Graoch Associates 33, L. P. v. Louisville/Jefferson County Metro Human Relations Commission

KAREN NELSON MOORE, Circuit Judge,

concurring in part and dissenting in part.

Defendant-Appellant the Louisville/Jefferson County Metro Human Relations Commission (“MHRC”) appeals from the district court’s order denying its motion for summary judgment and granting summary judgment to Plaintiff-Appellee Graoch Associates # 33 (“Graoch”). For the reasons set forth below, I concur in the lead opinion’s conclusion that a landlord’s withdrawal from the Section 8 program may give rise to a disparate-impact claim under the Fair Housing Act (“FHA”), Title VIII of the Civil Rights Act of 1968, codified as amended at 42 U.S.C. §§ 3601-3631. But I differ in the analytical framework that I think this court should apply to FHA disparate-impact claims against private entities. In the instant case, Graoch has conceded that MHRC has established a prima facie case of disparate impact under the FHA. I therefore think this court should reverse the grant of sum*380mary judgment to Graoch. Because the district court offered no reason for denying MHRC’s cross-motion for summary judgment, other than the grant of Graoch’s motion, I believe this court should vacate the part of the district court’s order denying MHRC’s motion for summary judgment. Insufficient evidence exists for us to rule as a matter of law on whether Graoch successfully rebutted MHRC’s pri-ma facie case by showing that withdrawal from the Section 8 program constitutes a business necessity. I therefore believe this court should remand for further proceedings.

I. BACKGROUND

The Section 8 tenant-based housing-assistance program aids low-income families in obtaining housing in the private market. 42 U.S.C. § 1437f(a). Local public-housing agencies receive federal funds to administer the program. 42 U.S.C. § 1437f(b)(l). These agencies issue vouchers to families who then independently find suitable rental housing provided by private owners voluntarily participating in the program. 42 U.S.C. § 1437f(o)(6)(B). The local housing agency pays the housing subsidy directly to the landlord. Participation in the Section 8 program places legal duties on owners related to the safety, sanitation, and lease-terms of the rental housing they provide. 42 U.S.C. § 1437f(d); 42 U.S.C. § 1437f(o)(7)-(8). Public-housing agencies set “the payment standard for each size of dwelling unit in a market area” at not less than ninety percent and not more than 110 percent of “the fair market rental.” 42 U.S.C. § 1437f(o)(1)(B). The monthly assistance for a family is equal to the amount by which the “payment standard” rent determined by the public housing agency exceeds either thirty percent of the monthly adjusted income or ten percent of the monthly income of the family, whichever is greater. A tenant, however, must pay the difference between his or her subsidy and actual rent, no matter the amount of the payment standard rent. 42 U.S.C. § 1437f(o)(2)(A); Alfred M. Clark III, Can America Afford to Abandon a National Housing Policy? 6 J. Afford. Hous. & Cmty. Dev. L. 185, 188 (1997).

The Housing Authority of Jefferson County administers the Section 8 program in the Louisville area. In March 2003, Graoch, the owner of Autumn Run apartments in Louisville, notified the Housing Authority that it would honor existing leases held by Section 8 tenants but would not renew these leases or sign any new Section 8 leases. Joint Appendix (“J.A.”) at 35. Both parties have stipulated the following facts: In 2003, 8,849 tenants in Louisville (including Jefferson County) received Section 8 voucher subsidies, and 6,270 of these tenants were African-Americans. According to the 2000 census, African-American households comprised approximately twenty-four percent and African-Americans comprised approximately nineteen percent of Louisville’s total population (also including Jefferson County). When Graoch withdrew from the Section 8 program, there were eighteen families at Autumn Run receiving Section 8 vouchers, of which seventeen, or approximately ninety-four percent, were African-American families. J.A. at 31-32.

The Kentucky Fair Housing Council, Inc. and three Autumn Run tenants receiving financial assistance through the Section 8 program filed a complaint with MHRC, which subsequently found probable cause that Graoch’s withdrawal constituted unlawful racial discrimination. J.A. at 56-57, 86. MHRC stayed administrative proceedings to allow Graoch to seek a declaratory judgment in the U.S. District Court for the Western District of Ken*381tucky that withdrawal from the Section 8 program cannot constitute a violation of the FHA. J.A. at 65. The current case involves only Graoch and MHRC and arises from the district court’s grant of summary judgment to Graoch.1

II. DISPARATE IMPACT UNDER THE FHA

A. Standard of Review

We must review the district court’s order granting summary judgment de novo. DiCarlo v. Potter, 358 F.3d 408, 414 (6th Cir.2004). “Summary judgment is proper if the evidence, taken in the light most favorable to the nonmoving party, shows that there are no genuine issues of material fact and that the moving party is entitled to a judgment as a matter of law.” Macy v. Hopkins County Sch. Bd., 484 F.3d 357, 363 (6th Cir.2007) (citing Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 587, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986) and Fed.R.Civ.P. 56(c)). We must determine “whether the evidence presents a sufficient disagreement to require submission to a jury or whether it is so one-sided that one party must prevail as a matter of law.” Bryson v. Regis Corp., 498 F.3d 561, 569 (6th Cir.2007) (citing Anderson v. Liberty Lobby Inc., 477 U.S. 242, 256, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986)). Regardless of the evidence put forth by the nonmoving party, the moving party has the burden of establishing that “there is an absence of evidence to support the nonmoving party’s case.” Patmon v. Mich. Supreme Ct., 224 F.3d 504, 508 (6th Cir.2000) (quoting Celotex Corp. v. Catrett, 477 U.S. 317, 325, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986)). Although both sides in this case made cross-motions for summary judgment before the district court, MHRC now appeals the grant of summary judgment to Graoch. Thus, we must view Graoch as the moving party and MHRC as the nonmoving party.

