Congress began debating fair housing legislation in 1966 during a period of widescale urban unrest,¹ but progress stagnated for almost three years. Then, in March 1968, the Kerner Commission Report found that America was “moving toward two societies, one Black, one White — separate and unequal”² and one of its major recommendations was enactment of a comprehensive federal fair housing law.³ Less than a month later, Dr. Martin Luther King Jr. was assassinated, sparking a new round of unrest. By April 11, one week after King’s death, Congress passed the Fair Housing Act (FHA).⁴

The disparate impact (DI) theory of liability was recognized as an important tool in remedying residential segregation.⁵ Unfortunately, in the 50 years since the FHA passed, the standard for determining DI liability remains unsettled, impeding the act’s effectiveness.

Though the U.S. Supreme Court held in Griggs v. Duke Power Co. and Smith v. City of Jackson⁶ that DI claims could be brought via Title VII and the Age Discrimination in Employment Act (ADEA),⁷ the Court for more than 40 years did not address whether such claims could be brought under the FHA.⁸ In 2012 and 2013, the Court granted certiorari in two cases to determine availability of DI claims under the FHA⁹ but both “unexpectedly settled.”¹⁰

During this time, all federal appeals courts considering DI claims under the FHA¹¹ recognized the theory of liability.¹² So, too, did the U.S. Department of Housing and Urban Development (HUD) in its adjudications.¹³ The courts and HUD also recognized an additional category of DI liability in policies shown to have “the unjustified effect of perpetuating segregation.”¹⁴ There are no appellate deci­sions in Michigan state court cases addressing fair housing cases involving DI claims. DI appears to be a wholly underutilized tool in Michigan to address housing policies that cause a disparate impact or perpetuate segregation.

HUD’S 2013 DI RULE

Establishing a uniform standard for disparate impact liability un­der the FHA proved elusive. In 2013, HUD clarified the frame­work by setting forth a uniform standard¹⁵ with a three-part, bur­den-shifting approach:

1) The plaintiff has the burden of proving a “[d]iscriminatory ef­fect” showing that the challenged practice “actually or predict­ably results in a disparate impact on a group of persons or creates, increases, reinforces, or perpetuates segregated hous­ing patterns because of race, color, religion, sex, handicap, familial status, or national origin.”¹⁶

2) The defendant then has the burden to prove that the challenged practice “[i]s necessary to achieve one or more substantial, le­gitimate, nondiscriminatory interests” of the defendant.¹⁷

3) If the defendant satisfies this burden, the “plaintiff may still 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.”¹⁸

THE SUPREME COURT ICP DECISION

In 2014, the Supreme Court agreed to review Texas Dep’t of Hous­ing & Community Affairs v. The Inclusive Communities Project, Inc. (ICP) to determine whether disparate impact claims are cognizable under the FHA.¹⁹ The Court held that DI claims could be brought un­der FHA sections 804(a) and 805(a),²⁰ noting that when Congress passed the amendments to the FHA in 1988, all courts of appeals considering DI claims found that such claims could be brought un­der the act.²¹ Congress, aware of this precedent, ratified the DI theory of liability as part of those amendments.²²

Further, the Court found DI liability was consistent with the FHA stat­utory purpose, saying that “unlawful practices include zoning laws and other housing restrictions that function unfairly to exclude mi­norities from certain neighborhoods without any sufficient justifica­tion”²³ and noted that recognizing DI liability “permits plaintiffs to counteract unconscious prejudices ... that escape easy classification as disparate treatment.”²⁴ And like HUD in its 2013 rule, the Court also found that actions leading to perpetuation of segregation can give rise to DI liability²⁵ and, after summarizing the burden-shifting framework in the HUD rule, agreed that the plaintiff had the burden of proving there were no less-discriminatory alternatives.²⁶

The Court articulated cautionary standards to govern the pleading and proofs for DI claims:

• A plaintiff must “allege facts at the pleading stage or produce statistical evidence demonstrating a causal connection” be­tween the policy and discriminatory effects.²⁷
• A robust causality requirement that ensures evidence of “ra­cial imbalance ... does not ... establish a prima facie case of disparate impact.” The robust causality requirement “protects defendants from being liable for racial disparities they did not create.”²⁸
• Policies of governmental or private actors will not give rise to DI liability unless they create “artificial, arbitrary, and unnec­essary barriers.”²⁹

As noted above, the Court granted review limited to whether DI claims were cognizable under the FHA, declining to require brief­ing on applicable standards and burdens of proof. Nonetheless, lower courts have sought to apply these cautionary standards.

