On Appeal from the United States District Court for the District of New Jersey (D.C. No. 07-cv-01683) District Judge: Honorable Dickinson R. Debevoise
The opinion of the court was delivered by: Hardiman, Circuit Judge.
Argued September 20, 2011
Before: FISHER, HARDIMAN and GREENAWAY, JR., Circuit Judges.
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.
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).*fn1 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 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.*fn2
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.*fn3 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, 22 A.3d 958, 959 (N.J. 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.
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, fortyninth, 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, 129 S. Ct. 2658 (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).*fn4
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 disparateimpact claims depend heavily on statistical proof of discriminatory effects, we review the experts' findings and the District Court's conclusions at length.
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]county [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."*fn5 The District Court agreed and found that "[t]he difference ...