Plaintiff-appellant Jesus Malave, an Hispanic male employee of the United States Postal Service ("Postal Service"), appeals from a summary judgment entered in the Unites States District Court for the District of Connecticut (Eginton, J.) in favor of defendant-appellee, Postmaster General of the Postal Service, dismissing Malave's disparate impact employment discrimination claim. In dismissing Malave's claim, the District Court found that the undisputed statistical analysis provided by Malave's expert failed to establish a prima facie case of disparate impact. In particular, the District Court held that, to make out a prima facie case of disparate impact, Malave's statistical analysis should have focused on the "applicant pool or the eligible labor pool" for the "at-issue" positions, rather than on the "overall number of Hispanics in the Connecticut Postal Service workforce." For the reasons that follow, we hold that such a per se rule is not appropriate in cases (such as this one) where the data concerning the number of qualified Hispanics who applied for the at-issue promotions are not available. Accordingly, we vacate the summary judgment and remand for proceedings consistent with this opinion.

The undisputed facts giving rise to Malave's appeal from the District Court's summary judgment are as follows: Malave is an Hispanic male who has been serving since 1994 as the Postmaster for Norfolk, Connecticut, an Executive and Administrative Schedule ("EAS")-15 level position.1 In 1996, Malave twice unsuccessfully applied for a promotion to the position of Manager of Customer Services, an EAS-21 level position, at the Murphy Road Post Office in Hartford, Connecticut. Between 1996 and 1997, Malave unsuccessfully applied for promotions to three EAS-20 level Postmaster positions in Connecticut, each of which was at a larger facility than the one in Norfolk. Each time Malave failed to receive a promotion, the successful candidate was not Hispanic.

In November 1998, after exhausting his administrative remedies,2 Malave commenced the action giving rise to this appeal in the District Court. He twice amended his original complaint. The gravamen of Malave's Second Amended Complaint is that, in continuously failing to promote him, the Postal Service discriminated against him on the basis of his race, in violation of Title VII of the Civil Rights Act of 1964, 42 U.S.C. §§ 2000e et seq.3 Malave claimed that the Postal Service discriminated generally against Hispanics in regard to promotions. In particular, he alleged that, "[o]f the more than 1,000 EAS-15 and above-grade positions" that existed within the Postal Service in Connecticut at the time he filed his complaint, "only eight (8) [were] occupied by Hispanics, despite a large pool of professionally qualified individuals" and that, "from a national perspective, roughly only 4.5% of the management employees within the Postal service [who] are at grade levels from EAS-15 through [EAS-]23 are Hispanic."

At the close of discovery, the Postmaster General moved for summary judgment, arguing inter alia, that Malave had failed to make out a prima facie disparate impact case. Malave opposed the Postmaster General's summary judgment motion and included with his opposition an expert's report containing a statistical analysis purporting to show that Hispanics were underrepresented in the upper-level management of the Postal Service's Connecticut workforce and that this underrepresentation was likely caused by discrimination. In particular, the report analyzed data regarding the ethnic composition of Connecticut's Postal Service employees for the period 1996 to 2000. The report, prepared by the firm of Bobo, Jaynes & McKinney, broke out these data into three groups by EAS level. The first group included EAS levels 1 through 14, the second group included EAS levels 15 through 18, and the third group included all EAS levels above 18. The report concluded that there were "large discrepancies between the number of Hispanic Americans actually occupying the two upper [groups] of EAS levels and the numbers that would be expected to occupy those [groups] if promotion opportunities were unbiased." Furthermore:

[I]n four out of the five years (1997 being the exception) the difference between the actual number of Hispanics in the various [groups] and the expected number of Hispanics in [the upper EAS groups] was statistically significant at the ten percent level. In one out of the four years, 1999, the difference between the actual representation of Hispanic Americans at the upper EAS level and the number expected from an unbiased promotion process was statistically significant at the five percent level. Based on these statistics we conclude that Hispanic Americans were statistically underrepresented at the higher EAS levels with the [Postal Service] in Connecticut during the late 1990s.

The report went on to conclude that Hispanics were underrepresented in the upper-level EAS positions of the Postal Service as a result of a selection process "contaminated with favoritism," where "both the pressures of friendship and ethnic loyalty operate to ensure that without formal objective procedures for identifying successful applicants, review boards and supervisors composed of individuals with largely homogenous backgrounds will as a matter of course make decisions that simply replicate the ethnic composition of the Postal Service already in existence." The Postmaster General did not attempt to rebut the report of Malave's expert with the report of an expert of his own and instead argued that the report of Bobo, Jaynes & McKinney did not establish a prima facie case.