HUD’s 2020 DI RULE AND RETURN TO THE 2013 RULE

Though the Supreme Court in ICP either left intact HUD’s 2013 rule or, at best, made it somewhat more demanding,³⁰ the agency in 2020 took the unusual step of amending its 2013 rule, claiming revisions were needed “to better reflect” the ICP ruling.³¹

Prior to implementing the 2020 regulation, three separate actions challenged it.³² In one, the district court enjoined HUD from imple­menting the regulation³³ and, as a result, it never went into effect. HUD subsequently reinstated its 2013 rule.³⁴

DI CLAIMS UNDER ICP AND HUD’s 2013 RULE

In post-ICP cases, courts have reached different conclusions as to whether the U.S. Supreme Court adopted HUD’s framework or modified it.³⁵ Courts have also struggled with the robust causation requirement,³⁶ but the Supreme Court analysis in ICP may be best understood as accepting HUD’s three-step, burden-shifting ap­proach from its 2013 regulation while offering guidance in analyz­ing each step of the claim.

To show a prima facie case at the first step of a claim, the plain­tiff must sufficiently allege (at the pleading stage) or establish (at summary judgment or trial) a robust causality requirement by spe­cifically identifying the policy or policies causing the disparity. A one-time decision ordinarily fails to constitute a policy,³⁷ evidence showing only preexisting statistical disparities between protected and non-protected groups is not sufficient, and the disparity must not have been caused by factors distinct from the challenged policy.

Most courts find robust causality if the plain­tiff presents statistical evidence showing a protected class “is disproportionately af­fected by a . . . policy.”³⁸ Other courts ap­ply more stringent standards requiring the plaintiff to show not just that the policy had a disproportionate adverse effect on a protected class, but also caused the sta­tistical disparity.³⁹

In the rebuttal step, courts agree that the defendant has the burden of showing the policy is necessary to achieve one or more substantial, legitimate, nondiscriminatory interests. This is not an onerous burden, and the defendant has leeway to identify the policy’s valid interests or priorities. Fur­ther, federal law or other regulatory require­ments that substantially limit the defendant’s discretion in fashioning a policy should also be considered, as should policies of govern­mental or private actors that create “artifi­cial, arbitrary, and unnecessary barriers.”⁴⁰

Analysis of the third stage under ICP does not vary from HUD’s formulation. Both re­quire the plaintiff to show the defendant’s interests could be served by an alternative practice that has less discriminatory impact or effect.⁴¹

POST-ICP CLAIMS IN HOUS­ING-RELATED CONTEXTS

Courts have addressed FHA DI claims in residential housing-related contexts using the framework set forth in ICP and HUD’s 2013 regulation. Five recent cases illus­trate the variety of policies that may give rise to DI claims and show they may be successfully prosecuted.

Zoning (Race)

In Mhany Mgmt, Inc. v. County of Nassau,⁴² the U.S. Court of Appeals for the Second Circuit considered a challenge to a city zoning ordinance as a perpetuation of seg­regation. The original multi-family residen­tial group zoning would have allowed the plaintiff to construct residential apartment units in a city that contained no affordable housing and where minorities constituted about 4% of the overall population. Due to public pressure, the city rezoned the prop­erty to residential-townhouse, which defined “townhouse” as a “single-family dwelling unit.”⁴³ The district court found that rezon­ing “significantly decreased the potential pool of minority residents likely to move into the new housing development.”⁴⁴

Analyzing the DI claim under ICP and HUD’s 2013 regulation, the Second Circuit found that the rezoning decision perpetuat­ed segregation within the city. In the sec­ond step of the claim, the city contended that rezoning advanced legitimate interests by reducing traffic and the strain on public schools. As to the third step, the court found that HUD regulations abrogated prior cir­cuit precedent regarding burden of proof to establish the absence of a less discriminato­ry alternative. Since the district court placed the burden on the city to establish the ab­sence of less-discriminatory alternatives, the appeals court reversed and remanded the ruling.⁴⁵

Citizenship Status (National Origin)

In Reyes v. Waples Mobile Home Park Ltd. Partnership,⁴⁶ a mobile home park began enforcing a policy requiring all occupants to document their legal status in the U.S. by presenting an original Social Securi­ty card or an original (foreign) passport, U.S. visa, and original arrival/departure form (I-94 or I-94W). Under the policy, the park would not renew the lease and seek eviction for any tenant with one or more occupants who did not provide the required documentation.⁴⁷

The plaintiffs alleged that Latinos comprised nearly 65% of the undocumented popula­tion in Virginia, undocumented immigrants comprised more than 36% of the Latino population in the state, and Latinos were 10 times more likely than non-Latinos to be adversely affected by the policy. The court said the plaintiffs satisfied ICP's robust cau­sality requirement by asserting the policy would disproportionately subject Latino ten­ants to eviction.⁴⁸

