Summary
holding summary judgment "patently inappropriate" when district court restricts plaintiff's ability to discover evidence necessary to establish disparate treatment prima facie case
Summary of this case from Hollander v. American Cyanamid Co.Opinion
No. 83-2652.
Argued and Submitted October 3, 1984.
Decided January 29, 1985.
Kathleen C. Roberts, P.C., Tucson, Ariz., for plaintiff-appellant.
Ruth V. McGregor, Anne L. Tiffen, Fennemore, Craig von Ammon, Udall Powers, Phoenix, Ariz., for defendant-appellee.
Appeal from the United States District Court for the District of Arizona.
Before BROWNING, ANDERSON, and REINHARDT, Circuit Judges.
Paul Silvestre Diaz, a Mexican-American, brought this action under Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e et seq. (1982), alleging that the failure by his employer, American Telephone Telegraph Company (AT T), to promote him to the position of Operations Supervisor resulted from discrimination on the basis of race or national origin. During discovery, AT T refused to provide certain employment statistics. Without ruling on Diaz's motion to compel discovery of these statistics, the district court granted AT T's motion for summary judgment, ruling that, as a matter of law, the promotion of another Mexican-American, Rebecca Gonzales, to the challenged position precluded Diaz from establishing a prima facie case of discrimination. We hold that the summary judgment order was based on an erroneous legal premise. However, we do not review the record to see if we could affirm that order on any other ground because Diaz did not receive, and the record does not contain, all the material to which he was entitled in framing his opposition to the summary judgment motion. The requested statistical material was properly subject to discovery.
I. BACKGROUND
Paul Diaz has been employed since 1969 at the Tucson facility of AT T, Long Lines Division, as a Communications Technician. At various times during his employment Diaz filled in temporarily as an Operations Supervisor. In March 1980, a permanent opening for Operations Supervisor at the Tucson facility became available. Again, Diaz filled in, this time for a period of three and one-half months, from mid-February to June 1, 1980.
As a policy, AT T does not disseminate information about promotional opportunities directly to the potentially eligible employees. When the Operations Supervisor position at the Tucson facility became available, in accordance with AT T's general hiring and promotion procedures the Tucson facility Operation Manager notified the Western Region personnel office of the job opening. Shortly afterwards, the personnel office compiled a list of all current job openings in the Western Region and sent the list to all Western Region supervisors. The supervisors then submitted lists of candidates they thought were qualified for the positions. The names of five candidates were submitted for the Tucson Operations Supervisor position, among them Rebecca Gonzales. The Operation Manager in Tucson, who was responsible for the ultimate promotion decision, was also responsible for selecting qualified candidates from the Tucson facility. He did not consider Diaz for the position although Diaz was deemed "promotable."
Because it became apparent to Diaz that, although he was serving temporarily as the Tucson Operations Supervisor, he was not going to be promoted to the position on a permanent basis, he filed a discrimination charge with the Arizona Civil Rights Division. His charge was filed on April 15, 1980. Thereafter, in late April or May, AT T decided to promote Gonzales to the Operations Supervisor position. Gonzales's promotion, effective in early June 1980, was apparently AT T's first promotion of a Mexican-American to that position at the Tucson facility. Gonzales was subsequently replaced by Peter Todd, a white male, in October 1981, only a year after her appointment.
We reach this interpretation of the facts when, as we are required to do on a motion for summary judgment, we view the evidence in the light most favorable to the plaintiff. See Twentieth Century-Fox Film Corp. v. MCA, 715 F.2d 1327, 1328-29 (9th Cir. 1983). In response to an interrogatory requesting the number and racial identification of Operations Supervisors at the Tucson facility since 1972, AT T provided information only for the period after October 1979. AT T did not provide any information to suggest that it had any Mexican-American Operations Supervisors at Tucson before Gonzales.
