Disparate-impact claims in employment and housing

Disparate-impact claims have the potential to grow in importance as a tool for reducing discrimination

2018 October

Segments of the economy still display racial and gender inequities. Certain job positions in the tech industry have been in the spotlight for this reason, but in many other highly valued fields – such as the finance and investment-management industries – female and minority employees are still often absent or are a small minority among their peers in the workplace.

Disparate-impact claims, under both Title VII and California’s Fair Employment and Housing Act (“FEHA”), allow employees to challenge disparities that can be tied to facially neutral policies, with no need to prove discriminatory intent. Disparate-impact liability is also available under the Fair Housing Act (“FHA”) to challenge housing policies that disproportionately harm people of color, women, and other protected groups. This article describes recent developments and the current state of certain types of disparate-impact claims. We discuss employment tests that screen out female job applicants, Fair Housing Act disparate-impact claims in the wake of two recent Supreme Court cases, the intersection between equal pay legislation and disparate-impact theories, and cases challenging criminal background checks by employers and landlords. Disparate- impact claims have the potential to grow in importance as a tool for reducing discrimination.

Employment application tests – exemplary new analysis by the Seventh Circuit

The Seventh Circuit issued a comprehensive analysis of a disparate-impact claim challenging a physical abilities test administered by the City of Chicago’s fire department to applicants for Emergency Medical Technician (“EMT”) positions in Ernst v. City of Chicago, (7th Cir. 2016) 837 F.3d 788. Male applicants passed the test at a rate of 98 percent, while female applicants passed at a rate of 60 percent. The district court had accepted the City’s defense of its test following a bench trial. The Seventh Circuit not only reversed but directed a verdict in favor of the plaintiffs.

The virtue of Ernst is the detailed analysis and level of care that the Seventh Circuit undertook in applying the Uniform Guidelines on Employee Selection Procedures, 29 Code of Federal Regulations (“C.F.R.”), section 1607, which govern the “validation” of such tests. In a Title VII disparate-impact claim, the plaintiff must make a prima facie showing that a facially neutral policy had an adverse impact on employees with a protected characteristic. The burden then shifts to the employer to show that the test is “job-related” with respect to the position in question, and consistent with “business necessity.” Employers may attempt to meet this burden by offering a “validation study,” which demonstrates that the test does, in fact, predict actual on-the-job performance. Such validation studies are governed by the technical standards in the Uniform Guidelines. (29 C.F.R. § 1607.14(B)(4).)

Ernst carefully examined the methodology of the expert who had devised Chicago’s test, and found numerous ways that it ran afoul of the validation regulations. The expert had begun by seeking volunteers from among Chicago’s incumbent EMTs to perform a series of physical ability exercises. The record showed that such volunteers had higher than average scores within the population of public- and private-sector employees (i.e., particularly fit employees volunteered), whereas the regulations require, as far as possible, the use of a representative sample population.

The expert then collected performance evaluations of the volunteers and attempted to correlate their general job performance with their scores on the physical exercises. The expert found that whether a volunteer had scored well on the exercises was not predictive of whether the volunteer had performed the job well. Rather than creating a new test, the expert disregarded the job evaluations and tried again, having the volunteers perform a series of “work sample” exercises, designed to mimic physical tasks required by the EMTs’ duties. She then attempted to correlate the scores on the physical ability exercises to the scores on the “work sample” exercises. Three of the physical ability exercises correlated with the work sample exercises, so she used them to create the City’s screening test. The Court found fundamental flaws in this method.

First, the fact that the physical test correlated to the work sample test was merely a “statistical form of self-affirmation” – that two tests correlated to each other did not validate either one as a predictor of job performance. Second, the work sample tests were strikingly different from the actual tasks that EMTs perform. One test was lifting a stair chair, but Chicago EMTs virtually never use stair chairs.