In an unpublished ruling dated November 2, 2001, the District Court granted the Postmaster General's summary judgment motion on the ground that Malave had failed to make out a prima facie disparate impact claim. In reaching this conclusion, the District Court cited the Supreme Court's decision in Wards Cove Packing Co. v. Atonio, 490 U.S. 642, 109 S. Ct. 2115, 104 L. Ed. 2d 733 (1989), and our decision in Waisome v. Port Authority of New York & New Jersey, 948 F.2d 1370 (2d Cir.1991), for the proposition that, in disparate impact cases, "the proper population for [statistical] analysis is the applicant pool or the eligible labor pool." Thus, the District Court opined that "the applicant pool or the number of eligible Hispanic candidates who applied for promotion is the relevant consideration rather than the overall number of Hispanics in the Connecticut Postal Service workforce." The District Court reasoned that, because Malave's expert provided "no information on how many Hispanics ha[d] applied for promotions to upper[-]level positions," the disparity noted by the expert "may well be attributed to the lack of qualified Hispanic candidates applying for positions rather than the challenged promotional practices." Consequently, the District Court concluded that Malave's "statistical evidence [was] not probative of the existence of a disparate impact" and Malave had failed as a matter of law to make out a prima facie disparate impact claim.

Final judgment was entered on November 7, 2001. Malave subsequently moved for reconsideration of the District Court's decision. The District Court granted the motion in February 2002 and, on reconsideration, adhered to its original decision. This timely appeal followed.

We review de novo the District Court's grant of summary judgment, construing the evidence in the light most favorable to Malave and drawing all reasonable inferences in his favor. See Mount Vernon Fire Ins. Co. v. Belize NY, Inc., 277 F.3d 232, 236 (2d Cir.2002).

"Title VII prohibits overt and intentional discrimination as well as discrimination resulting from employment practices that are facially neutral, but which have a `disparate impact' because they fall more harshly on a protected group than on other groups and cannot otherwise be justified." Waisome, 948 F.2d at 1374. Thus, "disparate impact theory targets `practices that are fair in form, but discriminatory in operation.'" Smith v. Xerox Corp., 196 F.3d 358, 364 (2d Cir.1999) (quoting Griggs v. Duke Power Co., 401 U.S. 424, 431, 91 S. Ct. 849, 28 L. Ed. 2d 158 (1971)). To make out a prima facie case of disparate impact, a plaintiff must "(1) identify a policy or practice, (2) demonstrate that a disparity exists, and (3) establish a causal relationship between the two." Robinson v. Metro-North Commuter R.R. Co., 267 F.3d 147, 160 (2d Cir.2001).

Statistical data may be submitted "to show a disparity in outcome between groups, but to make out a prima facie case[,] the statistical disparity must be sufficiently substantial to raise an inference of causation." Smith, 196 F.3d at 365. As we noted in Smith, "[a]lthough no bright line rules exist to guide courts in deciding whether [a] plaintiff['s] statistics raise an inference of discrimination, several overarching principles inform the issue." Id. First, a plaintiff must "identify a specific employment practice, rather than rely on bottom line numbers in an employer's workforce." Id. Second, a plaintiff must "present statistical evidence `of a kind and degree sufficient to show that the practice in question has caused the exclusion of applicants for ... promotions because of their membership in a protected group.'" Id. (quoting Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 994, 108 S. Ct. 2777, 101 L. Ed. 2d 827 (1988)). Third, any statistics relied upon "must be of a kind and degree sufficient to reveal a causal relationship between the challenged practice and the disparity" when combined with other evidence. Robinson, 267 F.3d at 160.

Here, the District Court found that Malave's statistical analysis failed to establish a prima facie case of disparate impact because that analysis did not focus on the "applicant pool or the eligible labor pool" for the "at-issue" promotions, and instead focused on the "overall number of Hispanics in the Connecticut Postal Service workforce." The District Court relied on the Supreme Court's decision in Wards Cove in adopting this per se rule. This reliance on Wards Cove was misplaced, however. In that case, the Court opined that, generally speaking,

the proper comparison [of statistical evidence in a disparate impact case is] between the racial composition of [the at-issue jobs] and the racial composition of the qualified ... population in the relevant labor market. It is such a comparison — between the racial composition of the qualified persons in the labor market and the persons holding at-issue jobs — that generally forms the proper basis for the initial inquiry in a disparate-impact case.