II. PROMOTION OF A MEMBER OF PLAINTIFF'S PROTECTED CLASS
In order to prevail in a Title VII disparate treatment case, a plaintiff must first establish a prima facie case of discrimination. The burden of production then shifts to the defendant to articulate a legitimate nondiscriminatory reason for the adverse employment decision. If the defendant carries its burden, the plaintiff is then afforded an opportunity to demonstrate that the "`assigned reason' was `a pretext or discriminatory in its application.'" Lynn v. Regents of the University of California, 656 F.2d 1337, 1341 (9th Cir. 1981) (quoting McDonnell Douglas Corp. v. Green, 411 U.S. 792, 807, 93 S.Ct. 1817, 1827, 36 L.Ed.2d 668 (1973)), cert. denied, 459 U.S. 823, 103 S.Ct. 53, 74 L.Ed.2d 59 (1982). In the present case, the district court did not examine AT T's articulated nondiscriminatory reason — that Gonzales was better qualified — or consider whether Diaz's evidence sufficiently rebutted this defense because it determined that Diaz could not, as a matter of law, establish a prima facie case. Its reason for reaching this conclusion was that the person ultimately hired was a member of the same protected class as Diaz.
Diaz claims that AT T's articulated reason for the decision is a pretext. He has attempted to show that the criteria under which Gonzales might have been "better qualified" do not match AT T's description of the skills most essential for the job and that, because of his own supervisory experience, he was better qualified for the position than Gonzales.
In order to establish a prima facie case, a plaintiff must offer evidence that "give[s] rise to an inference of unlawful discrimination." Texas Department of Community Affairs v. Burdine, 450 U.S. 248, 253, 101 S.Ct. 1089, 1094, 67 L.Ed.2d 207 (1981). One common way a plaintiff can establish an inference of discrimination is by demonstrating that the four requirements of the McDonnell Douglas test are met:
(1) that the plaintiff belongs to a class protected by Title VII;
(2) that the plaintiff applied and was qualified for a job for which the employer was seeking applicants;
(3) that, despite being qualified, the plaintiff was rejected; and
(4) that, after the plaintiff's rejection, the position remained open and the employer continued to seek applicants from persons of comparable qualifications.
See McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 93 S.Ct. 1817, 1824, 36 L.Ed.2d 668 (1973).
Under the traditional McDonnell Douglas test, the "inference of discrimination" is not dependent upon an examination of who, if anyone, was promoted instead of the plaintiff. Ordinarily, the fourth element of McDonnell Douglas, is met whenever the employer continues to consider other applicants whose qualifications are comparable to the plaintiff's after refusing to consider or rejecting the plaintiff. Diaz was deemed promotable but the Operation Manager did not consider him for the position although he subsequently considered comparably qualified applicants. Thus, Diaz satisfied the fourth element of the McDonnell Douglas test as it is ordinarily understood and applied.
Diaz alleged that he was at least as qualified as other candidates and the record does not require us to conclude that Diaz is wrong. Because we are reviewing a summary judgment order, we treat the other candidates as comparably qualified. See supra note 1.
The district court's determination that Diaz failed to establish a prima facie case was based on its implicit conclusion that the Supreme Court's articulation of the McDonnell Douglas test should not be read literally. The district court held that the fourth element of McDonnell Douglas cannot be met unless the challenged position is filled by someone outside the plaintiff's protected class. Thus, the district court concluded that Diaz failed to meet the McDonnell Douglas standard. AT T asks us to affirm the summary judgment order either by endorsing the district court's construction of McDonnell Douglas or by adopting a closely related line of reasoning. AT T's alternative argument is that even if McDonnell Douglas is read as the Supreme Court apparently meant it to be, and even if, therefore, Diaz met that test, it is not sufficient in all instances simply to satisfy the McDonnell Douglas standard. AT T contends, rather, that whenever the person who is selected for the position is a member of the same protected class, the plaintiff is precluded from establishing a prima facie case. We reject the reasoning both of the district court and of AT T.
AT T relies on numerous inapplicable cases. Some of the cases demonstrate merely that a prima facie case can be shown if the position is filled by a nonmember of plaintiff's class. See, e.g., Laborde v. Regents of the University of California, 686 F.2d 715, 717 (9th Cir. 1982); Lynn, 656 F.2d at 1340-41. In others the plaintiffs either failed to introduce any evidence indicating that they applied for openings that actually existed and for which they were qualified, see, e.g. Gay v. Waiters' and Dairy Lunchmen's Union, 694 F.2d 531, 546 (9th Cir. 1982); Burnett v. Chickasaw Area Development Commission, 662 F.2d 1210, 1213 (9th Cir. 1981), Nulf v. International Paper Co., 656 F.2d 553, 558 (10th Cir. 1981), or could not prove that the employers' actions were discriminatory, see, e.g., Coleman v. Braniff Airways, Inc., 664 F.2d 1282, 1285 (5th Cir. 1982). With one possible exception, see infra note 6, these cases do not support AT T's contention that a plaintiff who is a member of the same protected class as the individual who received the job cannot, as a matter of law, establish a prima facie case.
