers v. Lodge, 458 U. S. 613, 623 (1982) ; Blau v. Lehman, 368 U. S. 403 –409 (1962); Graver Tank & Mfg. Co. v. Linde Air Products Co., 336 U. S. 271, 275 (1949) , so the Court was confronted with the question whether Title VII always demands intentional discrimination.

Although Griggs involved a question of statutory interpretation, the body of the Court’s opinion—quite remarkably—does not even cite the provision of Title VII on which the plaintiffs’ claims were based. The only reference to §703(a)(2) of the 1964 Civil Rights Act appears in a single footnote that reproduces the statutory text but makes no effort to explain how it encompasses a disparate-impact claim. See 401 U. S., at 426, n. 1. Instead, the Court based its decision on the “objective” of Title VII, which the Court described as “achiev[ing] equality of employment opportunities and remov[ing] barriers that have operated in the past to favor an identifiable group of white employees over other employees.” Id., at 429–430.

That text-free reasoning caused confusion, see, e.g., Smith, supra, at 261–262 (O’Connor, J., concurring in judgment), and undoubtedly led to the pattern of Court of Appeals decisions in FHA cases upon which the majority now relies. Those lower courts, like the Griggs Court, often made little effort to ground their decisions in the statutory text. For example, in one of the earliest cases in this line, United States v. Black Jack, 508 F. 2d 1179 (CA8 1974), the heart of the court’s analysis was this: “Just as Congress requires ‘the removal of artificial, arbitrary, and unnecessary barriers to employment when the barriers operate invidiously to discriminate on the basis of racial or other impermissible classification,’ such barriers must also give way in the field of housing.” Id., at 1184 (quoting Griggs, supra, at 430–431; citation omitted).

Unlike these lower courts, however, this Court has never interpreted Griggs as imposing a rule that applies to all antidiscrimination statutes. See, e.g., Guardians Assn. v. Civil Serv. Comm’n of New York City, 463 U. S. 582 , n. 27 (1983) (holding that Title VI, 42 U. S. C. §2000d et seq., does “not allow compensatory relief in the absence of proof of discriminatory intent”); Sandoval, 532 U. S., at 280 (similar). Indeed, we have never held that Griggs even establishes a rule for all employment discrimination statutes. In Teamsters, the Court rejected “the Griggs rationale” in evaluating a company’s seniority rules. 431 U. S., at 349–350. And because Griggs was focused on a particular problem, the Court had held that its rule does not apply where, as here, the context is different. In Los Angeles Dept. of Water and Power v. Manhart, 435 U. S. 702 (1978) , for instance, the Court refused to apply Griggs to pensions under the Equal Pay Act of 1963, 29 U. S. C. §206(d) or Title VII, even if a plan has a “disproportion-ately heavy impact on male employees.” 435 U. S. at 711, n. 20. We explained that “[e]ven a completely neutral practice will inevitably have some disproportionate impact on one group or another. Griggs does not imply, and this Court has never held, that discrimination must always be inferred from such consequences.” Ibid.

B

Although the opinion in Griggs did not grapple with the text of the provision at issue, the Court was finally required to face that task in Smith, 544 U. S. 228 , which addressed whether the Age Discrimination in Employment Act of 1967 (ADEA), 29 U. S. C. §621 et seq., authorizes disparate-impact suits. The Court considered two provisions of the ADEA, §§4(a)(1) and 4(a)(2), 29 U. S. C. §§623(a)(1) and (a)(2).

The Court unanimously agreed that the first of these provisions, §4(a)(1), does not authorize disparate-impact claims. See 544 U. S., at 236, n. 6 (plurality opinion); id., at 243 (Scalia, J., concurring in part and concurring in judgment) (agreeing with the plurality’s reasoning); id., at 249 (O’Connor, J., concurring in judgment) (reasoning that this provision “obvious[ly]” does not allow disparate-impact claims).