Wards Cove, 490 U.S. at 650-51, 109 S. Ct. 2115 (internal quotation marks omitted). In the context of promotions, we have held that the appropriate comparison is customarily between the composition of candidates seeking to be promoted and the composition of those actually promoted. See Waisome, 948 F.2d at 1372. But the Supreme Court also cautioned in Wards Cove that "in cases where such labor market statistics will be difficult if not impossible to ascertain, we have recognized that certain other statistics — such as measures indicating the racial composition of `otherwise-qualified applicants' for at-issue jobs — are equally probative for this purpose." Wards Cove, 490 U.S. at 651, 109 S. Ct. 2115 (citing New York City Transit Auth. v. Beazer, 440 U.S. 568, 585, 99 S. Ct. 1355, 59 L. Ed. 2d 587 (1979)); see also Smith, 196 F.3d at 368 ("In the typical disparate impact case the proper population for analysis is the applicant pool or the eligible labor pool.") (emphasis added). In addition, the Court noted that when the "`figures for the general population might... accurately reflect the pool of qualified job applicants,' ... [the Court has] even permitted plaintiffs to rest their prima facie cases on such statistics as well." Wards Cove, 490 U.S. at 651 n. 6, 109 S. Ct. 2115. (internal quotations and citations omitted).

At oral argument, counsel for the Postmaster General conceded that data concerning the number of qualified Hispanics who applied for the at-issue promotions are not available. Therefore, it was error for the District Court to have rejected out of hand Malave's statistical analysis simply because it failed to conform to the preferred methodology described in Wards Cove, given the Supreme Court's express endorsement in that decision of alternative methodologies if the preferred statistics are "difficult" or "impossible" to obtain. Accordingly, we must vacate the summary judgment and remand the case to the District Court for proceedings consistent with this opinion.

On remand, the District Court must determine whether Malave makes out a prima facie case of disparate impact. To do so, the District Court must first decide what the most appropriate labor pool is and whether the expert's method utilized the best and most appropriate available labor pool information.4 Next, having determined the relevant labor pool (and therefore what statistics are directly relevant to Malave's prima facie case), the District Court must decide whether Malave's statistical analysis and other evidence is "of a kind and degree sufficient to show" that the Postal Service's promotion practices have "caused the exclusion of applicants for ... promotions" because they are Hispanic. Watson, 487 U.S. at 994, 108 S. Ct. 2777. In the course of making this decision, the District Court should be guided by established principles, beginning with the Supreme Court's teaching that "statistics come in infinite variety and ... their usefulness depends on all of the surrounding facts and circumstances." Id. at 996 n. 3, 108 S. Ct. 2777 (internal quotation marks omitted).

The District Court should also consider whether Malave can show "a statistically significant disparity of two standard deviations." Smith, 196 F.3d at 365. As we observed in Smith, although there is no one test that always answers the question whether a statistical disparity is sufficiently substantial to establish a prima facie case of disparate impact, courts "generally consider this level of significance sufficient to warrant an inference of discrimination." Id. at 365. Furthermore, the District Court must also keep in mind that a plaintiff "generally cannot attack an overall decision making process in the disparate impact context, but must instead identify the particular element or practice within the process that causes an adverse impact." Stout v. Potter, 276 F.3d 1118, 1124 (9th Cir.2002) (citing Wards Cove, 490 U.S. at 656-57, 109 S. Ct. 2115). Thus, a decisionmaking process "may be analyzed as a single employment practice if `the complaining party can demonstrate to the court that the elements of [the employer's] decisionmaking process are not capable of separation for analysis.'" Stout, 276 F.3d at 1124 (quoting 42 U.S.C. § 2000e-2(k)(1)(B)(i)). Of course, if the District Court determines that Malave has made out a prima facie case, it should then address the Postmaster General's remaining arguments in favor of summary judgment and proceed with the case as it deems appropriate.

Accordingly, our decision to remand this case to the District Court should not be construed as anything other than a rejection of the District Court's adoption of a rule that the lack of statistical information as to an applicant pool always renders it impossible to establish a prima facie disparate impact case.

For the foregoing reasons, we vacate the summary judgment entered by the District Court and remand the case for further proceedings consistent with this opinion.