B. Framework for Analysis

Section 3604 of the FHA prohibits discrimination because of, inter alia, race, col- or, or national origin in the sale or rental of housing by (1) refusing to rent or make available any dwelling; (2) offering discriminatory “terms, conditions, or privileges” of rental; (3) making, printing or publishing “any notice, statement, or advertisement” indicating a preference, limitation, or discrimination based on race, color, or national origin; and (4) representing to any person that “any dwelling is not available for ... rental when such dwelling is in fact so available.” 42 U.S.C. § 3604(a)-(d). We have previously held that a plaintiff may establish a violation under the FHA by showing that the defendant has an intent to discriminate (“disparate treatment”) or that an otherwise neutral practice has a disparate impact on a protected class (“disparate impact”). Larkin v. Mich. Dep’t of Soc. Servs., 89 F.3d 285, 289 (6th Cir.1996). Every one of the ten other circuits to consider the issue has allowed disparate-impact analyses under the FHA. John F. Stanton, The Fair Housing Act and Insurance: An Update *382and the Question of Disability Discrimination, 31 Hofstra L.Rev. 141, 174 n. 180 (2002) (listing cases); cf. 2922 Sherman Ave. Tenants’ Ass’n v. District of Columbia, 444 F.3d 673, 679 (D.C.Cir.2006) (assuming without deciding that plaintiffs may bring a disparate-impact claim under the FHA). To prove a disparate-impact claim, a plaintiff need not show any intent to discriminate on the part of the defendant. Bacon v. Honda of America Mfg., Inc., 370 F.3d 565, 576 (6th Cir.2004) (citing Wards Cove Packing Co. v. Atonio, 490 U.S. 642, 657, 109 S.Ct. 2115, 104 L.Ed.2d 733 (1989)), cert. denied, 543 U.S. 1151, 125 S.Ct. 1334, 161 L.Ed.2d 115 (2005); Nationwide Mut. Ins.Co. v. Cisneros, 52 F.3d 1351, 1362 (6th Cir.1995), cert. denied, 516 U.S. 1140, 116 S.Ct. 973, 133 L.Ed.2d 893 (1996).

The appropriate framework for analyzing an FHA disparate-impact claim against a private entity is an issue of first impression in this circuit. I believe that we should analyze these claims according to a framework derived from that which the Fourth Circuit uses to evaluate FHA disparate-impact claims against private defendants as well as that which we use to evaluate Title VII disparate-impact claims. First, we should require a plaintiff to establish a prima facie case by showing that a challenged housing practice or policy has caused a disparate effect on a protected class. Second, if a plaintiff has established a prima facie case, the defendant can avoid liability by proving that the challenged practice constituted a business necessity and that there existed no less discriminatory alternatives that could serve the business interest.

1. FHA Disparate-Impact Claims Against Government Defendants

In deriving this standard, I have found it helpful to survey how other circuits treat FHA disparate-impact claims against both government and private entities. The circuits evaluate the claim either by applying a multi-factor test, implementing a burden-shifting framework, or combining the two approaches. In Metropolitan Housing Development Corp. v. Village of Arlington Heights (Arlington Heights II), 558 F.2d 1283 (7th Cir.1977), cert. denied, 434 U.S. 1025, 98 S.Ct. 752, 54 L.Ed.2d 772 (1978), the first court of appeals case concerning disparate impact under the FHA, the Seventh Circuit stated: “we refuse to conclude that every action which produces discriminatory effects is illegal.” Id. at 1290.2 The Seventh Circuit accordingly set forth a four-factor test to determine whether an action producing a disparate effect violates the FHA. Its factors inquire:

(1) how strong is the plaintiffs showing of discriminatory effect; (2) is there some evidence of discriminatory intent, though not enough to satisfy the constitutional standard of Washington v. Davis [426 U.S. 229, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976)]; (3) what is the defendant’s interest in taking the action complained of; and (4) does the plaintiff seek to compel the defendant to affirmatively provide housing for members of minority groups or merely to restrain *383the defendant from interfering with individual property owners who wish to provide such housing.

Id. Five years later, the Fourth Circuit adopted the same four factors in considering an FHA disparate-impact claim against a government entity. See Smith v. Town of Clarkton, 682 F.2d 1055, 1065 (4th Cir.1982). The Sixth Circuit subsequently applied the Arlington Heights II analysis in a case involving a defendant government entity, but eliminated the assessment of evidence of discriminatory intent because it agreed with the Seventh Circuit that this was “the least important of the four factors.” Arthur v. City of Toledo, 782 F.2d 565, 575 (6th Cir.1986) (citing Arlington Heights II, 558 F.2d at 1292). See also Buckeye Cmty. Hope Found. v. City of Cuyahoga Falls, 263 F.3d 627, 640-41 (6th Cir.2001) (evaluating the factors articulated in Arthur to conclude that there existed genuine issues of material fact concerning defendants’ liability with respect to a disparate-impact claim under the FHA), vacated on abandonment of claim, 538 U.S. 188, 199-200, 123 S.Ct. 1389, 155 L.Ed.2d 349 (2003).

In response to confusion about the stage of litigation at which the Arlington Heights II factors came into play, the Second Circuit clarified that courts should consider the factors as part of a “final determination on the merits rather than as a requirement for a prima facie case.” Huntington Branch, NAACP v. Town of Huntington, 844 F.2d 926, 935 (2d Cir.), aff'd in part, 488 U.S. 15, 109 S.Ct. 276, 102 L.Ed.2d 180 (1988). The Second Circuit reasoned: “treating the four factors as steps necessary to make out a prima facie case places too onerous a burden on [plaintiffs].” Id. at 935-36. Drawing on earlier Third Circuit precedent,3 the Second Circuit established a burden-shifting framework, which considered the Arlington Heights II factors “as part of a defendant’s justification for its challenged action.” Id. at 935. The Second Circuit held that plaintiffs should first establish a prima fa-cie case “by showing that the challenged practice of the defendant ‘actually or predictably results in racial discrimination; in other words that it has a discriminatory effect.’ ” Id. at 934 (quoting United States v. City of Black Jack, 508 F.2d 1179, 1184-85 (8th Cir.1974), cert. denied, 422 U.S. 1042, 95 S.Ct. 2656, 45 L.Ed.2d 694 (1975)). “The plaintiff need not show that the decision complained of was made with discriminatory intent.” Id. Once a plaintiff has presented a prima facie case, a defendant “must prove that its actions furthered, in theory and in practice, a legitimate, bona fide governmental interest and that no alternative would serve that interest with less discriminatory effect.” Id. at 936 (citing Resident Advisory Bd. v. Rizzo, 564 F.2d 126, 148-49 (3d Cir.1977), cert. denied, 435 U.S. 908, 98 S.Ct. 1457, 1458, 55 L.Ed.2d 499 (1978)).

