Opinion
Nos. 10–3965 10–3983.
2011-12-12
The NATIONAL ASSOCIATION FOR THE ADVANCEMENT OF COLORED PEOPLE “NAACP”; the Newark Branch, NAACP; the New Jersey State Conference NAACP; Allen Wallace; Lamara Wapples; Altarik White v. NORTH HUDSON REGIONAL FIRE & RESCUE, a body corporate and politic of State of New Jersey; Alex M. DeRojas; Alexander Rodriguez; Randy Vasquez; Carlos Castillo; Orlando Duque; Pablo Claro, (Intervenors in D.C.)North Hudson Regional Fire & Rescue, a body corporate and politic of the State of New Jersey, Appellant.
David F. Corrigan [Argued], Keyport, NJ, for Appellants Carlos Castillo, Pablo Claro, Alex Matthew DeRojas, Orlando Duque, Alexander Rodriguez and Randy Vasquez. Thomas R. Kobin [Argued], John L. Shahdanian, II, Chasan, Leyner & Lamparello, Secaucus, NJ, Thomas B. Hanrahan, River Edge, NJ, for Appellant N. Hudson Regional Fire & Rescue.
David F. Corrigan [Argued], Keyport, NJ, for Appellants Carlos Castillo, Pablo Claro, Alex Matthew DeRojas, Orlando Duque, Alexander Rodriguez and Randy Vasquez. Thomas R. Kobin [Argued], John L. Shahdanian, II, Chasan, Leyner & Lamparello, Secaucus, NJ, Thomas B. Hanrahan, River Edge, NJ, for Appellant N. Hudson Regional Fire & Rescue. David L. Rose [Argued], Yuval Rubinstein, Rose Legal Advocates, Washington, DC, David Ben–Asher, Eugenie F. Temmler, Rabner, Allcorn, Baumgart, Ben–Asher & Tucker, Upper Montclair, NJ, for Appellees NAACP NJ Conference, NAACP Newark Branch, National Association for the Advancement of Colored People, Allen Wallace, Lamara Wapples and Altarik White.
Before: FISHER, HARDIMAN and GREENAWAY, JR., Circuit Judges.
OPINION OF THE COURT
HARDIMAN, Circuit Judge.
This appeal arises under Title VII of the Civil Rights Act of 1964, as amended in 1991. At issue is the legality of a residency requirement for firefighter candidates imposed by North Hudson Regional Fire and Rescue (North Hudson), a fire department comprising five New Jersey municipalities. The United States District Court for the District of New Jersey held the residency requirement invalid because it has a disparate impact on African–American applicants. North Hudson and six Hispanic firefighter applicants appeal the District Court's judgment.
I
A. Overview of Firefighter Hiring in New Jersey
New Jersey state law has regulated the hiring of firefighters for nearly a century. The current governing statute, the New Jersey Civil Service Act (Civil Service Act), was enacted in 1986 and establishes rules and procedures governing public employment that are intended to “advance employees on the basis of their relative knowledge, skills and abilities.” N.J. Stat. Ann. 11A:1–2. Among other things, the Civil Service Act requires New Jersey municipalities to fill civil service jobs, including firefighter, pursuant to a process controlled by the New Jersey Department of Personnel (NJDOP). The NJDOP periodically creates, administers, and scores a firefighter examination. It then publishes hiring lists for the individual fire departments, designing each list to account for the hiring preferences a given municipality applies. New Jersey regulations permit local governments to give a hiring preference to their own residents, N.J. Admin. Code § 4A:4–2.11, and numerous municipalities do so.
In 2008, the NJDOP was replaced by the Civil Service Commission. See Brief History, State of N.J. Civ. Serv. Comm'n, http:// nj. gov/ csc/ about/ about/ history/ (last visited Nov. 15, 2011). The testing and ranking tasks formerly performed by the NJDOP are now assigned to the Division of Selection Services. See Division of Selection Services, State of N.J. Civ. Serv. Comm'n, http:// nj. gov/ csc/ about/ divisions/ selection/ (last visited Nov. 15, 2011).
In some New Jersey municipalities, residency requirements are not only permitted, but judicially mandated. In 1977, the United States Department of Justice sued twelve municipalities—Atlantic City, Camden, East Orange, Elizabeth, Hoboken, Jersey City, New Brunswick, Newark, Passaic, Paterson, Plainfield, and Trenton—alleging that they engaged in race-based discrimination with respect to testing and appointments in the hiring and promotion of firefighters. See United States v. New Jersey, Nos. 77–2054, 79–0184 (D.N.J. filed Oct. 4, 1977). The case was settled on May 30, 1980 by a consent decree (Consent Decree) that mandated residency requirements in those twelve municipalities (Consent Decree Municipalities) and remains in effect over thirty years later.
Although the Consent Decree purported to establish that “[r]emedial actions and practices required by the terms or permitted to effectuate and carry out the purpose of [the Consent Decree] shall not be deemed discriminatory within the meaning of [the disparate treatment statute,] 42 U.S.C. [§ ] 2000e–2(a),” we note that disparate-impact claims had not been statutorily defined at the time of the Consent Decree and thus were unaffected by it.
B. North Hudson's Residency Requirement
Against this historical backdrop, North Hudson was formed in 1998 as a consortium of five municipalities: Guttenberg, North Bergen, Union City, Weehawken, and West New York. At the time North Hudson was formed, each of its member municipalities imposed a residency requirement, so North Hudson continued that practice. The validity of North Hudson's residency requirement is the subject of this appeal.
Like all New Jersey fire departments, North Hudson is subject to the Civil Service Act, which requires it to hire pursuant to statewide NJDOP testing and ranked lists derived therefrom. Applicants who pass the written exam undergo a physical test, and the ranked lists are derived from the results of both assessments. These rankings are used to publish eligibility lists. The Civil Service Act requires organizations to hire from the lists in rank order, and pursuant to New Jersey's “Rule of Three,” North Hudson must offer each open position to one of the three highest ranked candidates on the eligibility list provided by the NJDOP. N.J. Stat. Ann. § 11A:4–8; see also In re Foglio, 207 N.J. 38, 22 A.3d 958, 959 (2011) (describing the Rule of Three mandate). As is common practice in New Jersey, the eligibility lists created for North Hudson include only candidates who lived in one of its five member municipalities when they took the written exam (Residents–Only List). Accordingly, North Hudson never sees the names or scores of ineligible non-resident applicants. North Hudson's Residents–Only List also indicates a preference for veterans and accounts for volunteer firefighting experience and other statutorily mandated criteria. After a candidate is selected from the Residents–Only List, North Hudson battalion or deputy chiefs verify the candidate's residency at the time of hiring. Once the candidate has been hired, however, he may live anywhere; some North Hudson firefighters and officers live in neighboring counties or as far as sixty miles away.
The NJDOP firefighter exams relevant to this litigation were administered in 1999, 2002, 2003, and 2006. A more recent exam was given in October 2010. See Entry Level Firefighter Examination, State of N.J. Civ. Serv. Comm'n, http:// www. state. nj. us/ csc/ authorities/ faq/ safety/ firefighters. html (last visited Nov. 15, 2011). Because the expert reports in this case were prepared prior to the 2010 exam, the results of that test are not at issue.