Insurance (Race and Sex)⁴⁹

In National Fair Housing Alliance v. Trav­elers Indemnity Co.,⁵⁰ plaintiff’s testers⁵¹ in the District of Columbia posed as purchas­ers of an apartment complex and contact­ed various insurance companies to inform them that they participated in the Housing Choice Voucher (HCV)⁵² program. In each instance, insurance brokers told testers the companies would not insure the property if voucher recipients resided in the building.⁵³

According to statistics included in the com­plaint, D.C. households are approximately 45% non-Hispanic Black or African Amer­ican and 41% non-Hispanic white. More than 47% of D.C. households are headed by women. For households receiving HCVs, 92% are non-Hispanic Black or African American compared to roughly 45% of all households receiving HCVs in D.C. Only 1% of HVC recipients are non-Hispanic whites. The percentage of households re­ceiving HCVs headed by women totaled 81.5%, compared to 47% of all households headed by women in all of D.C. Further, residents participating in the HCV program were largely concentrated in four census tracts that were nearly 85% Black compared to the entire D.C. percentage of 51%.⁵⁴

The court held that the plaintiff’s allegations satisfied the ICP robust pleading require­ment. According to the court, the insurer’s policy ‘‘will exacerbate racial and sex-based disparities by having a disproportionate impact on African-American residents and members of women-headed households.”⁵⁵

Occupancy Limits (Familial Status)

In Rhode Island Commission for Human Rights v. Graul,⁵⁶ the tenants, a married couple renting a one-bedroom unit, had a child. The bedroom measured at least 150 square feet. The apartment complex gave the tenants six months to move to a two-bed­room unit or vacate due to a two-person-per-bedroom unit policy.⁵⁷

Applying the three-step analysis in ICP, the court found in step one that a prima facie case was established based on evidence showing disparity ratios for three-person households with children, four-person households, and five-person households well above the 1.25 ratio for statistical significance,⁵⁸ even when controlled by in­come range.

In the second step, the apartment complex claimed its policy complied with the view of a local official regarding the building code and asserted there was a minimum re­quirement of 170 square feet for bedroom occupancy by three persons. The court re­jected this assertion as misreading the build­ing code, which required bedroom units to have at least 70 square feet for the first occupant and at least 50 square feet per person for more than one occupant. A 150 square foot bedroom unit, therefore, could hold three people.⁵⁹ The complex failed to carry its burden in the second step, and the court did not need to reach the third step.

Use of Criminal Records

(Race and National Origin)

Connecticut Fair Hous. Ctr v Corelogic Rent­al Prop. Solutions, LLC⁶⁰ involved a chal­lenge to the defendant’s use of CrimSAFE criminal record screening. The defendant provides a housing provider with a form list­ing general categories of crimes for which it would like CrimSAFE to screen. Using those categories, the defendant generated a one-page report indicating whether dis­qualifying records were found, including ar­rests that that did not lead to conviction. The CrimSAFE report had a maximum “look­back” of 99 years for convictions and seven years for non-convictions and provides no additional information such as the underly­ing records, the nature of the alleged crime, the date of the offense or the outcome of the case, if any.⁶¹

One plaintiff, a Latino man, suffered de­bilitating injuries in a car accident, ren­dering him unable to speak, walk, or care for himself. The year before his injury, he was arrested for retail fraud but never convicted. Using CrimSAFE, the defen­dant informed the housing provider that the plaintiff was disqualified based on un­specified criminal records.⁶²

In denying cross motions for summary judgment on the DI claims, the court noted statistical evidence showing Blacks were more than four times as likely as whites and Latinos more than 2.5 times as likely as whites to have been jailed at some point in their lives.

With respect to the second step, the de­fendant claimed that federally subsidized properties required screening for arrests and convictions and the process accurate­ly categorizes risk levels and reduces bias. The court, however, found that the defen­dant did not provide legal or empirical support for treating a pending arrest record as sufficient to determine that a prospective tenant would threaten the health or safety of a residential community. The court noted that HUD, in informal guidance, treated ar­rest records as insufficient to warrant denial of admission.⁶³ In the third step, the court found that the plaintiffs met their burden un­der ICP and the 2013 regulation of show­ing less discriminatory alternatives.

CONCLUSION

The disparate impact theory of liability is well established as a cognizable theory of liability in fair housing cases. DI claims may challenge practices that result in discrimination. Still, the theory remains underutilized as a tool to combat policies that adversely impact one or more protected classes or perpetuate segregated housing patterns. The DI theory of liability should prompt all entities involved in any aspect of housing to carefully review all policies — without establishing unlawful quotas — to ensure they do not result in adverse discriminatory effects or perpetuate segregation.