By contrast, a majority of the Justices found that the terms of §4(a)(2) either clearly authorize disparate-impact claims (the position of the plurality) or at least are ambiguous enough to provide a basis for deferring to such an interpretation by the Equal Employment Opportunity Commission (the position of Justice Scalia). See 544 U. S., at 233–240 (plurality opinion); id., at 243–247 (opinion of Scalia, J.).

In reaching this conclusion, these Justices reasoned that §4(a)(2) of the ADEA was modeled on and is virtually identical to the provision in Griggs, 42 U. S. C. §2000e–2(a)(2). Section 4(a)(2) provides as follows:

“It shall be unlawful for an employer—

.     .     .     .     .

“(2) 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 age.” 29 U. S. C. §623(a) (emphasis added).

The provision of Title VII at issue in Griggs says this:

“It shall be an unlawful employment practice for an employer—

.     .     .     .     .

“(2) to limit, segregate, or classify his employees or applicants for employment 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, color, religion, sex, or national origin.” 42 U. S. C. §2000e–2(a)(2) (emphasis added).

For purposes here, the only relevant difference between these provisions is that the ADEA provision refers to “age” and the Title VII provision refers to “race, color, religion, or national origin.” Because identical language in two statutes having similar purposes should generally be presumed to have the same meaning, the plurality in Smith, echoed by Justice Scalia, saw Griggs as “compelling” support for the conclusion that §4(a)(2) of the ADEA authorizes disparate-impact claims. 544 U. S., at 233–234 (plurality opinion) (citing Northcross v. Board of Ed. of Memphis City Schools, 412 U. S. 427, 428 (1973) (per curiam)).

When it came to the other ADEA provision addressed in Smith, namely, §4(a)(1), the Court unanimously reached the opposite conclusion. Section 4(a)(1) states:

“It shall be unlawful for an employer—

“(1) to fail or refuse to hire or to discharge any individual or otherwise discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s age.” 29 U. S. C. §623(a)(1) (emphasis added).

The plurality opinion’s reasoning, with which Justice Scalia agreed, can be summarized as follows. Under §4(a)(1), the employer must act because of age, and thus must have discriminatory intent. See 544 U. S., at 236, n. 6.[9] Under §4(a)(2), on the other hand, it is enough if the employer’s actions “adversely affect” an individual “because of . . . age.” 29 U. S. C. §623(a).

This analysis of §§4(a)(1) and (a)(2) of the ADEA confirms that the FHA does not allow disparate-impact claims. Sections 804(a) and 805(a) of the FHA resemble §4(a)(1) of the ADEA, which the Smith Court unanimously agreed does not encompass disparate-impact liability. Under these provisions of the FHA, like §4(a)(1) of the ADEA, a defendant must act “because of” race or one of the other prohibited grounds. That is, it is unlawful for a person or entity to “[t]o refuse to sell or rent,” “refuse to negotiate,” “otherwise make unavailable,” etc. for a forbidden reason. These provisions of the FHA, unlike the Title VII provision in Griggs or §4(a)(2) of the ADEA, do not make it unlawful to take an action that happens to adversely affect a person because of race, religion, etc.

The Smith plurality’s analysis, moreover, also depended on other language, unique to the ADEA, declaring that “it shall not be unlawful for an employer ‘to take any action otherwise prohibited . . . where the differentiation is based on reasonable factors other than age.’ ” 544 U. S., at 238 (quoting 81Stat. 603; emphasis added). This “otherwise prohibited” language was key to the plurality opinion’s reading of the statute because it arguably suggesteddisparate-impact liability. See 544 U. S., at 238. This language, moreover, was essential to Justice Scalia’s controlling opinion. Without it, Justice Scalia would have agreed with Justices O’Connor, Kennedy, and Thomas that nothing in the ADEA authorizes disparate-impact suits. See id., at 245–246. In fact, even with this “otherwise prohibited” language, Justice Scalia merely concluded that §4(a)(2) was ambiguous—not that disparate-impacts suits are required. Id., at 243.

The FHA does not contain any phrase like “otherwise prohibited.” Such language certainly is nowhere to be found in §§804(a) and 805(a). And for all the reasons already explained, the 1988 amendments do not presuppose disparate-impact liability. To the contrary, legislative enactments declaring only that certain actions are not grounds for liability do not implicitly create a new theory of liability that all other facets of the statute foreclose.

