Summary
In Wilmore v. City of Wilmington, 699 F.2d 667 (3d Cir. 1983), "the exclusion of minority firefighters from `administrative jobs' because of their race with its consequent detrimental effect on their promotional test scores" was held to be "the kind of `artificial, arbitrary and unnecessary' barrier to professional development the [Supreme] Court found in [Connecticut v.] Teal [ 457 U.S. 440, 102 S.Ct. 2525, 2533, 73 L.Ed.2d 130 (1982)] to be prohibited by Title VII."
Summary of this case from Holtz v. Marcus Theatres Corp.Opinion
No. 82-1203.
Argued September 30, 1982.
Decided February 9, 1983. Rehearing and Rehearing In Banc Denied March 8, 1983.
Gary W. Aber (argued), Biggs Battaglia, Wilmington, Del., for appellants.
Robert D. Goldberg, City Sol., Frederick H. Schranck (argued), Asst. City Sol., City of Wilmington Law Dept., Wilmington, Del., for appellee.
Appeal from the United States District Court for the District of Delaware.
OPINION OF THE COURT
This case presents accusations of racial discrimination against the Wilmington Delaware Fire Department ("Fire Department"). Plaintiffs brought this class action in the United States District Court for the District of Delaware claiming that the Fire Department's promotion policies discriminate against racial minorities in violation of 42 U.S.C. §§ 1981, 1983 and Title VII of the 1964 Civil Rights Act, 42 U.S.C. § 2000e et seq. The district court rendered judgment in favor of defendants. We, however, find that the district court erred and will reverse as follows.
The Wilmington Firefighters Local 1590, International Association of Firefighters is a third-party defendant.
We hold that the district court erred in finding that the "administrative jobs" that were assigned on the basis of race did not create the kind of " barriers to professional development" that violated petitioners' right to equal employment opportunities under Title VII. Connecticut v. Teal, ___ U.S. ___, ___, 102 S.Ct. 2525, 2532, 73 L.Ed.2d 130 (1982) (emphasis in original).
We hold that the racially discriminatory assignment of "administrative jobs" affected the results of the 1978 and 1980 promotional tests in favor of white firefighters and to the detriment of minority firefighters. Therefore, we find that these tests significantly contributed to the disparate impact of the 1978 and 1980 promotion procedures in violation of Title VII of the Civil Rights Act of 1964 and 42 U.S.C. §§ 1981 and 1983.
I.
Wilmington's history is tainted with the stain of racial discrimination and court actions brought by racial minorities to eradicate discrimination. See, Wilmore v. City of Wilmington, 533 F. Supp. 844, 846, n. 4 (D.Del. 1982). This past litigation notwithstanding, racial discrimination has persisted. Even Wilmington's adoption of a policy of racial equality and the creation of an Affirmative Action Committee have failed to break the tenacious grasp of discrimination within city agencies such as the Fire Department. It is within this history of racial discrimination in Wilmington and the city government's attempts to bring about a more just society that the facts of this case must be understood.
Although the City of Wilmington established its Fire Department in 1921, minorities were not included in its ranks until 1961. Id. at 846. Despite recent efforts to recruit more blacks and hispanics, minorities continue to be under-represented in the Fire Department. Minorities make up 48% of Wilmington's labor force but only 15% of its firefighters force. 533 F. Supp. at 846; Joint Appendix ("J.A.") at 778. Minorities are even more seriously under-represented in the Fire Department's command structure. Of the fifty-one officers, only two are minorities. The elevation of these two into the officer ranks did not occur until 1980; and one of the two minority officers was promoted only after this complaint was filed.
The officer corps of the Fire Department includes in ascending order: Lieutenant, Captain, Battalion Chief, Deputy Chief and Chief of Fire. A firefighter must serve a minimum of three years in each rank before becoming eligible for promotion to the next rank. In addition, Wilmington requires firefighters seeking promotion to take promotion tests.
The Fire Department's promotion procedures have undergone various changes over the last decade, in part, because of earlier legal challenges. In 1973 and 1974, promotion procedures permitted an individual's superior officers and the Commissioner of Public Safety to determine who would be promoted. These procedures included subjective evaluations by superiors that carried sufficient weight to offset the results of written promotional examinations. 533 F. Supp. at 847. Litigation challenging the 1973 promotion procedures resulted in a judgment for petitioners. Malloy v. City of Wilmington, C.A. 78-374 (D.Del. filed Dec. 16, 1981). The firefighters' union subsequently challenged the disproportionate weight given to these subjective evaluations and the results of the 1974 Test. In a judicially-approved settlement, the City agreed to eliminate the subjective evaluations made by the Chief of Fire and Commissioner of Public Safety from the promotion process. 533 F. Supp. at 848.