As of 2000, the population of North Hudson's member municipalities was 69.6% Hispanic, 22.9% white non-Hispanic, and 3.4% African–American. In 2008, the Equal Employment Opportunity Commission (EEOC) reported that North Hudson employed 323 full-time employees, including 302 firefighters. North Hudson's firefighter ranks included 240 white non-Hispanics, fifty-eight Hispanics, and two African–Americans.
When this litigation began, North Hudson sought to fill thirty-five to forty new firefighter positions. The six Hispanic applicants who intervened in this case (Intervenors) earned passing scores on the 2006 NJDOP firefighter exam and satisfied North Hudson's residency requirement. Based on their scores on the 2006 exam, Intervenors were ranked twenty-first, twenty-fifth, twenty-sixth, forty-fifth, forty-ninth, and seventieth on North Hudson's Residents–Only List. Given their rankings, Intervenors claim they would have been offered a firefighting job.
C. The Rodriguez Settlement
Like that of many other New Jersey fire departments, the racial composition of North Hudson has been the subject of legal challenge. In 2001, thirteen Hispanic firefighters sued North Hudson for disparate-impact discrimination in promotions. See Rodriguez v. N. Hudson Reg'l Fire & Rescue, No. 01–3153 (D.N.J. filed July 2, 2001). After almost four years of litigation, the parties settled the case. In the settlement agreement ( Rodriguez Settlement), North Hudson agreed to promote four of the plaintiffs, waive length-of-service prerequisites for registering for the next chief/officer exam, and advertise in Spanish and English media to “attract additional qualified applicants of Hispanic/Latino origin.” The Rodriguez Settlement imposed no other hiring obligations on North Hudson, however. Although the Rodriguez Settlement dealt primarily with promotional practices, the advertising initiatives may have increased Hispanic hiring. Whereas in 2001 the Rodriguez plaintiffs alleged that only 7% of North Hudson firefighters were Hispanic, in 2007, 38% of new hires were Hispanic, and by 2008, the percentage of Hispanic North Hudson firefighters had climbed to 19%. None of the Plaintiffs or the Intervenors in this appeal was a party to the Rodriguez case.
D. NAACP's Disparate–Impact Claim
In April 2007, the Newark Branch of the NAACP, the New Jersey Conference of the NAACP, and firefighter candidates Allen Wallace, Lamara Wapples, and Altarik White (collectively, NAACP Plaintiffs) sued North Hudson alleging that its residency requirement causes a disparate impact on African–American applicants. In February 2009, the District Court certified the NAACP Plaintiffs' class and preliminarily enjoined North Hudson from hiring firefighters from its then-current eligibility list, which included only those candidates who were residents of the North Hudson municipalities when they took the statewide exam. NAACP v. N. Hudson Reg'l Fire & Rescue, 255 F.R.D. 374 (D.N.J.2009). North Hudson filed an interlocutory appeal.
While that appeal was pending, in June 2009, the Supreme Court decided Ricci v. DeStefano, 557 U.S. 557, 129 S.Ct. 2658, 174 L.Ed.2d 490 (2009), and in September 2009, the District Court permitted six Hispanic firefighters eligible for hiring based on North Hudson's then-current list to intervene. Because Ricci involved the interplay between disparate-impact and disparate-treatment claims, we remanded the case sua sponte to the District Court in March 2010.
On April 23, 2010, the District Court found that North Hudson's residency requirement might be lawful because of “business necessity.” Based on that finding and equitable considerations, the District Court vacated the preliminary injunction.
After the preliminary injunction was vacated, the parties moved for summary judgment. The NAACP Plaintiffs sought judgment on their disparate-impact claim and a permanent injunction against North Hudson's use of the Residents–Only List. North Hudson argued that the NAACP Plaintiffs failed to establish a causal relationship between the residency requirement and the statistical disparity in its African–American employment ratio. Alternatively, North Hudson claimed it had established the business-necessity defense. In addition, North Hudson and Intervenors claimed that Ricci provided a separate defense, and Intervenors sought attorney's fees for their role in causing vacatur of the District Court's February 2009 preliminary injunction.
The District Court granted the NAACP Plaintiffs' motion, permanently enjoined North Hudson's use of its Residents–Only List, and denied Intervenors' request for attorney's fees. NAACP v. N. Hudson Reg'l Fire & Rescue (North Hudson), 742 F.Supp.2d 501 (D.N.J.2010).
The District Court also referenced a state law claim, but “confined [its opinion] to discussion of Plaintiffs' Title VII claim because the state law claim is analyzed in essentially the same way as the federal claim.” North Hudson, 742 F.Supp.2d at 506 n. 2. Neither party has raised state law issues on appeal.
E. Statistical Evidence and the District Court's Findings
In adjudicating the parties' motions for summary judgment, Judge Debevoise conducted an extremely thorough analysis of the facts and expert reports. Because disparate-impact claims depend heavily on statistical proof of discriminatory effects, we review the experts' findings and the District Court's conclusions at length. 1. Dr. Richard Wright
The NAACP Plaintiffs presented the expert report of Dr. Richard Wright to establish their prima facie case of disparate-impact discrimination. In his 2008 report, Wright identified disparities between the percentage of qualified African–Americans in the relevant labor market, which he defined in several alternative ways, and the percentage of African–Americans employed by North Hudson.
The District Court first considered Wright's definition of the relevant labor market. The U.S. Census Bureau reported that in 2003 the average daily commute was twenty-four minutes nationally and twenty-nine minutes for North Hudson residents. By performing a geographic information system analysis, Wright concluded that most people living within a ten-mile radius of North Hudson's center would have no more than a twenty-nine-minute commute. Wright also opined that North Hudson could reasonably be expected to draw its employees from across New Jersey because state-regulated workers tend to search for jobs statewide. Wright concluded that “the appropriate labor market from which North Hudson may be expected to draw its protective service personnel is either the whole state or the neighboring three-county area.” The District Court accepted this definition because it was “based in sound reasoning, and because [North Hudson] did not dispute” it, present evidence to contradict it, or suggest an alternative definition. North Hudson, 742 F.Supp.2d at 516.
The District Court next considered whether Wright's comparisons demonstrated the kind of substantial statistical proof and causal relationship necessary to establish a prima facie disparate-impact claim. To do so, Wright compared the proportion of African–Americans employed by North Hudson with the percentages of African–Americans employed in “full time protective service” positions in the Tri-County Area and in New Jersey. The parties disputed—and continue to dispute—whether looking to the ratio of African–Americans in protective service jobs provides an accurate prediction of their expected presence in firefighting jobs. Although Wright parenthetically noted that protective service jobs consist of “mainly police officers [and] firefighters,” a 2000 list of North Hudson's “Protective Service Occupations” codes suggests that this category may also include a substantial number of non-analogous positions, including crossing guards, gaming surveillance officers, private detectives and investigators, animal control workers, parking enforcement workers, and fish and game wardens. Other positions with protective service codes, like lifeguards, transit and railroad police, and bailiffs and correctional officers, may involve some skills similar to those required for firefighting, but they unquestionably diverge in other significant ways. The District Court concluded that “[t]he qualifications for employment in full time protective service work in the Tri–[C]ounty [A]rea or in New Jersey would be similar to the qualifications to be employed by [North Hudson]” and that the protective service comparison “more closely track[ed] the requirement of a qualified population.” North Hudson, 742 F.Supp.2d at 517.