The Tenth Circuit takes the same approach as Huntington Branch in establishing a burden-shifting framework and then weighing the government defendant’s justifications against the plaintiffs showing of disparate impact. Reinhart v. Lincoln County, 482 F.3d 1225, 1229 (10th Cir.2007). Although the Fifth Circuit has not had occasion to develop fully the framework for assessing FHA disparate-impact *384claims against government entities,4 at least one district court within the Fifth Circuit follows the Huntington Branch analysis. In Dews v. Town of Sunnyvale, 109 F.Supp.2d 526, 531-32, 565 (N.D.Tex.2000), the U.S. District Court for the Northern District of Texas held that once a plaintiff has established a prima facie case of discriminatory effects, the defendant must prove that it acted to further a “legitimate, bona fide governmental interest” as well as prove the absence of a less discriminatory alternative. The court then evaluates the disparate-impact claim by balancing the showing of discriminatory effects against that of the defendant’s justifications and by assessing whether the plaintiff seeks to compel or to prevent interference with housing provision. Id.

The other circuits that have addressed FHA disparate-impact claims against government defendants have not followed Huntington Branch’s framework for incorporating the Arlington II factors into an evaluation of the merits of the defendant’s justification for the challenged action. The First, Third, and Eighth Circuits evaluate FHA disparate-impact claims against government entities according to burden-shifting frameworks and do not employ any final balancing test derived from the Arlington Heights II factors to evaluate the merits of the defendant’s justification. The burden-shifting standards employed by these circuits are similar, with the only significant variation being that the Eighth Circuit requires plaintiffs to prove the existence of a less discriminatory alternative, while the Third Circuit requires defendants to prove the absence of such an alternative; the First Circuit has not yet determined the allocation of this burden. See Darst-Webbe Tenant Ass’n Bd. v. St. Louis Hous. Auth., 417 F.3d 898, 902-03 (8th Cir.2005) (holding that once the plaintiff has established a prima facie case, the defendant “must demonstrate that the proposed action has ‘a manifest relationship’ to the legitimate, non-discriminatory policy objectives and ‘is justifiable on the ground it is necessary to’ the attainment of these objectives”; the burden then “shifts back to the plaintiffs to show that a viable alternative means is available to achieve those legitimate policy objectives without discriminatory effects”) (quoting Oti Kaga, Inc. v. South Dakota Hous. Dev. Auth., 342 F.3d 871, 883 (8th Cir.2003)); Lapid-Laurel, LLC v. Zoning Bd. of Adjustment, 284 F.3d 442, 467 (3rd Cir.2002) (holding that once a plaintiff has made out a prima facie case of disparate impact, “then the burden shifts to the defendant to show that it had a legitimate, non-discriminatory reason for the action and that no less discriminatory alternatives were available”); Langlois v. Abington Hous. Auth., 207 F.3d 43, 49-51 (1st Cir.2000) (finding that “practically all of the case law ... treats impact as doing no more than creating a prima facie case, forcing the defendant to proffer a valid justification” and that further balancing of objectives by judges is inappropriate in disparate-impact cases).

The Ninth and Eleventh Circuits have allowed FHA disparate-impact claims against government entities but have not *385specified whether they incorporate the Arlington Heights II multi-factor test into a burden-shifting framework. Gamble v. City of Escondido, 104 F.3d 300, 306 (9th Cir.1997); Jackson v. Okaloosa County, 21 F.3d 1531, 1543 (11th Cir.1994) (citing United States v. Mitchell, 580 F.2d 789, 791 (5th Cir.1978)); Keith v. Volpe, 858 F.2d 467, 483-84 (9th Cir.1988), cert. denied, 493 U.S. 813, 110 S.Ct. 61, 107 L.Ed.2d 28 (1989).

2. FHA Disparate-Impact Claims Against Private Defendants

Plaintiffs have brought to the federal courts far fewer FHA disparate-impact claims against private entities than they have against public entities. The Fifth Circuit acknowledges the availability of a disparate-impact claim against a private defendant under the FHA, but has not developed a framework for analyzing these claims. See Simms v. First Gibraltar Bank, 83 F.3d 1546, 1555 (5th Cir.) (citing Hanson v. Veterans Admin., 800 F.2d 1381, 1386 (5th Cir.1986)), cert. denied, 519 U.S. 1041, 117 S.Ct. 610, 136 L.Ed.2d 535 (1996). The Fourth and Tenth Circuits, however, have developed contrasting frameworks for evaluating FHA disparate-impact claims against private entities.5 The Fourth Circuit has applied a simple burden-shifting framework, holding that “when confronted with a showing of discriminatory impact, defendants must prove a business-necessity sufficiently compelling to justify the challenged practice.” Betsey v. Turtle Creek Assocs., 736 F.2d 983, 988 (4th Cir.1984). In applying the business-necessity standard to disparate-impact claims against private defendants under the FHA, the Fourth Circuit followed the framework for parallel claims under Title VII.

The Tenth Circuit, however, created a different framework for evaluating FHA disparate-impact claims against private defendants that departs from the business-necessity test. Instead, the Tenth Circuit’s evaluation of disparate-impact claims against private defendants follows the framework that the Sixth Circuit uses to evaluate disparate-impact claims against public defendants. In Mountain Side Mobile Estates Partnership v. Secretary of Housing & Urban Development, 56 F.3d 1243, 1252 (10th Cir.1995), the Tenth Circuit held that “a Title VIII prima fade ease, once established ... could alone suffice to prove a Title VIII violation unless the defendants justify the discriminatory effect which has resulted from their challenged actions.” The opinion reiterated that “unrebutted proof of discriminatory effect alone may justify a federal equitable response.” Id. To determine “whether a plaintiffs prima facie case of disparate impact makes out a violation of Title VIII,” the Tenth Circuit assesses the three factors we set forth in Arthur: the strength of the showing of discriminatory effect; the defendant’s interest in taking the action; and whether the plaintiff seeks to compel the defendant to provide housing or alternatively to restrain the defendant from interfering with the provision of housing. Id.