In order to preclude a plaintiff from recovery in all cases in which the challenged position is filled by a member of the same protected class, we would have to assume that Title VII protects only group rights. It is, of course, true that Title VII was designed to deter and remedy discrimination on the basis of group characteristics and to remove barriers that favor certain groups over others. See Griggs v. Duke Power Co., 401 U.S. 424, 429-30, 91 S.Ct. 849, 852-53, 28 L.Ed.2d 158 (1971); see also United Steelworkers v. Weber, 443 U.S. 193, 207-09, 99 S.Ct. 2721, 2729-30, 61 L.Ed.2d 480 (1979) (private employer's race-conscious affirmative action plan, designed to eliminate "conspicuous racial imbalance" and traditional patterns of segregation, not barred by Title VII); Johnson v. Transportation Agency, 748 F.2d 1308, 1314 (9th Cir. 1984) (public agency's affirmative action plan lawful and necessary to remedy long-standing imbalances in work force); cf. Fullilove v. Klutznick, 448 U.S. 448, 482, 100 S.Ct. 2758, 2776, 65 L.Ed.2d 902 (1980) (opinion of Burger, C.J.) (Congress' remedial actions need not be "colorblind"). But, at the same time, Title VII's focus on the rights of individual members of protected classes is "unambiguous." Arizona Governing Committee for Tax Deferred Annuity and Deferred Compensation Plans v. Norris, 463 U.S. 1073, 103 S.Ct. 3492, 3496, 77 L.Ed.2d 1236 (1983) (quoting Los Angeles Department of Water Power v. Manhart, 435 U.S. 702, 708, 98 S.Ct. 1370, 1375, 55 L.Ed.2d 657 (1978)). "It is clear that Congress never intended to give an employer license to discriminate against some employees on the basis of race or sex merely because he favorably treats other members of the employees' group." Connecticut v. Teal, 457 U.S. 440, 455, 102 S.Ct. 2525, 2535, 73 L.Ed.2d 130 (1982); see also Peters v. Lieuallen, 693 F.2d 966, 970 (9th Cir. 1982) ("The fact that a particular screening device admits some members of a protected class into pool of job candidates does not demonstrate an absence of discrimination."). Title VII protects individuals, as well as groups, from discrimination on the basis of group characteristics.
The Supreme Court's recent decision in Firefighters Local Union No. 1784 v. Stotts, ___ U.S. ___, 104 S.Ct. 2576, 81 L.Ed.2d 483 (1984), does not undermine the group-rights goals of Title VII. That decision simply holds that a court may not modify a pre-existing race-conscious consent decree in such a fashion that the resulting employment practices would divest another group of employees of entitlements under a bona fide seniority system.
We have already made it clear in another context that a plaintiff is not precluded from bringing suit merely because a person of the same protected class is selected for the challenged position. In Douglas v. Anderson, 656 F.2d 528 (9th Cir. 1981), we held that a plaintiff suing under the Age Discrimination in Employment Act (ADEA), 29 U.S.C. § 621-634 (1982), may establish a prima facie case under McDonnell Douglas even though the position was filled by a person who was also within the class protected by the ADEA. See 656 F.2d at 532 (requirement that person selected to fill challenged position not be a member of plaintiff's protected class "adds a gloss that is absent from McDonnell Douglas"). Other circuits have interpreted the fourth element of McDonnell Douglas in the same manner, both in ADEA cases, see, e.g., Loeb v. Textron, Inc., 600 F.2d 1003, 1013 n. 9 (1st Cir. 1979), and in Title VII cases, see, e.g., Jones v. Western Geophysical Co. of America, 669 F.2d 280, 284 (5th Cir. 1982) (Black plaintiff's replacement by another Black may have been a pretextual device and did not preclude finding "an inference of discrimination sufficient to satisfy the fourth element of McDonnell Douglas"). We see no reason to read McDonnell Douglas differently in the Title VII context than we do when suit is brought under the ADEA.