Wright's results in this comparison were compelling. In the Tri–County Area, 37.4% of protective service positions are held by African–Americans. Based on this percentage, one would expect 121 North Hudson firefighters to be African–American. Similarly, 20% of protective service workers statewide are African–American, so, based on that percentage, one would expect North Hudson to employ sixty-five African–American firefighters. The differences of 13 and 8.76 standard deviations in these comparisons leave “virtually no probability” that the discrepancies are the result of chance. Wright's calculations indicated that African–Americans are “significantly under-represented” in North Hudson. Given the near impossibility that the disparities are caused merely by chance, Wright concluded: “[T]his likely results from discriminatory hiring practices.” The District Court agreed and found that “[t]he difference between these expected numbers of African Americans—121 or 65—and the actual number employed by [North Hudson]—2—is striking and is sufficiently substantial to raise an inference of causation.” North Hudson, 742 F.Supp.2d at 517. 2. Dr. Bernard Siskin
Although 2008 data indicate that only 302 of North Hudson's 323 fire department personnel are firefighters, Wright used the 323 figure throughout his calculations. Given the standard deviations he describes, it is unlikely that amended calculations using the 302 figure would affect the ultimate conclusion that North Hudson's hiring practices are causing a disparately small African–American firefighter population.
Wright also compared the percentage of African–American firefighters in North Hudson's employ with the percentages of African–Americans in the general populations of several geographical areas around North Hudson and with the percentages of African–Americans employed full-time in state and local government in Hudson County and in New Jersey. The District Court found that “[t]hese metrics, however, do not fully account for the requirement that the population be qualified,” and that “[c]ertainly the population in general cannot be assumed to be qualified for a job with [North Hudson], and it is unclear whether the qualifications for employment in state and local government in New Jersey would be similar to those required for a ranking on the DOP list.” North Hudson, 742 F.Supp.2d at 516–17. Thus, the District Court did not rely on these broader measures of the qualified labor market.
After reviewing Wright's opinion, the District Court examined the report of North Hudson's expert, Dr. Bernard Siskin, to determine whether his findings undermined or contradicted the prima facie statistical evidence presented by Wright.
Siskin calculated the expected number of African–Americans in North Hudson using the scores and rankings of actual applicants in the 1999, 2002, and 2006 NJDOP testing cycles. Siskin created new eligibility lists using final test scores and veteran status to rank the candidates as though the candidate pool had included: (1) Hudson County, (2) the Tri–County Area, (3) a five-mile radius, or (4) a ten-mile radius. As the District Court noted, “[u]sing the actual DOP test results obviously accounts for the requirement that the population being compared is ‘qualified.’ ” Id. Although expanding the eligibility list to include Hudson County or a five-mile radius added only one to two African–Americans to the top thirty-five candidates depending on the exam year, when the list was expanded to include the Tri–County Area, which was the relevant labor market according to the District Court, six to twelve African–American applicants placed in the top thirty-five, and six to fifteen placed in the top fifty. Yet, the actual Residents–Only Lists from those years included no African–Americans in that range. Overall gains in the top ninety also were substantial, with expansion to include the Tri–County Area adding eleven to nineteen African–American candidates. Thus, the District Court found that including the Tri–County Area caused “a significant number of African Americans [to be] added to the DOP lists.” Id. Siskin's report noted that these gains came primarily at the expense of Hispanics, who moved down, often significantly, in the hypothetical expanded-list rankings. Siskin concluded that Caucasians would benefit most from an expansion of North Hudson's hiring area.
Siskin acknowledged that his original hypothetical rankings assumed two things: “(i) that [non-resident candidates] would necessarily prefer appointment to [North Hudson] compared to any other jurisdiction they sought; and (ii) that they would be as likely to receive an appointment offer from [North Hudson] as from any other jurisdiction.” Because Siskin considered these assumptions unrealistic, he next calculated the hypothetical rankings omitting all non-resident candidates who received appointments elsewhere during that hiring cycle, under the premise that they would have accepted those appointments instead of continuing to compete for a job with North Hudson. Removing these otherwise-appointed candidates from the hypothetical expanded eligibility lists had varied effects for the different years studied. For the 1999 hypothetical expanded list, it had no effect at all; the top thirty-five still contained twelve more African–Americans than North Hudson's Residents–Only List. For the 2002 expanded list, there was no change to the top thirty-five (still six more African–Americans than on North Hudson's 2002 Residents–Only List) and a decrease of only two in the top fifty (still six more African–Americans than on North Hudson's 2002 Residents–Only List).
Bolstering this assumption are NJDOP data that indicate a candidate is most likely to be appointed in the jurisdiction of residence; 60% of fire department appointments go to local residents. For African–Americans, 73% of those hired are hometown candidates.
As the District Court noted, “[t]he only tables that do not predict any added African Americans are Tables 7–2006 and 8–2006.” Id. at 518. Those 2006 tables showed the most substantial change, but they also excluded a class of candidates included in the 1999 and 2002 lists. Table 7–2006 contains calculations which “exclud[ed] those appointed or having a better rank outside [North Hudson's] local area.” This additional exclusion eliminated candidates who ranked “substantially better” on another municipality's eligibility list. By Siskin's definition, a candidate had a “substantially better” rank elsewhere if his rank on the North Hudson expanded list was at least twice his best rank on another municipality's list (where lower numbers mean higher rankings on the eligibility lists). “That is, a candidate ranked 12th on the expanded [North Hudson] list with a ‘best’ rank order number of six (6) or better on some other list would be excluded from the expanded [North Hudson] list.” Likewise, a candidate who ranked third on the North Hudson list but first on another municipality's list would be omitted, even though he had an identical chance of being hired by North Hudson under the Rule of Three. The candidates remaining on the hypothetical expanded lists for 2006 were then re-ranked. The result for the Tri–County expanded list was that no African–Americans were added to any of the ranges measured; all of them had a “substantially better” rank in another municipality.
The District Court doubted the value of the 2006 tables:
[W]hen the Court compares the results of Table 8–2006 to Tables 8–1999 and 8–2002, ... it appears that the assumption underlying Table 8–2006 may not bear out in reality. Since Tables 8–1999 and 8–2002 are based on the actual number of applicants who were hired by another department, the results in those tables are grounded in actual past events. Id. at 518. Ultimately, the District Court concluded that the “assumption in 8–2006 far over-emphasizes the impact of hiring in other jurisdictions on the number of African American applicants who would be highly ranked on the Tri-[C]ounty lists” and found that the 2002 and 1999 results were “more compelling.” Id. at 519.
Siskin conducted a second analysis, but it played little role in the District Court's decision and is unnecessary to our findings, so we summarize it only briefly. Siskin used the difference between the percentage of African–Americans in North Hudson's general population (3.4%) and the percentage of African–Americans in North Hudson's fire department (0.62%) to establish a baseline percentage of African–Americans who are both interested in and qualified for firefighter positions (17.22%). Siskin assumed that this relationship between race/ethnicity and interest/qualification holds true across all the geographical areas he studied and determined North Hudson would be expected to employ fifteen African–Americans if it hired from the Tri–County Area and eight if it hired from the entire state of New Jersey. Based on this second analysis, Siskin concluded that “removing the residency requirement would be expected to result in ... a small ultimate increase in the percent of African Americans in [North Hudson's] workforce.”