C

This discussion of our cases refutes any notion that “[t]ogether, Griggs holds[[10]] and the plurality in Smith instructs that antidiscrimination laws must be construed to encompass disparate-impact claims when their text refers to the consequences of actions and not just to the mindset of actors, and where that interpretation is consistent with statutory purpose.” Ante, at 10. The Court stumbles in concluding that §804(a) of the FHA is more like §4(a)(2) of the ADEA than §4(a)(1). The operative language in §4(a)(1) of the ADEA—which, per Smith, does not authorize disparate-impact claims—is materially indistinguishable from the operative language in §804(a) of the FHA.

Even more baffling, neither alone nor in combination do Griggs and Smith support the Court’s conclusion that §805(a) of the FHA allows disparate-impact suits. The action forbidden by that provision is “discriminat[ion] . . . because of” race, religion, etc. 42 U. S. C. §3605(a) (emphasis added). This is precisely the formulation used in §4(a)(1) of the ADEA, which prohibits “discriminat[ion] . . . because of such individual’s age,” 29 U. S. C. §623(a)(1) (emphasis added), and which Smith holds does not authorize disparate-impact claims.

In an effort to explain why §805(a)’s reference to “discrimination” allows disparate-impact suits, the Court argues that in Board of Ed. of City School Dist. of New York v. Harris, 444 U. S. 130 (1979) , “statutory language similar to §805(a) [was construed] to include disparate-impact liability.” Ante, at 11. In fact, the statutory language in Harris was quite different. The law there was §706(d)(1)(B) of the 1972 Emergency School Aid Act, which barred assisting education agencies that “ ‘had in effect any practice, policy, or procedure which results in the disproportionate demotion or dismissal of instructional or other personnel from minority groups in conjunction with desegregation . . . or otherwise engaged in discrimination based upon race, color, or national origin in the hiring, promotion, or assignment of employees.’ ” 444 U. S., at 132–133, 142 (emphasis added).

After stating that the first clause in that unusual statute referred to a “disparate-impact test,” the Harris Court concluded that “a similar standard” should apply to the textually “closely connected” second clause. Id., at 143. This was so, the Court thought, even though the second clause, standing alone, may very well have required discriminatory “intent.” Id., at 139. The Court explained that the Act’s “less than careful draftsmanship” regarding the relationship between the clauses made the “wording of the statute . . . ambiguous” about teacher assignments, thus forcing the Court to “look closely at the structure and context of the statute and to review its legislative history.” Id., at 138–140. It was the combined force of all those markers that persuaded the Court that disparate impact applied to the second clause too.

Harris, in other words, has nothing to do with §805(a) of the FHA. The “wording” is different; the “structure” is different; the “context” is different; and the “legislative history” is different. Id., at 140. Rather than digging up a 36-year-old case that Justices of this Court have cited all of twice, and never once for the proposition offered today, the Court would do well to recall our many cases explaining what the phase “because of” means.

V

Not only is the decision of the Court inconsistent with what the FHA says and our precedents, it will have unfortunate consequences. Disparate-impact liability has very different implications in housing and employment cases.

Disparate impact puts housing authorities in a very difficult position because programs that are designed and implemented to help the poor can provide the grounds for a disparate-impact claim. As Magner shows, when disparate impact is on the table, even a city’s good-faith attempt to remedy deplorable housing conditions can be branded “discriminatory.” 619 F. 3d, at 834. Disparate-impact claims thus threaten “a whole range of tax, welfare, public service, regulatory, and licensing statutes.” Washington v. Davis, 426 U. S. 229, 248 (1976) .