3. Sixth Circuit Treatment of Title VII Disparate-Impact Claims

In deciding what framework to apply in the instant case, the lead opinion does not look to other circuit courts’ treatment of FHA disparate-impact claims for guidance. *386Instead, the lead opinion integrates the three-factor analysis outlined in Arthur v. City of Toledo with the tripartite burden-shifting approach taken in a single Sixth Circuit Title VII disparate-impact case, Kovacevich v. Kent State Univ., 224 F.3d 806 (6th Cir.2000). In particular, the lead opinion suggests that the first two factors from Arthur — the strength of the showing of disparate effects and the strength of the defendant’s interest in the challenged practice — should be applied at the pretext stage of analysis. I part company with the lead opinion on two grounds. First, I believe that the appropriate burden-shifting framework for a disparate-impact case is found not in Kovacevich but in an earlier Sixth Circuit decision, Alexander v. Local 496, Laborers’ International Union of North America, 177 F.3d 394 (6th Cir.1999), cert. denied, 528 U.S. 1154, 120 S.Ct. 1158, 145 L.Ed.2d 1070 (2000). Second, I believe that the lead opinion improperly applies the factors outlined in Arthur v. City of Toledo to an FHA disparate-impact claim involving a private defendant.

The lead opinion is correct in looking to disparate-impact claims against private defendants under Title VII to derive a framework for assessing such claims under the FHA. See Larkin, 89 F.3d at 289. The Sixth Circuit’s cases on disparate-impact claims under Title VII, however, conflict. Kovacevich holds that a once a plaintiff has shown an adverse effect, the burden shifts to the defendants to “articulate a legitimate business reason” for the challenged practice. 224 F.3d at 830. Once the defendant has proffered such a reason, under Kovacevich “the burden shifts back to the plaintiff to show either that the [defendant’s] reason is a pretext for discrimination, or that there exists an alternative ... practice that would achieve the same business ends with a less discrimina,-tory impact.” Id. In contrast, Alexander holds that once a plaintiff has established a prima facie case by showing adverse effect, “the burden shifts to the [defendant] to produce evidence that the challenged practice is a business necessity.” 177 F.3d at 405-06. The difference between the two opinions lies in the fact that by requiring the plaintiff to prove pretext, Kovacevich both imports an intent-oriented principle from the disparate-treatment test into the disparate-impact analysis and places the ultimate burden of persuasion on the plaintiff rather than on the defendant.

Alexander offers the better standard for assessing disparate-impact claims against private entities. Moreover, Alexander is the earlier opinion, and “when a later decision of this court conflicts with one of our prior published decisions, we are still bound by the holding of the earlier case.” Darrah v. City of Oak Park, 255 F.3d 301, 310 (6th Cir.2001) (citing Sowards v. Loudon County, 203 F.3d 426, 431 n. 1 (6th Cir.), cert. denied, 531 U.S. 875, 121 S.Ct. 179, 148 L.Ed.2d 123 (2000), and Brentwood Acad. v. Tenn. Secondary Sch. Athletic Ass’n, 180 F.3d 758, 765 (6th Cir.1999), rev’d on other grounds, 531 U.S. 288, 121 S.Ct. 924, 148 L.Ed.2d 807 (2001)). Pretext does not belong in a disparate-impact case as a matter of principle. As a general matter, the plaintiffs burden to prove pretext involves a showing that the defendant’s proffered justification for the challenged action masks actual discriminatory intent. Amini v. Oberlin Coll., 440 F.3d 350, 359-60 (6th Cir.2006). Because a plaintiff does not need to establish discriminatory intent in a disparate-impact case, proof of pretext in such a case would be an unnecessary and indeed absurd undertaking.

Third, the genealogy of the pretext prong in Kovacevich renders it suspect. Kovacevich devised the requirement that a plaintiff prove pretext in a Title VII disparate-impact claim from an earlier Sixth *387Circuit Title VII disparate-impact case, Scales v. J.C. Bradford & Co., 925 F.2d 901 (6th Cir.1991). Scales, in turn, derived the pretext requirement from a Supreme Court opinion, Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 108 S.Ct. 2777, 101 L.Ed.2d 827 (1988). The relevant portion of Watson, however, does not make a plaintiffs proof of pretext a third stage of the burden-shifting in disparate-impact cases. Instead, the opinion merely observes that the factors used to assess a plaintiffs showing of a less discriminatory alternative in disparate-impact cases are also ones that determine whether “the challenged practice has operated as the functional equivalent of a pretext for discriminatory treatment.” Watson, 487 U.S. at 998, 108 S.Ct. 2777 (emphasis added).

Moreover, the section of Watson cited by Scales for the proposition that a plaintiff retains the burden of proof in a disparate-impact claim was issued by a plurality rather than a majority of the Court and is therefore not binding. In this section of the Watson opinion joined only by three other Justices, Justice O’Connor wrote that the plaintiff retains the burden of persuasion at each stage of the burden-shifting inquiry in a disparate-impact case. Id. at 997, 108 S.Ct. 2777. Justice White’s majority opinion in Wards Cove, 490 U.S. at 660, 109 S.Ct. 2115, adopted the position taken by the O’Connor plurality in Watson. The Civil Rights Act of 1991, § 105, 105 Stat. 1074-1075 (1994) (codified as amended at 42 U.S.C. § 2000e-2(k)), however, rejected Wards Cove’s holding regarding the allocation of burdens of production and proof in disparate-impact cases. The interpretive memorandum for the Act, which Congress declared to be its exclusive legislative history, states: “The terms ‘business necessity’ and ‘job-related’ are intended to reflect the concepts enunciated by the Supreme Court in Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971), and in the other Supreme Court decisions prior to Wards Cove Packing Co. v. Atonio.” 137 Cong. Rec. 28,680 (1991). We have clearly interpreted the Civil Rights Act of 1991 to mandate that once a plaintiff has established a prima facie case of disparate effects, “the burden of persuasion, not just the burden of production, shifts to the defendant to show a business necessity for the practices with disparate impact.” Phillips v. Cohen, 400 F.3d 388, 398 (6th Cir.2005) (citing 42 U.S.C. § 2000e-2(k)(1)(A)).