Siskin's overall assessment of the impact on African–Americans was that “with the exception of the larger expanded ( i.e., Tri–County) area, there is a trivial increase in African Americans.” 3. District Court's Conclusions
In the District Court's view, Wright's statistical evidence sufficed to establish a prima facie case of disparate impact, including both substantial statistical disparities and a causal link to North Hudson's residency requirement. The District Court then searched for contravening evidence from North Hudson's own expert, but Siskin's report revealed that “[his] own results predict[ed] that a significant number of qualified African Americans would be eligible and qualified for employment with [North Hudson] if the labor market were expanded to the Tri-[C]ounty [A]rea.” Id. Thus, “both experts' reports support[ed] a statistical finding of disparate impact,” and North Hudson had “failed to raise a genuine dispute that the 2006 data predicts that no African Americans would be in the top-ranked candidates on a DOP list.” Id.
The District Court next considered North Hudson's business-necessity defense and rejected it because living in North Hudson was not a “mandatory minimum requirement” for familiarity with local geography, swift response times, or a bilingual firefighter force and because less discriminatory alternative means of achieving these goals were apparent. Id. at 522–25. It also found the residency requirement was not compelled by the Rodriguez Settlement. Id. at 523–24. Nor did the residency requirement ensure that North Hudson firefighters would live in the North Hudson municipalities after they were hired. Id. at 523. Finally, the District Court determined that the Supreme Court's decision in Ricci v. DeStefano, 557 U.S. 557, 129 S.Ct. 2658, 174 L.Ed.2d 490 (2009), did not afford North Hudson an alternative defense because the expansion of its hiring list, to the detriment of Intervenors, would not be “because of race-base[d] statistics alone; ... [but rather] because the residency requirement causes a disparate impact that is not justified by business necessity.” North Hudson, 742 F.Supp.2d at 528. North Hudson and Intervenors filed this appeal.
II
The District Court exercised jurisdiction over the NAACP Plaintiffs' Title VII claims pursuant to 28 U.S.C. § 1331. We have jurisdiction over this appeal under 28 U.S.C. § 1291.
We exercise plenary review over a district court's summary judgment decision, and we “apply the same test required of the district court.” Groman v. Twp. of Manalapan, 47 F.3d 628, 633 (3d Cir.1995). On a motion for summary judgment, the movant must show that there is “no genuine issue as to any material fact,” such that he is “entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(a). To be material, a fact must have the potential to alter the outcome of the case. See Kaucher v. Cnty. of Bucks, 455 F.3d 418, 423 (3d Cir.2006). As the District Court correctly stated, “[d]isputes over irrelevant or unnecessary facts will not preclude a grant of summary judgment.” North Hudson, 742 F.Supp.2d at 509. Faced with a summary judgment motion, the court must view the facts “in the light most favorable to the nonmoving party.” Scott v. Harris, 550 U.S. 372, 380, 127 S.Ct. 1769, 167 L.Ed.2d 686 (2007). After the movant shows that there is no genuine issue for trial, the non-moving party then bears the burden of identifying evidence that creates a genuine dispute regarding material facts. See Celotex Corp. v. Catrett, 477 U.S. 317, 323, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986). Ultimately, “[w]here the record taken as a whole could not lead a rational trier of fact to find for the non-moving party, there is no genuine issue for trial.” Matsushita Elec. Indus. Co., Ltd. v. Zenith Radio Corp., 475 U.S. 574, 587, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986) (internal quotation marks omitted).
We review a district court's grant of a permanent injunction for abuse of discretion. United States v. Bell, 414 F.3d 474, 478 (3d Cir.2005). An abuse of discretion occurs when the District Court's decision “rests upon a clearly erroneous finding of fact, an errant conclusion of law, or an improper application of law to fact.” ACLU of N.J. v. Black Horse Pike Reg'l Bd. of Educ., 84 F.3d 1471, 1476 (3d Cir.1996) (quoting Int'l Union, UAW v. Mack Trucks, Inc., 820 F.2d 91, 95 (3d Cir.1987)). We do not interfere with a district court's discretionary judgments, unless it clearly erred in weighing the relevant factors and reaching a conclusion. Morgan v. Perry, 142 F.3d 670, 683 (3d Cir.1998).
III
A
Title VII is intended to ensure that “the workplace [is] an environment free of discrimination, where race is not a barrier to opportunity.” Ricci v. DeStefano, 557 U.S. 557, 129 S.Ct. 2658, 2674, 174 L.Ed.2d 490 (2009). It endeavors “to promote hiring on the basis of job qualifications, rather than on the basis of race or color.” Griggs v. Duke Power Co., 401 U.S. 424, 434, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971) (quoting 110 Cong. Rec. 7247 (1964)). In furtherance of this objective, Title VII makes it illegal for an employer to “limit, segregate, or classify his employees or applicants for employment in any way which would deprive or tend to deprive any individual of employment opportunities ... because of such individual's race, color, religion, sex, or national origin.” 42 U.S.C. § 2000e–2(a)(2). Even “practices that are not intended to discriminate but in fact have a disproportionately adverse effect on minorities (known as ‘disparate impact’)” are unlawful. Ricci, 129 S.Ct. at 2672. “The touchstone is business necessity.” Griggs, 401 U.S. at 431, 91 S.Ct. 849. “If an employment practice which operates to exclude [minorities] cannot be shown to be related to job performance, the practice is prohibited.” Id.
Title VII's disparate-impact provision prohibits employment practices that have the unintentional effect of discriminating based on race. See Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 987, 108 S.Ct. 2777, 101 L.Ed.2d 827 (1988); Newark Branch, NAACP v. City of Bayonne, N.J., 134 F.3d 113, 121 (3d Cir.1998). The statute provides:
An unlawful employment practice based on disparate impact is established ... only if ... a complaining party demonstrates that a respondent uses a particular employment practice that causes a disparate impact on the basis of race, color, religion, sex, or national origin and the respondent fails to demonstrate that the challenged practice is job related for the position in question and consistent with business necessity[.] 42 U.S.C. § 2000e–2(k)(1)(A)(i). “By enacting § 2000e–2(k)(1)(A)(i), Congress allowed claims to be brought against an employer who uses a practice that causes disparate impact, whatever the employer's motives and whether or not he has employed the same practice in the past.” Lewis v. City of Chicago, Ill., ––– U.S. ––––, 130 S.Ct. 2191, 2200, 176 L.Ed.2d 967 (2010).
Disparate-impact litigation proceeds in three steps. First, a plaintiff must establish a prima facie case by “demonstrat[ing] that application of a facially neutral standard has caused a ‘significantly discriminatory hiring pattern.’ ” Bayonne, 134 F.3d at 121 (quoting Newark Branch, NAACP v. Town of Harrison, N.J., 940 F.2d 792, 798 (3d Cir.1991)). This prima facie showing requires the plaintiff to prove a significant statistical disparity and to “demonstrate that the disparity [he] complain[s] of is the result of one or more of the employment practices that [he is] attacking.” Id. (quoting Wards Cove Packing Co., Inc. v. Atonio, 490 U.S. 642, 657, 109 S.Ct. 2115, 104 L.Ed.2d 733 (1989)).