We need not now decide whether identity of class membership may ever be relevant to a determination whether a plaintiff has established a prima facie case under the McDonnell Douglas test. Rather, we need say only that, contrary to the district court's assumption, such identity does not automatically preclude a plaintiff from meeting the test. When the individual who was promoted receives the challenged position only after the plaintiff has filed a discrimination charge, the fact that both individuals are members of the same protected class does not rebut the otherwise established inference of discrimination. In the case before us, there is no bar to plaintiff's establishing a prima facie case under McDonnell Douglas.
Contrary to AT T's suggestion, we think it clear that a plaintiff who satisfies the four elements of McDonnell Douglas has necessarily established the requisite inference of discrimination. Equally important, however, it is possible for a plaintiff to establish a prima facie case without satisfying the McDonnell Douglas test. Compliance with that standard is only one method of establishing a prima facie case. International Brotherhood of Teamsters v. United States, 431 U.S. 324, 358, 97 S.Ct. 1843, 1866, 52 L.Ed.2d 396 (1977); Gay v. Waiters' Dairy Lunchmen's Union, 694 F.2d 531, 550 (9th Cir. 1982); Lynn v. Regents of the University of California, 656 F.2d 1337, 1341 (9th Cir. 1981). In McDonnell Douglas itself the Supreme Court noted that, because of the variations in factual situations, plaintiffs in some cases may seek to establish a prima facie case by other means. See 411 U.S. at 802 n. 13, 93 S.Ct. at 1824 n. 13; see also Note, Sexual Harassment Claims of Abusive Work Environment Under Title VII, 97 Harv.L.Rev. 1449, 1454-59 (1984) (inappropriateness of McDonnell Douglas test in sexual harassment cases). A plaintiff may demonstrate an inference of discrimination in whatever manner is appropriate in the particular circumstances. See Loeb v. Textron, Inc., 600 F.2d 1003, 1014 n. 12 (1st Cir. 1979).
Even when a panel of the District of Columbia Circuit phrased the fourth element of the McDonnell Douglas test as a requirement that "employees . . . who were not members of the protected group were . . . promoted," Freeman v. Lewis, 675 F.2d 398, 400 (D.C.Cir. 1982), it noted that a plaintiff might be able to establish a prima facie case without satisfying the McDonnell Douglas test by presenting statistical or testimonial evidence of discrimination. See id. at n. 14.
The burden of establishing a prima facie case is not designed to be "onerous." Burdine, 450 U.S. at 253, 101 S.Ct. at 1094. In order to establish a prima facie case of discrimination, the plaintiff need only provide evidence that suggests that the "employment decision was based on a discriminatory criterion illegal under the [Civil Rights] Act." Teamsters, 431 U.S. at 358, 97 S.Ct. at 1866. Diaz presented substantial evidence to support the inference that AT T acted in a discriminatory fashion. Although he has frequently filled in temporarily as an Operations Supervisor, Diaz has never been offered this position on a permanent basis. He is as qualified as other candidates who were considered for the job, see supra note 3, and may even be able to demonstrate that he was more qualified than the individual who was promoted, see supra note 2. Furthermore, Diaz has introduced evidence of AT T's complete failure in the past to promote Mexican-Americans to the Position of Operations Supervisor in the Tucson facility. See supra note 1. The fact that AT T eventually promoted a Mexican-American to the position that Diaz sought, after Diaz filed his state discrimination claim, does not automatically remove the presumption of discrimination created by this other evidence. Cf. EEOC v. Brown Root, Inc., 688 F.2d 338, 340 (5th Cir. 1982) (female plaintiff who was replaced by another woman could establish an inference of discrimination without meeting traditional elements of McDonnell Douglas test).
The district court erred in holding that because the person promoted was a member of the same protected class as Diaz, Diaz was barred from establishing a prima facie case.
III. DISCOVERY OF THE EMPLOYMENT STATISTICS
When we review a district court's order granting summary judgment, we normally examine the record to ascertain whether it provides any basis for affirmance. See Jewel Companies v. Payless Drug Stores Northwest, Inc., 741 F.2d 1555, 1564-65 (9th Cir. 1984). We affirm such orders when it is clear that, even if the evidence is viewed in the light most favorable to the non-moving party, no genuine issue of material fact exists and the moving party is entitled to prevail as a matter of law. See Twentieth Century-Fox Film Corp. v. MCA, Inc., 715 F.2d 1327, 1328 (9th Cir. 1983); Fed.R.Civ.P. 56(c). We can make these determinations because, ordinarily, summary judgment orders are not entered until after the non-moving party has had an opportunity to obtain through discovery whatever evidence is necessary or relevant (at least for the purposes of summary judgment) and to place that evidence in the record.