4. Burdens of Proof and the Business-Necessity Defense

“Business necessity” is similarly the appropriate standard for private entities’ defense against disparate-impact claims under the FHA. See 2922 Sherman Ave. Tenants’ Ass’n, 444 F.3d at 679 (observing that Griggs is the “seminal case” on the issue of disparate-impact claims and that, in the employment context, Griggs establishes “business necessity” as the sole defense to a showing of disparate effects). The Secretary of Housing and Urban Development (“HUD”) takes the position that implementing regulations promulgated in 1989 established a business-necessity defense to disparate-impact claims. The Secretary of HUD’s Regulation of the Federal National Mortgage Association (Fannie Mae) and the Federal Home Loan Mortgage Corporation, 60 Fed.Reg. at 61,-866-61,868 (Dec. 1, 1995) (codified at 24 C.F.R. Part 81).

“Business necessity” holds defendants to a higher standard than the more lenient “business justifications” test set forth in Wards Cove, 490 U.S. at 659, 109 S.Ct. 2115, which defined “the dispositive issue [to be] whether a challenged practice serves, in a significant way, the legitimate [business] goals of the [defendant].” Be*388tween 1989 and 1991, some courts applied the Wards Cove standard to FHA cases. Cases decided after the passage of the Civil Rights Act of 1991, however, implement the “business necessity” test in recognition that the Act overruled the weakened disparate-impact standards set forth in Wards Cove. Stanton, The Fair Housing Act and Insurance, 31 Hofstra L.Rev. at 186-87. While a consensus exists that business necessity is the appropriate test, there remains some disagreement as to the content of that standard. The Fourth Circuit, HUD, and district courts within the Ninth Circuit define business necessity to mean a “compelling” business interest; in their allocation of the burdens of proof, the Second and Third Circuit’s apply a “substantive equivalent” of this test.6 Id. at 187. I agree that “when confronted with a showing of discriminatory impact, defendants must prove a business necessity sufficiently compelling to justify the challenged practice.” See Betsey, 736 F.2d at 988. Requiring the defendant to prove anything less would contravene the plain meaning of the word “necessity,” defined as “an imperative requirement or need for something.” Random House Dictionary 1284 (Stuart Berg Flexner, ed., 2d ed.1993). Moreover, requiring the defendant to offer a mere business justification would confuse defendants’ burden of proving business-necessity in disparate-impact cases with defendants’ burden of producing a legitimate, nondiscriminatory reason for the challenged action in disparate-treatment cases. Compare Macy, 484 F.3d at 364 (discussing the shifting burdens in disparate-treatment cases), with Alexander, 177 F.3d at 405-06 (discussing the shifting burdens in disparate-impact cases).

In articulating the framework for our evaluation of FHA disparate-impact claims against private defendants, I believe that we should rely on our earlier precedent Alexander, rather than Kovacevich, and seek guidance as well from the Fourth Circuit decision in Betsey v. Turtle Creek. Thus, we should first require that plaintiffs establish a prima facie case of disparate effects. Second, upon the showing of a prima facie case, the burden of persuasion then shifts to the defendants, who must prove that the challenged housing practice constitutes a business necessity and that no less discriminatory alternatives are available. In the Title VII context, plaintiffs bear the burden of proving that a less discriminatory alternative exists. Isabel v. City of Memphis, 404 F.3d 404, 411 (6th Cir.2005) (citing Albemarle Paper Co. v. Moody, 422 U.S. 405, 425, 432, 95 S.Ct. 2362, 45 L.Ed.2d 280 (1975)). The consensus across the circuits, however, is that in the housing context defendants have the burden of proving the absence of a less discriminatory alternative. See Lapid-Laurel, LLC, 284 F.3d at 467 (discussing the defendant’s burden of proof regarding the absence of less discriminatory alternatives); Langlois, 207 F.3d at 50 (same); Huntington Branch, NAACP, 844 F.2d at 935 (same); Dews, 109 F.Supp.2d at 531-32, 565 (same); see also Betsey, 736 F.2d at 988 (holding that once plaintiffs have established a prima facie case, defendants must prove business necessity).

Our analysis should stop with an evaluation of whether a private defendant has *389successfully rebutted the plaintiffs showing of disparate effects by proving “business necessity” and the unavailability of a less discriminatory alternative. Considerations of doctrine and policy preclude this court’s use of the Arthur factors for a government defendant to evaluate an FHA disparate-impact claim against a private defendant. First, the lead opinion applies factors from a case concerning a public defendant to cases against private defendants, without explaining why they are transferable. That the lead opinion needed to excise the third Arthur factor from its analysis only highlights the unsuitability of the framework in the context of a case against a private defendant. I believe that the Tenth Circuit incorrectly integrated Arthur’s multi-factor test into the “business necessity” framework applicable to private defendants in disparate-impact suits; it did so without even acknowledging that the case concerned a private defendant while Arthur concerned a public defendant. See Mountain Side, 56 F.3d at 1252. In contrast, the Fourth Circuit has explicitly stated that the multi-factor analysis derived from Arlington Heights II and its progeny should not apply in disparate-impact claims against private defendants. Betsey, 736 F.2d at 988 n. 5.