The employer may defend against a prima facie showing of disparate impact only by demonstrating that the challenged practice is “job related for the position in question and consistent with business necessity.” See 42 U.S.C. § 2000e–2(k)(1)(A)(i). We have interpreted the business-necessity defense to mean that employers may not use criteria which have a discriminatory effect unless those criteria define the minimum qualifications necessary to perform the job. See Lanning v. Se. Pa. Transp. Auth., 181 F.3d 478, 481, 489 (3d Cir.1999) (“[I]n order to show the business necessity of a discriminatory cutoff score an employer must demonstrate that its cutoff measures the minimum qualifications necessary for successful performance of the job in question.”). The employer bears the burdens of production and persuasion, see id. at 487, and must assert actual reasons why the challenged employment practice is important to the position; “the mere assertion of ... conceivable bases is not sufficient,” Harrison, 940 F.2d at 804. The Supreme Court has “refused to accept bare or ‘common-sense'-based assertions of business necessity and instead require[s] some level of empirical proof that challenged hiring criteria accurately predict[ ] job performance.” El v. Se. Pa. Transp. Auth., 479 F.3d 232, 240 (3d Cir.2007). “[E]mployers cannot rely on rough-cut measures of employment-related qualities; rather they must tailor their criteria to measure those qualities accurately and directly for each applicant.” Id. Ultimately, if the employer cannot present compelling evidence of business necessity, “the plaintiff wins simply by showing the stated elements.” Lewis, 130 S.Ct. at 2198.
Prior to the enactment of the Civil Rights Act of 1991, codified at 42 U.S.C. § 2000e–2(k), the Supreme Court held that an employment practice did not constitute “an artificial, arbitrary, or unnecessary barrier, [so long as] it measured skills related to effective performance in [the job],” Connecticut v. Teal, 457 U.S. 440, 451, 102 S.Ct. 2525, 73 L.Ed.2d 130 (1982), and required an employer to show only that “the practice further[ed] legitimate business goals ‘in a significant way,’ ” Harrison, 940 F.2d at 803 (quoting Wards Cove, 490 U.S. at 659, 109 S.Ct. 2115). Our current, more stringent “minimum qualifications” standard comports with the legislative repeal of the Supreme Court's decision in Wards Cove.
Finally, a plaintiff can overcome an employer's business-necessity defense by showing that alternative practices would have less discriminatory effects while ensuring that candidates are duly qualified. See 42 U.S.C. § 2000e–2(k)(1)(A)(ii), (C). Proving a less discriminatory, viable alternative requires supporting evidence. One witness's testimony identifying potential alternatives, standing alone, “does not raise a genuine issue of material fact that [the alternative is] available to the [employer] ... and that [it] would have produced less adverse impact.” Ricci, 129 S.Ct. at 2680.
B
Statistical disparities alone can raise an inference of causation, but only when those disparities are substantial and the statistical evidence is reliable. Bayonne, 134 F.3d at 121. In showing statistical disparity, the relevant comparison is “between the racial composition of [the at-issue jobs] and the racial composition of the qualified ... population in the relevant labor market.” Id. (alterations in original) (quoting Wards Cove, 490 U.S. at 650, 109 S.Ct. 2115) (internal quotation marks omitted). Two of our prior decisions guide our analysis.
In Harrison, we considered a residency requirement's effect on minority firefighter and police officer applicants. We found significant statistical proof of disparate impact, even though the African–American population of Harrison was only 0.2%, because no African–American had ever held a uniformed position while Harrison's private work force was 22% African–American. 940 F.2d at 796. “[W]here black persons represent[ed] 0% of Harrison's workforce, it matter[ed] little whether there [wa]s a disparity of 30%, 20% or even 10% or 15%.” Id. at 800 (quoting NAACP, Newark Branch v. Town of Harrison, N.J., 749 F.Supp. 1327, 1340 (D.N.J.1990)). Potential causes for the disparity, other than the residency requirement, had been eliminated. “There [we]re multiple means to travel quickly between most of Essex County and Harrison,” and “qualified black persons would seek positions with Harrison's municipal government” if the residency requirements were removed, so the disparity could not be attributed to a lack of interest or an unwillingness to commute. Id. at 797 (quoting Harrison, 749 F.Supp. at 1341) (internal quotation marks omitted).
In Bayonne, we also examined the characteristics and degree of statistical evidence required to make out a prima facie disparate impact claim. In response to a lawsuit filed by the NAACP, Bayonne suspended its residency requirement for four years. Bayonne, 134 F.3d at 115. During the suspension, the number of African–American firefighters in Bayonne did not increase, and the number of African–American police officers actually decreased. Id. When the four-year suspension ended, Bayonne reinstated its residency requirement. In its challenge to the reinstatement, the NAACP presented statistical evidence showing that African–Americans comprised only 2.6% of the non-residents and only 5.5% of the residents hired as firefighters during the suspension of the residency requirement. Id. at 118. Compared with the 3.5% African–American representation in Bayonne's municipal workforce, however, it “would [have been] surpris[ing] if that [difference] was statistically significant.” Id. at 118 n. 10 (third alteration in original) (quoting the NAACP's expert). Nevertheless, the NAACP's expert opined that the residency requirement was discriminatory. Id. The district court denied the NAACP's request to enjoin the residency requirement, and we affirmed. Id. at 115. As the district court explained, real statistical data were lacking:
[W]e have had for the last day and a half evidence produced by David Griffin primarily which consists, with all due respect, of statistical possibilities relying upon data wherein he assumes that Hudson County is the employment market by which comparisons shall be made as to whether blacks are being disparately impacted by the residency ordinance that was recently reinstated by the municipality of Bayonne....
With due respect to the doctor, his opinions, as he admits in one circumstance, are nothing more than his expression of common sense. They are speculative. They involve speculative contingencies and possibilities without any evidential basis.
...
We have well intentioned statistical platitudes. Id. at 119 (emphasis added).
On appeal, we concluded that causation was lacking because four years of data showed no change in the African–American employment ratio while the residency requirement was suspended. Moreover, the NAACP's claim that an expanded eligibility list would contain more African–Americans was insufficient to establish that the residency requirement was causing the disparity because Bayonne was required to hire in rank order from the NJDOP's certified list; it was not allowed to select freely from the entire pool of candidates who had passed the test. Id. at 122.
Unlike in Bayonne, the NAACP Plaintiffs have provided ample evidence of not only a statistical disparity, but also a causal connection. In more than a decade since North Hudson's inception, it has hired only two African–American firefighters (0.62% of its firefighters), despite an African–American population of 3.4%. For the reasons we explained in Harrison, a minority workforce representation that low suggests discrimination.
That suggestion is borne out by both expert reports in this case. They prove NAACP Plaintiffs' prima facie case, and they do so with non-speculative data that do not suffer from the deficiencies of the expert testimony in Bayonne. Wright's comparison of the proportion of African–Americans employed in Tri–County Area protective service positions (37.4%) with the proportion of African–Americans employed as firefighters by North Hudson (0.62%) shows a disparity that raises an inference of causation. It suggests that North Hudson should employ sixty-five African–American firefighters, and it employs only two.
Although North Hudson now contests Wright's definition of the relevant labor market to include the Tri–County Area and even the entire state of New Jersey, it offers no alternative analysis to explain why the market should be defined more narrowly. On the other hand, as the District Court acknowledged, Wright bolstered his definition of the labor market by pointing to commute times that do not exceed the average for North Hudson residents and the tendencies of those in this type of employment to seek positions statewide.