Another test was lifting, holding, and lowering a series of increasingly heavy stretchers 13 times in four-and-a-half minutes. In reality, stretchers are lifted by two EMTs, not one; there is rarely occasion to lift and lower them repeatedly, and weights in the test far exceeded the normal demands of the job. Such an extreme hypothetical scenario was not within the scope of “primary” EMT skills, which is what the regulations require the test to measure.

Third, the Seventh Circuit highlighted the important requirement that tests be not only accurate, but also “reliable.” One of the three physical exercises that correlated with the work sample tests did so only 50 percent of the time, meaning that the likelihood of a correlation was the same as a coin toss. Following Ernst, the reliability of tests is a fruitful area for plaintiffs to challenge.

The Court reiterated why careful enforcement of the regulations governing statistical validation studies is important: “We recognize that, in itself, there is nothing unfair about women characteristically obtaining lower physical-skills scores than men. But the law clearly requires that this difference in score must correlate with a difference in job performance. To guard against this unfairness, the law requires that the physical exam must validly test job-related skills.” (837 F.3d at 804.)

Plaintiffs’ counsel can use Ernst as an example of the detailed scrutiny courts must give to the methodology behind employment tests. Courts that might have been tempted to rubber-stamp an employer’s explanation should view Ernst as an example of their duty to ensure that the Uniform Guidelines are being followed.

Disparate-impact housing discrimination claims

The federal Fair Housing Act (“FHA”), adopted in 1968 to eradicate discrimination and segregation in housing, is similar in many ways to Title VII, and indeed is often interpreted consistently with Title VII jurisprudence. FHA cases run the gamut from individual claims about a single transaction to class cases challenging pervasive discriminatory practices or policies, and plaintiffs in impact-oriented FHA cases may be organizations or municipalities injured by the discriminatory practice at issue.

In July 2015, the much-anticipated Supreme Court decision in Texas Department of Housing and Community Affairs v. Inclusive Communities Project, 135 S.Ct. 2507 (2015) (“Inclusive Communities”) confirmed that disparate-impact claims are available under the Fair Housing Act. Although Inclusive Communities was widely, and justly, considered a victory for fair housing plaintiffs, Justice Kennedy’s majority opinion also contained language circumscribing disparate-impact claims.

The two-and-a-half years following the Inclusive Communities ruling have highlighted several key challenges that fair housing plaintiffs must overcome under that case. The challenges are derived from three limitations on disparate- impact liability highlighted in Inclusive Communities, all drawn from pre-existing disparate-impact jurisprudence. First, “disparate impact liability mandates the removal of artificial, arbitrary, and unnecessary barriers.” (Id., 135 S.Ct. at 2522.) Second, plaintiffs must identify the “policy or policies” they are challenging. (Id., at 2523.) Third, the opinion announces a “robust causality requirement” connecting the challenged policy to the disparate impact on members of a protected class. (Id., at 2523-24.) Defendants have seized on these limitations to fight off disparate- impact claims. Cases built on statistically disparate outcomes but not tightly focused on particular policies are especially vulnerable to criticisms that they fail to identify the policy at issue or to show how the disparities result from a particular policy. (See, e.g., Cobb County v. Bank of America Corp., 183 F.Supp.3d 1332 (N.D. Ga. 2016) (allegations that “minority borrowers were more likely than non-minority borrowers to receive undesirable loans and loan servicing” were inadequate to state a disparate-impact claim).)

The language from Inclusive Communities stating that disparate-impact liability must be limited to removing “artificial, arbitrary, and unnecessary barriers” is more of a cipher. On the surface, it seems to be a straightforward description of the function of the burden-shifting framework in disparate-impact claims, which serves to limit liability to policies that “are otherwise unjustified by a legitimate rationale.” (Id.,135 S.Ct. at 2513.) Yet some courts have treated it as a distinct pleading requirement (see Cobb County, 183 F.Supp.3d at 1347), and others have construed it to limit disparate- impact liability to affirmative policies rather than omissions. (See, e.g., City of Los Angeles v. Wells Fargo & Co., 2015 WL 4398858, at *8 (C.D. Cal. July 17, 2015) [(rejecting a disparate-impact claim premised on the lack of a monitoring policy) aff’d 691 F. App’x 453 (9th Cir. 2017].)