Second, by applying the Arthur factors, the lead opinion converts an inquiry into the business justification for a practice bearing a disparate impact on a protected class into a balancing test. The lead opinion suggests that this court should weigh the disparate effects of the challenged action against the strength of the defendant’s interest in taking the challenged action. Courts appropriately countenance governments’ weighing of competing public considerations of deep importance, and they may engage in such balancing acts themselves. See, e.g., Arthur, 782 F.2d at 575 (holding that because of “the strong policy considerations underlying referendums ... absent highly unusual circumstances, the discriminatory effect of a referendum cannot establish a violation of the Fair Housing Act”). In limited instances, Congress has reduced the burden of fair housing laws on government entities, with respect to those entities’ pursuit of important public interests, but Congress did not do the same for private landlords. See Pfaff v. U.S. Dep’t of Hous. & Urban Dev., 88 F.3d 739, 746 (9th Cir.1996) (holding that “Congress chose to give special deference to government-imposed occupancy limits only ... [and] made no comparable provision for private occupancy policies”). Even if weighing relative strengths of interests were appropriate, we must follow the longstanding rule that the weighing of litigants’ competing interests belongs at trial and not at the summary judgment stage. Crestview Parke Care Ctr. v. Thompson, 373 F.3d 743, 755 (6th Cir.2004) (“In evaluating whether summary judgment is proper, we do not weigh the evidence, but rather view the evidence in the light most favorable to [the non-moving party] to divine the existence of a genuine dispute of material fact.”)

Third, the lead opinion’s framework mistakenly places the ultimate burden of persuasion on the plaintiff. The lead opinion states that the final step in the burden-shifting analysis involves evaluating the plaintiffs showing by considering “the strength of the plaintiffs showing of discriminatory effect against the strength of the defendant’s interest in taking the challenged action.” Lead Op. at 374. Under the business-necessity test, however, the landlord will successfully defend against a showing of disparate effects only if he or she proves that the challenged action constitutes a business necessity. If the landlord fails to carry this burden of proof, then he or she should lose once the plaintiff has established a statistically signifi*390cant showing of disparate effects regardless of its strength. Conversely, if the defendant does indeed prove business necessity, he or she should not be liable under a disparate-impact theory of the FHA no matter how strong the showing of disparate effects.

In conclusion, we are bound by the factors set forth in Arthur when evaluating FHA disparate-impact claims against government defendants. But the framework outlined in Arthur is not well suited to the evaluation of FHA disparate-impact claims against private defendants. The burden-shifting framework for FHA disparate-impact claims against private defendants is straightforward: first, the plaintiff establishes a prima facie case by showing that a challenged housing practice or policy has caused a statistically significant disparate effect on a protected group; second, the housing owner can assert the affirmative defense that the challenged action constitutes a business necessity and that there are no less discriminatory alternatives.

C. Withdrawal from the Section 8 Program and Disparate Impact

I concur in the lead opinion’s holding that a landlord’s withdrawal from the Section 8 program, when it has a disparate impact on members of a protected class, may violate the FHA. Courts should give the “broad and inclusive” language of the FHA a “generous construction” to effectuate its remedial purpose. Trafficante v. Metro. Life Ins. Co. 409 U.S. 205, 209, 212, 93 S.Ct. 364, 34 L.Ed.2d 415 (1972).

No compelling jurisprudential or policy reasons exist to make a categorical exception to our recognition of disparate-impact claims under the FHA for the withdrawal by landlords from the Section 8 program. Graoch argues that the voluntary nature of the Section 8 program should counsel against allowing a plaintiff to bring a disparate-impact claim based on a landlord’s decision to withdraw. Appellee Br. at 11-14. Graoch derives this analysis from the Seventh Circuit opinion in Knapp v. Eagle Property Management Corp., 54 F.3d 1272 (7th Cir.1995), which held that the refusal by a landlord who participated in the Section 8 program to accept a new Section 8 tenant could not give rise to liability under a disparate-impact analysis of the FHA. Id. at 1280. The Seventh Circuit found no principled basis upon which to distinguish between a landlord’s decision not to participate in the Section 8 program and the decision of a participating housing owner not to accept a new Section 8 tenant. Id. Accordingly, the Seventh Circuit held that because nonparticipation is not actionable, neither can a landlord’s withdrawal from the program give rise to a disparate-impact claim. Id. (“The actions of both nonparticipating and participating owners have the same impact on minorities and to hold only the latter liable for racial discrimination for that conduct would deter them from joining or remaining involved in the program.”)

The voluntary nature of the Section 8 program does caution us to consider carefully whether a landlord’s decision to withdraw from the program may give rise to liability under the FHA. Congress’s revision of the FHA has only reinforced that body’s intent to make Section 8 a voluntary program. Before 1996, some courts interpreted 42 U.S.C. § 1437f(d)(l)(B), which provided “that the owner shall not terminate the tenancy except for serious or repeated violation of the terms and conditions of the lease, for violation of applicable Federal, State, or local law, or for other good cause” to prescribe an “endless lease.” Some courts read this provision to mean that the owner could not refuse to renew a lease at the conclusion of the lease period, except for the specified *391reasons. 30 Eastchester LLC v. Healy, Index SP-2002-77, 2002 WL 553709, at *3 (N.Y.City Ct.2002) (unpublished opinion) (citing Salute v. Stratford Greens Garden Apartments, 136 F.3d 293, 300 n. 5 (2d Cir.1998); Tann Realty Co. v. Thompson, 112 Misc.2d 392, 446 N.Y.S.2d 959 (Civ.Ct., Kings Co.1981)). In 1996, however, Congress amended this section of the FHA as part of the Omnibus Consolidated Rescis-sions and Appropriations Act of 1996, Pub.L. No. 104-134, § 203, 110 Stat. 1321, at *1321-281 (1996), by inserting “during the term of the lease” into both clause (ii) and clause (hi) of the section. “The effect of the statutory change was to eliminate the ‘endless lease’ interpretation of such provision and clarify that a landlord could terminate a Section 8 tenant’s lease and, a fortiori, its participation in the Section 8 program with respect to that tenant, when the term of the tenant’s lease expires.” 30 Eastchester, 2002 WL 553709, at *4.

The same Omnibus Budget Act also repealed the “take one, take ah” provision of the United States Housing Act of 1937. Omnibus Consolidated Rescissions and Appropriations Act § 203. That provision, previously codified at 42 U.S.C. § 1437f(t)(l)(A) was entitled “Nondiscrimination Against Certificate Holders and Voucher Holders” and provided that no owner who had entered into a contract for housing assistance payments under the section “shall refuse to lease any available dwelling unit in any multifamily housing project of such owner to a holder of a voucher.” Glover v. Crestwood Lake Section 1 Holding Corps., 746 F.Supp. 301, 308 (S.D.N.Y.1990).