North Hudson also objects to Wright's definition of the “qualified” population. North Hudson points out an obvious flaw: neither the entire African–American population nor the whole pool of African–American state and local government employees can be considered qualified for a firefighter position. In Siskin's words, Wright “never attempts to control for differences in the likelihood of being qualified for the position, as determined by the results of the [NJDOP] process for firefighter eligibility.” See Bayonne, 134 F.3d at 123 n. 21 (“[L]ooking to the general population is not necessarily sufficient in situations ... where the claim involves jobs with ‘special qualifications.’ ” (citing Hazelwood Sch. Dist. v. United States, 433 U.S. 299, 309, 97 S.Ct. 2736, 53 L.Ed.2d 768 (1977))). But this argument is unavailing to North Hudson on appeal because the District Court agreed with Siskin on this point and instead based its summary judgment on the protective services comparison.
Wright's analysis of full-time protective service employees—though imperfect because it includes several non-analogous positions—fairly, and as nearly as possible, approximates the pool of qualified African–Americans. See, e.g., Green v. USX Corp., 896 F.2d 801, 804 (3d Cir.1990) ( “[W]here labor market statistics are unavailable, ‘certain other statistics—such as measures indicating the racial composition of “otherwise-qualified applicants” for at-issue jobs—are equally probative for this purpose.’ ” (quoting Wards Cove, 490 U.S. at 650, 109 S.Ct. 2115)); Carpenter v. Boeing Co., 456 F.3d 1183, 1197 (10th Cir.2006) (“To be sure, the population selected for statistical analysis need not perfectly match the pool of qualified persons. Such perfection may be impossible to obtain. When reliable data regarding that pool are unavailable, a different population may be used if it adequately reflects the population of qualified persons.”); cf., e.g., Peightal v. Metro. Dade Cnty., 26 F.3d 1545, 1554–55 (11th Cir.1994) (finding a statistical comparison of the composition of the general 18–to–55–year–old population with the composition of the firefighter workforce sufficient after concluding that firefighter is an entry-level position). But cf., e.g., Carpenter, 456 F.3d at 1197–99 (finding insufficient disparate-impact evidence where the expert's “qualified population” failed to account for several overtime eligibility requirements specified in the controlling collective bargaining agreement and the expert did not explain why proxy measures (like “job” and “grade”) adequately approximated the actual requirements).
The NAACP Plaintiffs explain—and North Hudson and the Intervenors do not contest—that there are no EEOC–4 data breaking down the full-time protective service positions into smaller clusters from which Wright might have been able to select employment data for a more closely analogous set of positions.
The full-time protective service positions are a sufficient proxy even though firefighting is a specialized job. Many of the full-time protective service positions require emergency medical training, physical fitness, calmness under pressure, and strategic decisionmaking in emergencies. Though none of the included positions—other than firefighter, of course—encompasses the full breadth of special skills required for firefighting, the law does not demand a perfect analog. Cf., e.g., Dothard v. Rawlinson, 433 U.S. 321, 330–31, 97 S.Ct. 2720, 53 L.Ed.2d 786 (1977) (finding disparate impact based on a comparison of height and weight requirements for prison guards with height and weight statistics based on national data and explaining that “reliance on general population demographic data was not misplaced where there was no reason to suppose that physical height and weight characteristics of Alabama men and women differ markedly from those of the national population”); Harrison, 940 F.2d at 799–800 (finding disparate impact based on a comparison of the percentage of African–Americans in Harrison's public uniformed positions with that in its private workforce). We agree with the District Court that the statistics yielded by Wright's comparisons still show a sufficient statistical disparity to establish a prima facie case of disparate impact.
As the District Court concluded, North Hudson's own expert not only failed to create a real dispute regarding Wright's findings, but himself bolstered the causal link between the residency requirement and the disparity demonstrated in Wright's calculations. For each of the three years studied, Siskin's tables indicate that if North Hudson had hired from a Tri–County List rather than its Residents–Only List, it would almost certainly have hired many more African–Americans. If North Hudson's recent hiring needs are any indication, it might seek to hire thirty-five to forty firefighters in a given exam cycle. Siskin's hypothetical lists for the Tri–County Area predicted that between eight and sixteen of those hires would have been African–American in 1999, 2002, and 2006. Even assuming that African–Americans were more likely to receive and accept firefighter positions in their hometowns during those years, six to twelve African–American candidates still would have placed within the top thirty-five candidates available to North Hudson.
Siskin's only data suggesting that African–Americans would not have been hired from a Tri–County List appear in Tables 7–2006 and 8–2006, which inexplicably added an assumption which had not been part of the 1999 and 2002 lists. In these lists, Siskin eliminated candidates who were hired by other municipalities that year as well as candidates who ranked “substantially better” on another municipality's eligibility list. As the District Court noted, “the assumption underlying Table 8–2006 may not bear out in reality.” North Hudson, 742 F.Supp.2d at 518. The 1999 and 2002 hypothetical lists eliminating those candidates who were actually hired by other municipalities showed that six to twelve African–Americans ranked in the top thirty-five would still have been available for hiring by North Hudson. Assuming that a 2006 table eliminating only those actually hired by other municipalities would have yielded similar results, Siskin's data suggest that all the available African–Americans ranked thirty-fifth and above in that type of 2006 table were ranked seventeenth or above on other municipalities' 2006 lists but did not receive jobs. The additional elimination from Tables 7–2006 and 8–2006 of those who were better ranked on other municipalities' lists but who were not ultimately hired purportedly would have removed all African–American candidates from the top ninety, thus making none available to North Hudson. Siskin's adjustment of the data is not accompanied by any explanation for why the adjustment was necessary in 2006 but not in 1999 or 2002, why his definition of “substantially better” rank yields accurate predictions of hometown preferences, or why candidates who were not ultimately hired elsewhere would have been disinterested in working for North Hudson.
In light of the ample evidence of record, including Siskin's other expanded lists, the District Court did not err in concluding that no genuine dispute of material fact exists as to whether North Hudson's residency requirement creates a disparate impact on African–American firefighter applicants.
C
Having found no error in the District Court's conclusion that the NAACP Plaintiffs proved a prima facie case of disparate impact, we turn to its rejection of North Hudson's defense that business necessity justifies the residency requirement. As outlined above, we have interpreted the business-necessity defense to apply only when an employer can show that its challenged hiring criteria define minimum qualifications for the position. See Lanning, 181 F.3d at 481, 489 (holding that strong aerobic ability was not a necessary qualification for transit officers when many incumbent officers who never passed the challenged test later received commendations and promotions). Even if business necessity is shown, the plaintiff will prevail if there are less discriminatory alternative means of selecting for the crucial qualification. See, e.g., id. at 485.
In Harrison, we rejected the town's purported justifications for hiring firefighters from a residents-only eligibility list. We held that although emergency preparedness, i.e., the speed with which an employee firefighter could respond to an emergency call, was potentially crucial to the position, there were alternative ways to ensure short response times, such as establishing an acceptable response distance. Harrison, 940 F.2d at 804–05. We also held that ensuring loyalty to the Harrison community was not an essential qualification. See id. at 805. Similarly, we rejected Harrison's plea to avoid an expensive and time-consuming increase in applications because that inconvenience was not actually related to firefighting. See id. at 804–05.