If the argument that FHA disparate-impact claims can only challenge affirmative policies gains traction, fair-housing plaintiffs in California may consider turning to disparate-impact claims under the state Fair Employment and Housing Act. (See Sisemore v. Master Financial, Inc., 151 Cal.App.4th 1386 (2007) [holding that FEHA permits disparate-impact claims].) Significantly, FEHA’s legislative history describes an intent to encompass “an act or failure to act [that] has the effect, regardless of intent, of unlawful discrimination.” (Id. at 1419 (emphasis added).)

Despite Inclusive Communities’ emphasis on the limits of disparate-impact liability, recent fair-housing cases show that disciplined and well-supported disparate-impact claims can still succeed. One notable victory was notched in National Fair Housing Alliance v. Travelers Indem. Co. (D.D.C. 2017) 261 F.Supp.3d 20, when the court refused to dismiss NFHA’s disparate-impact claim challenging Travelers’ policy of refusing to insure buildings where landlords rented to Section 8 tenants, rejecting a challenge premised on the “robust causality” requirement from Inclusive Communities. As the district court explained, “NFHA pleaded facts that show that because of the different composition of the affected population (voucher recipients) as compared to the District’s population as a whole, members of a particular class are more likely to be harmed by Traveler’s policy than are other individuals.” (Id. at 34.)

In another recent decision that approved the viability of disparate-impact claims in the mortgage lending arena, the City of Philadelphia stated a claim against Wells Fargo based on lending policies that disproportionately harmed minority homebuyers. (Philadelphia v. Wells Fargo, 2:17-cv-02203-AB, Slip Op. at 6-7 (E.D. Pa. Jan. 16, 2018).) The Court noted that the complaint identified “at least seven specific policies” that, according to the City’s statistical analysis, resulted in a disparate impact on African-American and Latino borrowers, as compared with similarly situated white borrowers.

Although FHA claims may be brought as class claims, organizational plaintiffs have long been permitted to challenge discriminatory practices and often are at the forefront of enforcement efforts. Such organizational claims received the blessing of the Supreme Court in Havens Realty Corp. v. Coleman (1982) 455 U.S. 363, 378-79, to the extent that organizations can show that the alleged discriminatory practices harm their operations by forcing them to divert resources to investigate and counteract the discriminatory conduct and by frustrating the organization’s work in furtherance of its mission.

The Supreme Court revisited standing under the FHA last spring in Bank of America Corp. v. City of Miami, Fla. (2017) 137 S.Ct. 1296. The Court avoided addressing a challenge to longstanding Supreme Court precedent that standing under the FHA reaches the Article III limits and held that even if the narrower zone of interests test did apply, the City of Miami’s alleged injuries from discriminatory mortgage lending – including harm to African-American and Latino neighborhoods, hindrance to Miami’s efforts to “create integrated, stable neighborhoods,” and the diminution of property tax revenues – fell within the FHA’s zone of interests. (Id. at 1303-05.) The Court then turned to proximate cause and held that FHA plaintiffs must show “some direct relation between the injury asserted and the injurious conduct alleged.” (Id. at 1306 (quotation marks omitted).) But rather than apply that standard to the facts at hand, the Court remanded the case and announced that “[t]he lower courts should define, in the first instance, the contours of proximate cause under the FHA.” (Id. at 1306.)

Although the Bank of America case undoubtedly heralds increased litigation around standing under the FHA, organizational plaintiffs are well positioned to succeed in establishing their right to sue. Miami’s injuries, deemed adequate by the Court, are strongly analogous to the kinds of injuries suffered by organizational plaintiffs like the NFHA in the Travelers case. Additionally, the Court cited Havens Realty as good law relevant to the broad interpretation of an “aggrieved” person under the FHA. (Id. at 1303.) Nonetheless, organizational plaintiffs (and others) should be particularly attentive to ensuring they can show injury proximately caused by the discriminatory policy when contemplating new FHA disparate-impact claims.