Ultimately, however, I disagree with the Seventh Circuit’s reasoning that the voluntary character of the Section 8 program precludes liability for withdrawal from the program under a disparate-impact theory of the FHA. The repeal of the “endless lease” and “take one, take all” provisions of 42 U.S.C. § 1437f evinced Congress’ intention that a landlord’s decision to participate in the Section 8 program should not require him or her to accept any particular Section 8 tenant or to renew endlessly the lease of a given tenant. The statutory repeals, however, did not address the issue of a complete withdrawal from the program and a decision not to accept any new Section 8 tenants. The FHA imposes a non-discrimination mandate on actors engaging in many different types of voluntary actions. See, e.g., Betsey, 736 F.2d at 986-87 (allowing a disparate-impact claim to proceed where a private landlord issued eviction notices to famihes with children in an effort to institute an all-adult rental policy). Landlords who choose not to accept any Section 8 vouchers may do so for all kinds of legitimate or biased reasons that remain behind the veil of nonparticipation, far removed from legal scrutiny. Once a landlord decides to enroll in the Section 8 program, however, he or she necessarily accepts a new set of responsibilities toward tenants under the FHA. These include duties both to refrain from intentional discrimination on the basis of a protected characteristic and, in the case of a business practice that has a disparate impact on a protected class, to implement when possible an alternative business practice with less discriminatory effects. See Alexander, 177 F.3d at 405-06. I agree with the lead opinion that the Sixth Circuit should decline to follow the Seventh Circuit and should instead allow plaintiffs to use evidence of disparate impact to show that a landlord violated the FHA by withdrawing from the Section 8 program.

In addition to Knapp, Graoch relies on the Second Circuit’s decision in Salute v. Stratford Greens Garden Apartments, 136 F.3d 293 (2d Cir.1998), to argue against the propriety of allowing a disparate-im*392pact claim for withdrawal from the Section 8 program. Appellee Br. at 11-12. In Salute, however, the majority based its opinion on the fact that the defendant landlord had never accepted tenants who, at the time of application, participated in the Section 8 program. Id. at 302 (“[N]on-participation constitutes a legitimate reason for their refusal to accept section 8 tenants and ... we therefore cannot hold them liable for ... discrimination under the disparate impact theory.”) (quoting Knapp, 54 F.3d at 1280) (alterations in the original). Therefore, Salute is not applicable in the instant case because of the distinction between nonparticipation, at issue in Salute, and withdrawal from the Section 8 program, at issue here.7

Graoch further suggests that allowing disparate-impact claims based on a landlord’s withdrawal from the Section 8 program will deter landlords from entering the program. Appellee Br. at 13 (“If a landlord is held to a perpetual lease — and all of the administrative burdens which Section 8 compliance entails — once it accepts a Section 8 tenant, many landlords may never agree to enter into the program at all, thereby decreasing the available private housing stock for low income tenants.”) (quoting 30 Eastchester LLC, 2002 WL 553709, at *4). The Eastchester case upon which Graoch relies, however, did not involve a disparate-impact claim under the FHA. Instead, a tenant-plaintiff claimed that her landlord’s allegedly illegal withdrawal from the Section 8 program, at the termination of her lease, constituted a defense to her own nonpayment of rent. Id. at *2. Allowing disparate-impact claims for withdrawal does not yoke landlords to a “perpetual lease.” A plaintiffs showing that a withdrawal from the Section 8 program has a disparate effect on a protected class merely establishes a plaintiffs prima facie case and does not automatically result in liability. Such claims require only that a landlord show business necessity. See Green v. Sunpointe Assocs., Ltd., No. C96-1542C, 1997 WL 1526484, at *4 (W.D.Wash.1997).

Graoch argues that a landlord would incur significant expense defending itself against a disparate-impact claim at the summary-judgment stage. Appellee Br. at 16. But the Supreme Court has not shied away from allowing innovative disparate-impact claims even though newly-recognized causes of action may potentially impose costs on businesses. See, e.g., Smith v. City of Jackson, 544 U.S. 228, 232, 125 S.Ct. 1536, 161 L.Ed.2d 410 (2005) (holding that a disparate-impact theory is cognizable under the Age Discrimination in Employment Act). Landlords have a considerable financial incentive to accept Section 8 tenants because the federal government’s subsidy of a portion of the market-based rent expands the pool of available tenants. Clark, 6 J. Afford. Hous. & Cmty. Dev. L. at 187. This economic reality makes it less likely that the potential of disparate-impact claims upon withdrawal from Section 8 will deter landlords from entering the program.

III. MHRC’S DISPARATE-IMPACT CLAIM

A. Prima Facie Case

Because the district court ruled as a matter of law that a landlord’s withdrawal *393from the Section 8 program cannot alone constitute a violation of the FHA, it did not discuss the statistical evidence that MHRC would need to produce to establish a prima facie case of disparate impact. Graoch appears to have admitted, however, “that the uncontested statistical evidence would in theory support a ‘disparate impact’ claim against Graoch if ... Graoch’s mere withdrawal from the voluntary Section 8 program could subject it to such a claim ... and Graoch’s business necessity defense were to be rejected by the Court....” R. 28, Graoch’s Response to MHRC’s Cross-Motion for Summary Judgment, p. 2 n. 2. Because all members of this panel are in agreement that a landlord’s withdrawal from the Section 8 program can give rise to a disparate-impact claim, Graoch’s admission obligates this court to hold that MHRC has established a prima facie showing of disparate effects.8 As a result of Graoch’s concession in the district court, we need not engage in any further assessment of MHRC’s statistical evidence. In the absence of disagreement and briefing on this issue, I think it is unwise to set forth any framework for evaluating the statistical evidence necessary to prove disparate effects.