Diaz sought to use statistical data to support both his prima facie case and his argument that AT T's articulated reason for hiring Gonzales instead of him was pretextual. Diaz filed a discovery motion, seeking to compel AT T to provide him with information about its hiring and promotion patterns in its Western Region (covering the Western portion of the United States). AT T claimed that the employment statistics for the portion of its Western Region outside of Tucson were irrelevant and refused to supply them. Because the district judge believed he could dispose of this case on the ground discussed in the preceding section of this opinion, he granted AT T's summary judgment motion without ruling on the discovery motion.
When we review a case in which summary judgment was ordered but the non-moving party was improperly denied discovery, we are not in a position to view all the necessary evidence in the light most favorable to that party. In such a case, there is no purpose to reviewing the record to determine whether the order could be affirmed on a basis other than that relied upon by the district court, unless, of course, there is a separate and independent ground to which the material sought in discovery would be irrelevant. Here, the statistical data, if discoverable at all, would be relevant to the only other ground on which the district court's order might be affirmed. Accordingly, we must next determine whether Diaz was entitled to the requested statistical information.
AT T has advanced only two grounds for summary judgment. First, as discussed in the preceding section of this opinion, AT T contends that Diaz has not established a prima facie case. Second, according to AT T, Diaz has failed to produce evidence sufficient to convince a factfinder that AT T's articulated reason for hiring Gonzales instead of him was pretextual. But the requested statistical information is probative on both of these points. See infra note 8 and accompanying text.
Statistical evidence is unquestionably relevant in a Title VII disparate treatment case. Statistical information is helpful in establishing a prima facie case "despite the fact that [it] may not be directly probative of any of the four specific elements set forth by McDonnell Douglas." Lynn, 656 F.2d at 1342-43. In fact, summary judgment is patently inappropriate when a plaintiff needs statistical data to substantiate the inference of discrimination and has been denied the opportunity to discover this data. See Cedillo v. International Association of Bridge Structural Iron Workers, 603 F.2d 7, 12 (7th Cir. 1979). A plaintiff is also entitled to use statistical evidence to show that a defendant's articulated nondiscriminatory reason for the employment decision in question is pretextual. See McDonnell Douglas, 411 U.S. at 804-05, 93 S.Ct. at 1825-26.
In fact, the same statistical evidence that a plaintiff introduces to establish a prima facie case may be considered by the trier of fact on the issue of whether the defendant's explanation for the employment decision is pretextual. See Burdine, 450 U.S. at 255 n. 10, 101 S.Ct. at 1094 n. 10; Lynn, 656 F.2d at 1346 n. 13.
Statistical data is relevant because it can be used to establish a general discriminatory pattern in an employer's hiring or promotion practices. Such a discriminatory pattern is probative of motive and can therefore create an inference of discriminatory intent with respect to the individual employment decision at issue. In some cases, statistical evidence alone may be sufficient to establish a prima facie case. See O'Brien v. Sky Chefs, Inc., 670 F.2d 864, 866 (9th Cir. 1982) ("[s]tatistical data is one way to establish a prima facie case"); cf. Gay v. Waiters' Dairy Lunchmen's Union, 694 F.2d 531, 552 (9th Cir. 1982) (statistical evidence alone is only rarely sufficient to establish a prima facie disparate treatment case); Laborde v. Regents of University of California, 686 F.2d 715, 718 (9th Cir. 1982) (statistical evidence not necessarily sufficient to establish prima facie disparate treatment case), cert. denied, 459 U.S. 1173, 103 S.Ct. 820, 74 L.Ed.2d 1017 (1983). Even when not sufficient to establish a prima facie case, statistical evidence is helpful in showing that an employer's articulated reason for the employment decision is pretextual. In some cases it may be essential:
Statistics showing racial or ethnic imbalance are probative . . . because such imbalance is often a telltale sign of purposeful discrimination; absent explanation, it is ordinarily to be expected that nondiscriminatory hiring practices will in time result in a work force more or less representative of the racial and ethnic composition of the population in the community from which employees are hired.
. . . "In many cases the only available avenue of proof is the use of racial statistics to uncover clandestine and covert discrimination by the employer or union involved."