This case illustrates the point. The Texas Department of Housing and Community Affairs (the Department) has only so many tax credits to distribute. If it gives credits for housing in lower income areas, many families—including many minority families—will obtain better housing. That is a good thing. But if the Department gives credits for housing in higher income areas, some of those families will be able to afford to move into more desirable neighborhoods. That is also a good thing. Either path, however, might trigger a disparate-impact suit.[11]

This is not mere speculation. Here, one respondent has sued the Department for not allocating enough credits to higher income areas. See Brief for Respondent Inclusive Communities Project, Inc., 23. But another respondent argues that giving credits to wealthy neighborhoods violates “the moral imperative to improve the substandard and inadequate affordable housing in many of our inner cities.” Reply Brief for Respondent Frazier Revitalization Inc. 1. This latter argument has special force because a city can build more housing where property is least expensive, thus benefiting more people. In fact, federal law often favors projects that revitalize low-income communities. See ante, at 2.

No matter what the Department decides, one of these respondents will be able to bring a disparate-impact case. And if the Department opts to compromise by dividing the credits, both respondents might be able to sue. Congress surely did not mean to put local governments in such a position.

The Solicitor General’s answer to such problems is that HUD will come to the rescue. In particular, HUD regulations provide a defense against disparate-impact liability if a defendant can show that its actions serve “substantial, legitimate, nondiscriminatory interests” that “necessar[ily]” cannot be met by “another practice that has a less discriminatory effect.” 24 CFR §100.500(b) (2014). (There is, of course, no hint of anything like this defense in the text of the FHA. But then, there is no hint of disparate-impact liability in the text of the FHA either.)

The effect of these regulations, not surprisingly, is to confer enormous discretion on HUD—without actually solving the problem. What is a “substantial” interest? Is there a difference between a “legitimate” interest and a “nondiscriminatory” interest? To what degree must an interest be met for a practice to be “necessary”? How are parties and courts to measure “discriminatory effect”?

These questions are not answered by the Court’s assurance that the FHA’s disparate-impact “analysis ‘is analogous to the Title VII requirement that an employer’s interest in an employment practice with a disparate impact be job related.’ ” Ante, at 4 (quoting 78 Fed. Reg. 11470). See also ante, at 18 (likening the defense to “the business necessity standard”). The business-necessity defense is complicated enough in employment cases; what it means when plopped into the housing context is anybody’s guess. What is the FHA analogue of “job related”? Is it “housing related”? But a vast array of municipal decisions affect property values and thus relate (at least indirectly) to housing. And what is the FHA analogue of “business necessity”? “Housing-policy necessity”? What does that mean?

Compounding the problem, the Court proclaims that “governmental entities . . . must not be prevented from achieving legitimate objectives, such as ensuring compliance with health and safety codes.” Ante, at 21. But what does the Court mean by a “legitimate” objective? And does the Court mean to say that there can be no disparate-impact lawsuit if the objective is “legitimate”? That is certainly not the view of the Government, which takes the position that a disparate-impact claim may be brought to challenge actions taken with such worthy objectives as improving housing in poor neighborhoods and making financially sound lending decisions. See Brief for United States as Amicus Curiae 30, n. 7.

Because HUD’s regulations and the Court’s pronouncements are so “hazy,” Central Bank, 511 U. S., at 188–189, courts—lacking expertise in the field of housing policy—may inadvertently harm the very people that the FHA is meant to help. Local governments make countless decisions that may have some disparate impact related to housing. See ante, at 19–20. Certainly Congress did not intend to “engage the federal courts in an endless exercise of second-guessing” local programs. Canton v. Harris, 489 U. S. 378, 392 (1989) .

Even if a city or private entity named in a disparate-impact suit believes that it is likely to prevail if a disparate-impact suit is fully litigated, the costs of litigation, in-cluding the expense of discovery and experts, may “push cost-conscious defendants to settle even anemic cases.” Bell Atlantic Corp. v. Twombly, 550 U. S. 544, 559 (2007) . Defendants may feel compelled to “abandon substantial defenses and . . . pay settlements in order to avoid the expense and risk of going to trial.” Central Bank, supra, at 189. And parties fearful of disparate-impact claims may let race drive their decisionmaking in hopes of avoiding litigation altogether. Cf. Ricci, 557 U. S., at 563. All the while, similar dynamics may drive litigation against private actors. Ante, at 19.