In the District Court, North Hudson reprised several of the arguments we found wanting in Harrison. North Hudson claimed that residency is essential to its fire department operations because it: (1) increases the likelihood that its firefighters will be able to respond to emergencies more quickly because they will be more familiar with the buildings and streets in the served community; (2) comports with the terms of the Rodriguez Settlement; (3) increases the number of Spanish-speaking firefighters in a department that serves a 69% Hispanic population; and (4) fosters community pride.
We have no quarrel with the notion that a critical aspect of firefighting is the ability to respond quickly and that familiarity with the streets and buildings of a locale is important to achieving that goal. But this valid point cannot be reconciled with the fact that North Hudson does not require its firefighters to reside in the North Hudson municipalities after they are hired. In fact, in 2008, only 34% to 36% of North Hudson firefighters were residents. See North Hudson, 742 F.Supp.2d at 523 & n. 18.
Nor does the Rodriguez Settlement provide a valid business justification for the residency requirement. Although that settlement encourages minority hiring and promotion, it does not compel the use of a residency requirement; it merely requires targeted advertising of exams. The other provisions of the Rodriguez Settlement were satisfied years ago when certain remedies were provided to the particular plaintiffs in that case. Furthermore, North Hudson cannot claim that the residency requirement is part of its attempt to comply with the broad purpose of the Rodriguez Settlement because the residency requirement was in place before that lawsuit was filed.
As for employing a certain number of Spanish-speaking firefighters in a region that is 69% Hispanic, this is a plausible justification in the abstract. But North Hudson failed to establish that the residency requirement leads to a greater number of Spanish-speaking firefighters. Rather, this purported justification is a “conceivable” basis, which is insufficient to invoke the business-necessity defense. See Harrison, 940 F.2d at 804. Moreover, as the District Court noted, there are non-discriminatory ways to ensure the hiring of Spanish-speaking firefighters. Rather than seek out Spanish-speakers by making the imprecise assumption that North Hudson residents are more likely to speak Spanish, North Hudson could, and actually does, seek out bilingual candidates.
Finally, as Harrison and the business-necessity standard itself make clear, community pride is not a sufficient justification for a discriminatory hiring practice.
In sum, the District Court properly concluded that North Hudson's purported business-necessity arguments fail. They are not tied to minimum firefighter qualifications in North Hudson and, in some cases, less discriminatory alternatives are available.
D
North Hudson argues that, even if it cannot show business necessity, the Supreme Court's decision in Ricci offers it safe harbor. Ricci established a defense for defendants in disparate-treatment suits who can show a “strong basis in evidence” to believe that failing to engage in the discriminatory action being challenged would lead to disparate-impact liability. See 129 S.Ct. 2658. North Hudson seeks to establish a defense in the converse situation, where an employer is charged with disparate-impact discrimination but fears disparate-treatment liability if it ceases the employment practice that is causing the disparate impact. Specifically, North Hudson asserts that removing its residency requirement to eliminate the disparate impact on African–Americans would subject it to suit by Hispanics alleging disparate treatment.
The distinction between disparate impact and disparate treatment lies in the disparate-treatment requirement that the discrimination be intentional. Title VII's disparate-treatment provision forbids an employer “to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual's race, color, religion, sex, or national origin.” 42 U.S.C. § 2000e–2(a)(1). Disparate treatment “occur[s] where an employer has ‘treated [a] particular person less favorably than others because of’ a protected trait.” Ricci, 129 S.Ct. at 2672 (first alteration in original) (quoting Watson, 487 U.S. at 985–86, 108 S.Ct. 2777). The plaintiff must prove that the employer had a “discriminatory intent or motive.” Watson, 487 U.S. at 986, 108 S.Ct. 2777. Where the employment action was based on a non-discriminatory reason, disparate treatment is, by definition, lacking.
In Ricci, 118 firefighters in New Haven, Connecticut, took lieutenant and captain examinations to qualify for promotions. 129 S.Ct. at 2664. When the exam results showed that no African–American candidate was eligible for promotion for failure to place in the top ten of either exam, the city faced competing threats of suit. Id. at 2664–66. Several minority firefighters claimed the exam results evidenced a discriminatory test. Id. at 2664. Others threatened litigation if the city discarded the results or denied promotion to the top performers on the exam. Id. After New Haven Civil Service Board hearings and discussions with city attorneys, New Haven decided to nullify the exam scores. Id.
The white firefighters who were denied promotions following the nullification of the test results sued, alleging disparate treatment. The district court granted and the Second Circuit upheld summary judgment for New Haven. Id. at 2671–72. The Supreme Court reversed because New Haven's nullification of the test results was a decision based on race, namely “too many whites and not enough minorities would [have been] promoted were the lists to be certified.” Id. at 2672–73 (quoting Ricci v. DeStefano, 554 F.Supp.2d 142, 152 (D.Conn.2006)).
The Court considered “whether the purpose to avoid disparate-impact liability excuses what otherwise would be prohibited disparate-treatment discrimination.” Id. at 2674. The Court rejected the argument that an employer's mere good-faith belief that it must discriminate against individuals to avoid disparately impacting a protected group should be sufficient to shield the employer from disparate-treatment liability. Id. at 2675 (“[W]hen Congress codified the disparate-impact provision in 1991, it made no exception to disparate-treatment liability for actions taken in a good-faith effort to comply with the new, disparate-impact provision in subsection (k).”). Instead, the Court adopted the “strong basis in evidence” standard, which excuses an employer from disparate-treatment liability only if the employer shows a credible possibility of disparate-impact liability if it were not to engage in the intentional discrimination being challenged. Id. at 2676. The Court's holding was specific:
We hold only that, under Title VII, before an employer can engage in intentional discrimination for the asserted purpose of avoiding or remedying an unintentional disparate impact, the employer must have a strong basis in evidence to believe it will be subject to disparate-impact liability if it fails to take the race-conscious, discriminatory action. Id. at 2677; see also id. at 2664 (“We conclude that race-based action like the City's in this case is impermissible under Title VII unless the employer can demonstrate a strong basis in evidence that, had it not taken the action, it would have been liable under the disparate-impact statute.”).
Complying with the Supreme Court's mandate, New Haven certified the exam results and promoted white firefighters consistent with the ranked list of eligible candidates. Id. As expected, an adversely affected African–American firefighter sued, alleging that the test had a discriminatory effect, primarily because its 60/40 weighting of the written and oral portions of the test conflicted with the industry norm of 30/70. See Briscoe v. New Haven, 654 F.3d 200, 201 (2d Cir.2011). New Haven claimed that it could not be held liable because (1) the Supreme Court's resolution of the viability of the prospective African–American firefighters' disparate-impact claim as part of its Ricci decision precluded the suit, and (2) there was a “strong basis in evidence” that failing to certify the test results would lead to disparate-treatment liability. Id. at 202–03. The Second Circuit refused to extend Ricci's “strong basis in evidence” defense to the disparate-impact suit against New Haven. Id. at 205–09. We likewise see no reason to extend Ricci's “strong basis in evidence” defense to the NAACP Plaintiffs' disparate-impact suit against North Hudson.