Using disparate-impact claims to complement equal pay claims

Disparate-impact claims under Title VII and FEHA can complement equal-pay claims by filling in gaps in the California and, particularly, the federal equal-pay laws.

California’s Fair Pay Act, which went into effect on January 1, 2016, bars employers from paying workers of one sex less than those of the other sex for “substantially similar work, when viewed as a composite of skill, effort, and responsibility, and performed under similar working conditions.” The Act provides for four affirmative defenses, permitting wage differentials caused by a (1) seniority, (2) merit, or (3) productivity system, or that are caused by (4) a “bona fide factor other than sex.” (See Lab. Code, § 1197.5.)

The Act states that such a factor other than sex may not be “derived from” a sex-based differential in compensation (e.g., prior salary, when prior salaries include a sex-based differential) and must be “job-related with respect to the position in question, and consistent with business necessity.” (Ibid.) This defense does not apply if the plaintiff can show that “alternative business practice exists that would serve the same business purpose without producing the wage differential.” (Ibid.) As of January 1, 2017, the law also applies to wage differentials by race or ethnicity. As of yet, there are few decisions applying the new law.

The Federal Equal Pay Act of 1963 (29 U.S.C. § 206(d)(1)), is structured similarly, but is less favorable for plaintiffs in several ways. First, it applies only to pay differentials for substantially “equal” work, which is narrower than the California law’s application to workers performing “substantially similar” work. Second, it applies only to differentials among employees at a given “establishment,” rather than all employees working “under similar working conditions.” And third, although its affirmative defenses are similar, its “bona fide factor other than sex” defense does not include the express prohibitions on factors “derived from” sex-based differentials in compensation, nor that the employer prove job-relatedness with respect to the position in question and business necessity. Federal circuits are split on whether the Equal Pay Act’s affirmative defense for reliance on a “bona fide factor other than sex” requires the employer to prove something akin to business necessity and job relatedness. (Compare Kouba v. Allstate Ins. Co. (9th Cir. 1982) 691 F.2d 873, 876 (requiring an “acceptable business reason”), and Belfi v. Prendergast (2d Cir. 1999) 191 F.3d 129, 136 (requiring a “legitimate business reason”), with Wernsing v. Dep’t of Human Servs. (7th Cir. 2005) 427 F.3d 466, 469 (no such requirement).) The Ninth Circuit recently heard en banc argument in Rizo v. Yovino, 16-15372 (9th Cir.), which will provide new guidance on the “factor other than sex” prong in the Ninth Circuit.

The Fair Pay Act’s incorporation of the business necessity and job-relatedness-with-respect-to-the-position-in-question requirements means that Title VII and FEHA case law applying those concepts over past decades may be useful, particularly given the lack of case law applying the new act. For example, one key question is whether the employer bears the burden of proving that it considered reasonable alternatives with less discriminatory impact to establish its affirmative defense. Title VII’s Uniform Guidelines require such a showing. (See 29 C.F.R. § 1607.3(b) (“[W]henever a validity study is called for by these guidelines, the user should include, as part of the validity study, an investigation of suitable alternative procedures … which have as little adverse impact as possible….”); Officers for Justice v. Civil Serv. Comm’n of City & Cnty. of San Francisco (9th Cir. 1992) 979 F.2d 721, 728.) Placing this burden on the employer makes good sense: the employer is in the better position to determine which alternatives will allow it to satisfy its business goals. Practitioners helping to create the initial jurisprudence interpreting California’s new act should look to Title VII disparate-impact law and encourage courts to adopt such helpful features from it.