This is not the Fair Housing Act that Congress enacted.

VI

Against all of this, the Court offers several additional counterarguments. None is persuasive.

A

The Court is understandably worried about pretext. No one thinks that those who harm others because of pro-tected characteristics should escape liability by conjuring up neutral excuses. Disparate-treatment liability, however, is attuned to this difficulty. Disparate impact can be evidence of disparate treatment. E.g., Church of Lukumi Babalu Aye, Inc. v. Hialeah, 508 U. S. 520 –542 (1993) (opinion of Kennedy, J.); Hunter v. Underwood, 471 U. S. 222, 233 (1985) . As noted, the facially neutral requirements in Griggs created a strong inference of discriminatory intent. Nearly a half century later, federal judges have decades of experience sniffing out pretext.

B

The Court also stresses that “many of our Nation’s largest cities—entities that are potential defendants in disparate-impact suits—have submitted an amicus brief in this case supporting disparate-impact liability under the FHA.” Ante, at 23–24.

This nod to federalism is puzzling. Only a minority of the States and only a small fraction of the Nation’s municipalities have urged us to hold that the FHA allowsdisparate-impact suits. And even if a majority supported the Court’s position, that would not be a relevant consideration for a court. In any event, nothing prevents States and local government from enacting their own fair housing laws, including laws creating disparate-impact liability. See 42 U. S. C. §3615 (recognizing local authority).

The Court also claims that “[t]he existence of disparate-impact liability in the substantial majority of the Courts of Appeals for the last several decades” has not created “ ‘dire consequences.’ ” Ante, at 24. But the Court concedes that disparate impact can be dangerous. See ante, at 18–22. Compare Magner, 619 F. 3d, at 833–838 (holding that efforts to prevent violations of the housing code may violate the FHA), with 114 Cong. Rec. 2528 (1968) (remarks of Sen. Tydings) (urging enactment of the FHA to help combat violations of the housing code, including “rat problem[s]”). In the Court’s words, it is “paradoxical to construe the FHA to impose onerous costs on actors who encourage revitalizing dilapidated housing.” Ante, at 19. Our say-so, however, will not stop such costly cases from being filed—or from getting past a motion to dismiss (and so into settlement).

C

At last I come to the “purpose” driving the Court’s analysis: The desire to eliminate the “vestiges” of “residential segregation by race.” Ante, at 5, 23. We agree that all Americans should be able “to buy decent houses without discrimination . . . because of the color of their skin.” 114 Cong. Rec. 2533 (remarks of Sen. Tydings) (emphasis added). See 42 U. S. C. §§3604(a), 3605(a) (“because of race”). But this Court has no license to expand the scope of the FHA to beyond what Congress enacted.

When interpreting statutes, “ ‘[w]hat the legislative intention was, can be derived only from the words . . . used; and we cannot speculate beyond the reasonable import of these words.’ ” Nassar, 570 U. S., at ___ (slip op., at 13) (quoting Gardner v. Collins, 2 Pet. 58, 93 (1829)). “[I]t frustrates rather than effectuates legislative intent simplistically to assume that whatever furthers the statute’s primary objective must be the law.” Rodriguez v. United States, 480 U. S. 522, 526 (1987) (per curiam). See also, e.g., Board of Governors, FRS v. Dimension Financial Corp., 474 U. S. 361 –374 (1986) (explaining that “ ‘broad purposes’ ” arguments “ignor[e] the complexity of the problems Congress is called upon to address”).

Here, privileging purpose over text also creates constitutional uncertainty. The Court acknowledges the risk that disparate impact may be used to “perpetuate race-based considerations rather than move beyond them.” Ante, at 21. And it agrees that “racial quotas . . . rais[e] serious constitutional concerns.” Ante, at 20. Yet it still reads the FHA to authorize disparate-impact claims. We should avoid, rather than invite, such “difficult constitutional questions.” Ante, at 22. By any measure, the Court today makes a serious mistake.

*  *  *

I would interpret the Fair Housing Act as written and so would reverse the judgment of the Court of Appeals.