International Brotherhood of Teamsters v. United States, 431 U.S. 324, 340 n. 20, 97 S.Ct. 1843, 1857 n. 20, 52 L.Ed.2d 396 (1977) (quoting United States v. Ironworkers Local 86, 443 F.2d 544, 551 (9th Cir.), cert. denied, 404 U.S. 984, 92 S.Ct. 447, 30 L.Ed.2d 367 (1971)).
AT T argues that even if statistical data is generally relevant in Title VII cases, information about its hiring and promotion patterns in the portion of the Western Region outside of Tucson is irrelevant. According to AT T, the decision about Diaz cannot be considered in the context of this larger geographic labor market because (1) promotions outside of Tucson were made by decision makers other than the Operation Manager who made the challenged promotion decision, and (2) Diaz had told his supervisors he was unwilling to leave Tucson to accept a promotion.
Both of AT T's arguments ignore the underlying purpose of statistical information — to provide otherwise unavailable indications of an employer's conscious or unconscious motives. Diaz contends that his employer, AT T, engages in a region-wide policy of discrimination. The existence of a pattern of racial disparity in AT T's employment decisions would allow for an inference about its motives. This would bolster Diaz's prima facie case and would support his contention that the articulated reason for AT T's failure to promote Diaz is pretextual. Thus, Diaz is entitled to attempt to prove that such a pattern exists.
It is irrelevant whether the same Operations Manager made all of the hiring and promotion decisions about which Diaz seeks information. One way of reaching conclusions about an employer's motive is by ascertaining whether the employer's explicit or implicit policies encourage or permit discriminatory employment decisions by its supervisory personnel. The employer is responsible for such decisions, see Miller v. Bank of America, 600 F.2d 211, 213 (9th Cir. 1979), because its policies control the manner in which its supervisory employees make them. We have previously recognized that the promotion patterns of the employer, as a whole, are relevant to such an analysis of motive. See Lynn, 656 F.2d at 1342 (defendant's university-wide tenure-granting pattern, not just tenure patterns in individual departments, relevant in disparate treatment case).
Similarly, it is irrelevant whether Diaz would have been willing to leave the Tucson area if he had been offered a promotion. Even if Diaz could not use the Western Region statistics as evidence that he was discriminated against with regard to other potential promotions, the evidence of a pattern of racial disparity would nevertheless be of probative value with regard to his employer's motive. In Lynn, we evaluated statistical evidence about the defendant university's tenure-granting pattern from the time of its founding until the time of the complaint; we did not restrict ourselves to a consideration of tenure practices during the period in which the plaintiff was eligible for tenure. See 656 F.2d at 1342.
Diaz seeks the Western Region statistics for two purposes. First, he intends to use them to demonstrate a discriminatory pattern in AT T's employment practices. Second, Diaz may have been qualified for other Operations Supervisor openings, approximately four of which were available weekly in AT T's Western Region. Because AT T does not inform nonsupervisory personnel about these openings, Diaz cannot specifically identify openings for which he was qualified but was not considered. If he was qualified for other openings that were available within 300 days prior to the time he filed his charge with the EEOC on June 14, 1980, see 42 U.S.C. § 2000e-5(e) (1982), the failure to promote Diaz in those instances could also form a basis for this suit. Diaz could not, however, maintain that he was qualified for those openings unless he were willing to leave the Tucson area.
Furthermore, even if we assumed that the statistical information would be relevant only if it pertained to jobs in areas in which Diaz was willing to work, the Western Region statistics would not, on the basis of the record before us, be irrelevant. AT T argues that Diaz refused to leave Tucson to accept a promotion. However, the only evidence before us that in any way supports this contention is the evidence that, in the past, Diaz had refused to leave Tucson to accept transfers to positions at the same level as his current position. Diaz contends, however, that he would have been willing to leave Tucson to accept a promotion; nothing in the record before us serves to refute his argument.
AT T refused to provide Diaz with relevant statistical information which might support both the prima facie inference of discrimination and Diaz's allegation that AT T's articulated reason for failing to promote him was pretextual. Thus, we do not have before us all of the material that would be necessary before we could determine whether the summary judgment order could be affirmed on any of the potential grounds presented by the record.
IV. CONCLUSION
The district court relied on an erroneous legal premise when it concluded that Diaz was precluded from establishing a prima facie case of discrimination. For that reason, and because Diaz was improperly denied the opportunity to discover material that is relevant to the only bases on which summary judgment could have been granted, we reverse.
REVERSED AND REMANDED.