Under federal law, Title VII disparate- impact claims can be brought separately to fill gaps left by the Equal Pay Act. First, collective actions under the Federal Equal Pay Act are “opt-in” actions governed by the same requirements as other Fair Labor Standards Act cases, meaning that only those employees who affirmatively opt in to the case can participate. (See 29 U.S.C. § 216(b) [Stating a disparate-impact claim under Title VII allows the plaintiff to seek to certify a Rule 23 opt-out class, encompassing a wider group of employees].) Second, if a policy has resulted in lower pay for women, but the plaintiff cannot prove that the women perform “substantially equal” work, or that they worked at the same “establishment” as higher-paid men, a Title VII disparate-impact claim may succeed where an Equal Pay Act claim would fail. And third, a Title VII claim may be brought for pay disparities based on protected categories other than gender.

Background check policies

Successful disparate-impact challenges to criminal background check policies have recently been brought in both the housing and employment arenas. At least two district courts have recently certified classes challenging policies disqualifying job applicants with certain criminal records. (See Little v. Washington Metropolitan Area Transit Authority (D.D.C. 2017) 249 F.Supp.3d 394, 419; Houser v. Pritzker (S.D.N.Y. 2014) 28 F.Supp.3d 222, settlement approved in Gonzalez v. Pritzker, 2016 WL 5395905 (S.D.N.Y. Sep. 20, 2016).)

Such claims are supported by recent administrative guidance. In April 2012, the EEOC issued Enforcement Guidance 915.002, advising that reliance on criminal records in employment decisions will likely result in a disparate racial impact and may be unlawful until Title VII to the extent reliance on criminal history “is not job related and consistent with business necessity.” Four years later, HUD issued similar guidance regarding housing decisions based on criminal records.

Both guidance documents agree that excluding applicants because of a prior arrest – absent any conviction – cannot satisfy defendants’ burden of offering a legitimate rationale for the challenged policy. And both documents strongly suggest that in most cases individualized assessments are an appropriate and less discriminatory alternative to blanket bans on candidates with conviction records.

The case law in this emerging area is uneven, however, with decisions coming down both ways. In one recent FHA case, the district court held that the plaintiff adequately alleged that refusal to rent to tenants with criminal records would disproportionately impact African-American residents in D.C. (Alexander v. Edgewood Management Corp., 2016 WL 5957673 (D.D.C. Jul. 25, 2016).) Similarly, the court in Williams v. Compassionate Care Hospice, 2016 WL 4149987 (D.N.J. Aug. 3, 2016), rejected a motion to dismiss a Title VII disparate-impact challenge to a criminal background check where the plaintiff relied on statistics from EEOC’s Enforcement Guidance 915.002 to plausibly allege that such a policy will disproportionately affect African-American applicants.

On the other side of the ledger are two decisions from the Ninth Circuit and the Third Circuit holding that challenged criminal background checks are consistent with business necessity and withstand a disparate-impact challenge. (Hardie v. NCAA, 876 F.3d 312 (9th Cir. 2017) [rejecting challenge to NCAA’s policy barring individuals with felony convictions from coaching NCAA certified youth tournaments]; El v. S.E. Pa. Transp. Auth., 479 F.3d 232 (3d Cir. 2007) [rejecting challenge to policy disqualifying applicants with criminal records].) Both courts relied heavily on expert testimony proffered by defendants opining that someone with a criminal record had a higher chance – even if it was in absolute terms very small – of committing a crime. Both courts also highlighted the vulnerability of the specific populations the policies at issue were designed to protect: minor athletes in Hardie, and disabled individuals relying on paratransit in El. (876 F.3d at 322, n.11; 479 F.3d at 245.) While this may provide a basis for distinguishing these cases, plaintiffs should be ready to address Hardie and El.

Between the strong administrative guidance in favor of disparate-impact claims and compelling evidence of disparities in the criminal justice system, on the one hand, and the adverse decisions in Hardie and El on the other, this developing area of disparate-impact litigation is one to watch.

Time for renewed disparate-impact focus

The past several years have seen virtually no detailed discussion of disparate- impact claims in reported California
appellate decisions. As the foregoing developments and discussion suggest, plaintiffs’ attorneys should keep this tool in mind as they watch for policies that are perpetuating troubling disparities throughout the economy.