The promotion procedures were modified again in 1978 in an attempt to restore credibility to a process that firefighters had come to regard as unfair and controlled by cronyism. Id. at 847. The promotion process adopted by the Fire Department excluded all subjective ratings by superiors. The components of this three-part process included a written examination (worth 50 points), an oral examination (worth 40 points) and experience (worth 8 to 10 points depending upon length of service). Id. at 848. Combining the points of this process yielded a list of individuals who qualified for promotion in a strict rank-ordering with each vacancy to be filled by the next available highest scorer.
In 1977, the City hired the William L. Clark Company to develop new promotion procedures.
Sixty-five white and eight minority candidates completed the 1978 exam. Only one minority candidate finished within the top twenty, and he ranked eighteenth. Because only eleven candidates were promoted from this promotion list, none was a minority. Id. at 848-49.
Wilmington's Affirmative Action Officer, Gale Templin, considered the absence of any minority promotions an undesirable result of the 1978 exam. Id. at 849. Moreover, the district court found that Chief of Fire, Jerome Donahue, is committed to racial equality and to fair and unbiased promotion procedures. Id. at 859. Templin consequently hired Dr. Barry Morstain and his consulting firm to develop fair procedures for the 1980 promotion exam. However, Dr. Morstain was retained in January 1980, too late to revise completely the 1978 Test or to conduct a complete validation study of the 1980 exam. Id. at 849.
Pursuant to Dr. Morstain's recommendations, the City adopted a two-step, four-part process that displaced the three-part 1978 process. First, a candidate was required to pass a written test of one hundred questions in order to be eligible for the rest of the exam. The second step included three parts. The point value for the oral portion of the exam was raised to 60% of the overall rating. The experience factor counted for 10%. A separate Administrative Task Analysis (ATA) portion of the exam accounted for the remaining 30% of the final score.
Dr. Morstain placed this weight on the ATA because of a pre-1980 Lieutenant's Job Analysis which concluded that lieutenants devoted 30% of their time handling administrative matters. Id. at 849-50. Dr. Morstain also assumed that all firefighters received administrative experience and were familiar with the forms included in the ATA. Id. This assumption proved to be erroneous and detrimental to the test performance of minority firefighters. Sixty-one white, one hispanic and ten black firefighters completed the 1980 exam. Although four of the top twenty finishers were minority, only one minority firefighter, a black, finished high enough to be promoted. He ranked seventh. Eight whites were promoted.
Petitioners claim that the 1978 and 1980 promotion procedures are racially discriminatory in that they had a disparate impact upon minorities in two primary ways. First, the percentage of minorities in the Fire Department's officers corps is significantly smaller than the proportion of minorities in the Wilmington labor force. Id. at 850. Second, no minorities and eleven whites were promoted in 1978 and one minority and eight whites were promoted in 1980. Id. at 851-52. Petitioners also claim that the assignment of firefighters to administrative positions within the Fire Department was made, in part, on the basis of race, and that experience acquired in these positions contributed significantly to the assignees' performance on the promotional tests. Id. at 851. They claim, therefore, that the racially discriminatory treatment of minorities gave white firefighters an unfair advantage over minority firefighters and accounts, in part, for the disparate impact of the promotional tests upon promotions in 1978 and 1980. Petitioners claim that the promotion processes used by the Fire Department in 1978 and in 1980 violated their civil rights under Title VII of the 1964 Civil Rights Act and under 42 U.S.C. §§ 1981 and 1983.
II.
Liability under Title VII of the 1964 Civil Rights Act and 42 U.S.C. §§ 1981 and 1983 can be shown under two separate legal theories. The first and older theory under Sections 1981 and 1983 and Title VII refers to the discriminatory treatment of minorities. This theory requires proof of a discriminatory motive. General Building of Contractors Association, Inc. v. Pennsylvania, et al., ___ U.S. ___, 102 S.Ct. 3141, 73 L.Ed.2d 835 (1982); Texas Dept. of Comm. Affairs v. Burdine, 450 U.S. 248, 259, 101 S.Ct. 1089, 1096, 67 L.Ed.2d 207 (1981); Croker v. Boeing Co. (Vertol Div.), 662 F.2d 975, 988-89 (3d Cir. 1981) (en banc). The second and more recent theory under Title VII alone is the "disparate impact" theory that originated in Griggs v. Duke Power Co., 401 U.S. 424, 431, 91 S.Ct. 849, 853, 28 L.Ed.2d 158 (1970). Thus, the two theories differ in that the former is directed to the discriminatory motive underlying the "treatment" of minorities whereas the latter focuses upon the "outcome" regardless of how benign the motive underlying the treatment of minorities might have been.