Here, there can be no doubt that removing the residency requirement will adversely affect Intervenors. They will lose their high rankings on North Hudson's candidate hiring list and find it significantly more difficult to secure a firefighter position with North Hudson. For example, if, as the NAACP Plaintiffs advocate, North Hudson hires from a list incorporating candidates from nearby Essex and Union counties, Intervenors likely would not even rank within the top 180 candidates. The Intervenors currently ranked twenty-first, twenty-fifth, and twenty-sixth would drop to 189th, 261st, and 292nd respectively, and those ranked forty-fifth, forty-ninth, and seventieth would not rank within the top 400 candidates.
Nevertheless, the facts and claims in this case are dissimilar to those in Ricci. In Ricci, New Haven had already administered the purportedly illegal exam (the basis for the disparate-impact claim) and attempted to remedy the exam results by denying ten white firefighters the right to promotions (the basis for the disparate-treatment claim). New Haven had to choose between irreconcilable alleged errors before it secured judicial guidance regarding the merits of the competing discrimination claims. Here, North Hudson faces no such quandary. At this juncture, North Hudson's only action is the use of its Residents–Only List, which has been demonstrated to cause a disparate impact in violation of Title VII. It has taken no steps to eliminate the residency requirement or otherwise adjust its policies to reduce the adverse effect. Thus, North Hudson faces a classic disparate-impact claim, one that we have resolved based on the three-step inquiry dictated by the statute.
Moreover, North Hudson has no basis for believing it will be liable to Intervenors or other North Hudson resident applicants under a disparate-treatment theory. First, this Court, rather than North Hudson, is responsible for eliminating the residency requirement. A government employer's compliance with a judicial mandate does not constitute an official policy or employment practice of the employer, see, e.g., Wolfe v. City of Pittsburgh, 140 F.3d 236, 240 (3d Cir.1998), and it is an employer's deliberate discrimination that the disparate-treatment provision of Title VII prohibits. Removal of the residency requirement can hardly be viewed as a race-based decision when it is motivated by the imperative to comply with a judicial order. Besides, North Hudson would have a “legitimate, nondiscriminatory reason” for hiring more broadly, namely, the superior merit of higher-ranked non-resident applicants. See McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973) (articulating the standards for disparate-treatment liability). As the NAACP Plaintiffs correctly point out, Intervenors will not be removed from the eligibility lists; rather, they will compete based on objective criteria against candidates who were unlawfully excluded from the Residents–Only List. Cf., e.g., Harrison, 940 F.2d at 810 (rejecting amicus curiae's due process claims that forcing them to compete with non-resident firefighter candidates would reduce their chances of being hired).
Finally, North Hudson's repeated reliance on the possibility that it will face a lawsuit by its Hispanic resident candidates is unavailing. “Fear of litigation alone” does not suffice; a demonstrated potential for liability is required. Ricci, 129 S.Ct. at 2681.
In sum, we conclude that the District Court properly granted summary judgment to the NAACP Plaintiffs on their disparate-impact claim. The NAACP Plaintiffs presented sufficient evidence to establish that North Hudson's residency requirement causes a disparate impact by excluding well-qualified African–Americans who would otherwise be eligible for available firefighter positions. North Hudson failed to present evidence to create any genuine dispute regarding this disparate impact or adduce a valid business necessity for the residency requirement. And Ricci is unavailing to North Hudson.
IV
North Hudson and Intervenors claim that the inequities Hispanics will face if the residency requirement is removed counsel against imposing a permanent injunction against the use of the Residents–Only List. In particular, North Hudson and Intervenors point to residency requirements compelled by the Consent Decree in many of the municipalities surrounding North Hudson. Because of those residency requirements, North Hudson residents will be eligible for employment only in North Hudson while residents of other jurisdictions will be eligible in their hometowns and in North Hudson. In other words, North Hudson will be forced to open its hiring to non-residents while other municipalities exclude those, like Intervenors, who reside within North Hudson. North Hudson and Intervenors certainly have a point in this regard. But we have no authority to endorse discrimination against firefighter candidates who do not live in North Hudson in order to protect those who do. Once a court identifies racial discrimination, it must order relief that will remedy past discrimination and curb the potential for future discrimination. See Albemarle Paper Co. v. Moody, 422 U.S. 405, 418, 95 S.Ct. 2362, 45 L.Ed.2d 280 (1975) (quoting Louisiana v. United States, 380 U.S. 145, 154, 85 S.Ct. 817, 13 L.Ed.2d 709 (1965)).
Although district courts are afforded substantial discretion in fashioning injunctive relief, it should not be broader than required to provide a full remedy to the injured party. See Meyer v. CUNA Mut. Ins. Soc'y, 648 F.3d 154, 170 (3d Cir.2011). North Hudson and Intervenors do not argue that the District Court's permanent injunction exceeds the bounds of its discretion or provides relief that is disproportional to the harms the NAACP Plaintiffs have proven. Nor could they. The permanent injunction against use of the Residents–Only List is properly circumscribed to eliminate the employment practice that the expert reports establish is causing the disparate impact.
Title VII cases demonstrate that inequities to one group accruing from remedies for discrimination against another group cannot forestall those remedies. As we have explained, “Title VII itself clearly makes each employer responsible for its own employment practices, and does not support the proposition that one employer is free to discriminate against a class of employees as long as other employers are willing to hire them.” Harrison, 940 F.2d at 800 (internal citation omitted); see also id. at 801 (“[R]esidency requirements imposed by other municipalities ... were appropriately excluded from the district court's definition of the relevant labor market.”). The unavailability of firefighter positions outside North Hudson to North Hudson residents does not negate the discrimination perpetrated by the exclusion of non-residents from North Hudson positions. Moreover, the municipalities around North Hudson that impose their own residency requirements are not parties in this case, so we cannot adjudicate the legality of their policies. If Intervenors or any other North Hudson residents believe the residency requirements of neighboring communities unlawfully discriminate against them, they remain free to challenge those employment practices. Accordingly, we will affirm the District Court's injunction.
Intervenors claim entitlement to attorney's fees for influencing the District Court in April 2010 to vacate the preliminary injunction it granted in February 2009. The District Court concluded that Intervenors should not receive attorney's fees. We see no abuse of discretion in the District Court's decision. A party prevails when it secures a “material alteration of the legal relationship of the parties in a manner which Congress sought to promote in the fee statute.” Tex. State Teachers Ass'n v. Garland Indep. Sch. Dist., 489 U.S. 782, 792–93, 109 S.Ct. 1486, 103 L.Ed.2d 866 (1989). Here, the District Court's vacatur of the preliminary injunction was based on a “preliminary finding” that the residency requirement might be related to North Hudson's business objectives and that equitable principles disfavored maintaining an injunction (an extreme form of relief) while a full consideration of the disparate-impact claim was undertaken. See North Hudson, 742 F.Supp.2d at 507. This type of temporary victory, especially when it is followed by the permanent injunction Intervenors opposed, is not a “material alteration of [the parties'] legal relationship.” Cf., e.g., Sole v. Wyner, 551 U.S. 74, 82–83, 127 S.Ct. 2188, 167 L.Ed.2d 1069 (2007) (holding that one who obtains a preliminary injunction but fails to secure a permanent injunction is not a prevailing party); West v. Keve, 721 F.2d 91, 94 (3d Cir.1983) (“If the fee award is based on a ‘prevailing party’ standard ... then an appellate court's reversal on the merits may well require a redetermination of the fee issue.” (internal citation omitted)).
V
For the reasons stated, we will affirm the District Court's judgment in all respects.