B. Business Necessity

Because Graoch has conceded that its withdrawal from the Section 8 program would have a disparate effect on African-Americans, to avoid liability under a disparate-impact theory of the FHA Graoch needs to prove that the withdrawal amounted to a business necessity and that no less discriminatory alternatives were available. Graoch stated that it decided to withdraw from the Section 8 program because participation had required that Graoch agree to a Housing Assistance Payments Contract “which imposed terms and conditions more burdensome than those contained in Autumn Run’s standard Lease.” J.A. at 70 (Hamasaki Aff.) In addition, Graoch stated that the Housing Authority of Jefferson County held the company “to an impossible standard” in enforcing requirements for the quality of the housing provided. Id. Specifically, according to Graoch the Housing Authority abated rental payments “for conditions which Graoch was either not made aware of prior to inspection, or which it attempted to fix only to be cited upon re-inspection for not making its repairs to the arbitrary satisfaction of the inspector.” Id. Neither party devoted attention on appeal to arguing whether a business necessity existed to justify Graoch’s withdrawal from the Section 8 program. Insufficient evidence exists for us to rule as a matter of law regarding whether business necessity justified Graoch’s withdrawal from the Section 8 program. Under these circumstances, we should remand to the district court for further development of the business necessity factor.

Upon remand, the district court should give the parties an opportunity to brief the issue of business necessity and to present relevant evidence. To survive any motion that Graoch might make for summary judgment, MHRC would need to create a genuine issue of material fact that Graoch’s withdrawal from the Section 8 program did not constitute a business ne-*394eessity. MHRC would be able to discharge this burden by showing “that there is an absence of evidence to support [Graoch’s] case.” Wright v. Murray Guard, Inc., 455 F.3d 702, 706 (6th Cir.2006) (citing Celotex Corp., 477 U.S. at 322-23, 106 S.Ct. 2548). Conversely, to survive a motion for summary judgment made by MHRC, Graoch would need to establish a genuine issue of material fact that the withdrawal constituted a business necessity.

IV. CONCLUSION

For the foregoing reasons, I concur in the lead opinion’s conclusion that a landlord’s withdrawal from the Section 8 program may give rise to liability under the disparate-impact theory of the FHA. I believe that the appropriate action for this court to take is to REVERSE the district court’s grant of summary judgment to Graoch, VACATE the denial of summary judgment to MHRC, and REMAND the case for further proceedings.

. The procedural idiosyncracies of this case involving a request for declaratory relief have placed the landlord, Graoch, in the position of Plaintiff-Appellee and MHRC in the position of Defendant-Appellant. In a more standard FHA claim, tenants or a local fair-housing council would be in the position of plaintiff and private landlords or a government entity in the position of defendant. Because Graoch sought declaratory relief on the question whether a landlord’s withdrawal from the Section 8 program could ever constitute a violation of the FHA, my analysis proceeds to consider the appropriate framework for analyzing disparate-impact claims against private defendants. That Graoch is here technically plaintiff, rather than defendant, is not significant.

. Arlington Heights II marked the second time the Seventh Circuit had faced the same action concerning constitutional and statutory challenges to Arlington Heights's zoning law. In Vill. of Arlington Heights v. Metro. Hous. Dev. Corp. (Arlington Heights I), 429 U.S. 252, 97 S.Ct. 555, 50 L.Ed.2d 450 (1977), the Supreme Court had overturned the Seventh Circuit's recognition of a discriminatory-effects violation of the Equal Protection Clause of the Fourteenth Amendment, holding that the plaintiffs did not show that discriminatory intent motivated the defendants’ challenged rezoning decision. Id. at 270-71, 97 S.Ct. 555. The Supreme Court, however, remanded the question of liability under the FHA, which the Seventh Circuit had not earlier addressed. Id. at 271, 97 S.Ct. 555.

. Shortly after the Arlington Heights II decision, the Third Circuit held in Resident Advisory Board v. Rizzo, 564 F.2d 126, 146 (3d Cir.1977), cert. denied, 435 U.S. 908, 98 S.Ct. 1457, 1458, 55 L.Ed.2d 499 (1978), that a plaintiff may establish a prima facie case under the FHA “by proving that the [defendants’] acts had a discriminatory effect and that the [defendants] have failed to justify the discriminatory results of their actions."

. The Fifth Circuit recognizes disparate-impact claims under the FHA but has at least twice found insufficient evidence to support plaintiffs’ prima facie cases and, therefore, has not proceeded beyond that stage of analysis. United States v. Mitchell, 580 F.2d 789, 791-92 (5th Cir.1978) (holding that a plaintiff may produce statistical evidence of discriminatory effects to establish a disparate-impact claim under the FHA); Bonvillian v. Lawler-Wood Hous., LLC, 242 Fed.Appx. 159 (5th Cir.2007) (unpublished opinion) (holding that the closing of an apartment building following Hurricane Katrina did not have a disparate impact on protected classes).

. In United States v. Starrett City Associates, 840 F.2d 1096, 1101 (2d Cir.), cert. denied, 488 U.S. 946, 109 S.Ct 376, 102 L.Ed.2d 365 (1988), the Second Circuit refrained from deciding whether the defendant was a private or state actor and evaluated the case as if the defendant were a state actor.

. The Tenth Circuit has rejected an attempt by HUD to define business necessity as a ‘'compelling need or necessity.” Mountain Side, 56 F.3d at 1254. But in making this determination, the Tenth Circuit closely paraphrased language from Justice O’Connor's plurality opinion in Watson. Because I believe that relying on the Watson plurality is inappropriate following the Civil Rights Act of 1991, I also refrain from following the standard for business necessity set forth in Mountain Side.

. In his dissent to the majority opinion in Salute, Judge Calabresi argued that the defendant landlord’s acceptance of Section 8 vouchers from pre-existing tenants, even though he refused new Section 8 tenants, made the landlord a partial participant in the program rather than a nonparticipant. 136 F.3d at 312 (Calabresi, J., dissenting). Nevertheless, the majority viewed the landlord as a nonparticipant and based the court's decision on the view that nonparticipation does not incur liability.

. Graoch's concession that the statistical evidence would support a disparate-impact claim, if a withdrawal from the Section 8 program could give rise to liability under the FHA and if we rejected Graoch's business-necessity defense, implicitly concedes that the evidence establishes a prima facie showing of disparate effects. Accordingly, although we do not decide the issue of business necessity today, as I explain below, we must nevertheless assume that Graoch has conceded that its withdrawal from the Section 8 program had a disparate effect on African-Americans.