The interweaving of these two, separate legal theories of liability under the applicable civil rights statutes complicates this case. Plaintiffs claim that, while the 1978 and 1980 promotional tests are neutral on their face, they are part of a promotional process that is tainted by discriminatory treatment of racial minorities with respect to administrative experience. The end result is that these otherwise facially neutral tests produced a disparate impact. They allege that administrative experience assisted promotional test-takers, and that this experience was provided to white, but withheld from minority, firefighters, in part, because of race.
III.
This court has applied the disparate impact theory to "`practices that are facially neutral in their treatment of different groups but in fact fall more harshly on one group than another and cannot be justified by business necessity.'" 662 F.2d at 991. Croker specified that if "a plaintiff makes out a prima facie case of disproportionate impact . . . a defendant must offer evidence to show that the challenged requirement or device has a manifest relation to employment." Id.
A prima facie case of disparate impact can be shown by statistical indicia. Id. The court below used the 5% standard to determine whether promotion rates in 1978 and 1980 evinced disparate impact. The court expressed this rule as follows:
The five percent rule focuses directly on statistical significance . . . . If the probability that the observed result of the selection process could occur in a racially neutral manner is less than five percent, then the Court will conclude that there is a disparate impact as a result of race.
The district court rejected plaintiffs' statistical evidence of disparate impact based upon their work force/officer corps comparison. The court reasoned that the population/work force data did not bear a sufficiently direct relationship to the applicant pool and the positions for promotion. Instead, it concluded that, because of the special qualifications required to fill particular officers' jobs, the relevant work force must be limited to the firefighter ranks just as it was limited to school teachers in Hazelwood School District v. U.S., 433 U.S. 299, 307, 97 S.Ct. 2736, 2741, 53 L.Ed.2d 768 (1977). However, the Court did not decide whether, in a relevant work force of 224 firefighters of whom 28 are minority, the statistical probability of 2 minorities out of 51 officers constituted conclusive evidence of disparate impact in the promotional procedures. Nonetheless, Dr. Norfleet W. Rives, petitioners' statistical expert, determined this probability at 2.9%, well within the 5% standard for showing the presence of disparate impact. J.A. at 780.
Dr. Rives is Associate Professor of Mathematics, Urban Affairs and Public Policy at the University of Delaware.
Petitioners agree that the statistical evidence relating to rates of promotion among white and minority firefighters is inconclusive. Dr. Rives' rationale for the unreliability of this statistical evidence standing alone is persuasive. He asserted that the statistical universe is simply too small numerically to provide valid conclusions. Id. at 786. Thus, he demonstrated that the results would fall outside the 5% threshold even if no minorities were promoted in 1978 and 1980. Id. However, this case presents other evidence, statistical and testimonial, that demonstrates the presence of disparate impact and discriminatory treatment in the promotion procedures employed by the Fire Department.
The court did reject, as numerically insignificant because the universe was too small, data showing the presence of disparate impact in the promotion rate under the EEOC 80% guideline. Id. at 853.
The court did find that assignment to "some `administrative' positions" was tainted by racial discrimination. Id. at 855. However, it rejected petitioners' contention that the experience acquired in these jobs contributed to higher test scores achieved by white firefighters on promotional tests. The court concluded that this discriminatory treatment was irrelevant as a matter of law to petitioners' claim of disparate impact. It therefore entered judgment for appellees.
IV.
Last June, the Supreme Court decided a case with circumstances similar to this case. Connecticut v. Teal, supra. The Court's discussion of the focus of the disparate impact theory and of Congress' objectives in enacting Title VII are germane to this case. Quoting Griggs, the Court declared:
In Teal, a black employee of a Connecticut state agency challenged, as an absolute requirement for promotion, that applicants pass a written test that disproportionately excluded blacks and that was not job-related. Before trial, the state promoted 22.9% of the black candidates and only 13.5% of the white candidates as a result of the state's affirmative action program. The Court held that the states' nondiscriminatory "bottom-line" promotions did not preclude petitioners from establishing a prima facie case of disparate impact in violation of Title VII, nor did it provide the state with a defense to such an action.
A disparate impact claim reflects the language of § 703(a)(2) and Congress' basic objectives in enacting that statute: `to achieve equality of employment opportunities and remove barriers that have operated in the past to favor an identifiable group of white employees over other employees.'. . . In other words, § 703(a)(2) prohibits discriminatory `artificial, arbitrary, and unnecessary barriers to employment,' . . . that `limit . . . or classify . . . applicants for employment . . . in any way which would deprive or tend to deprive any individual of employment opportunities.'
Connecticut v. Teal, 102 S.Ct. at 2531-32 (emphasis in original). The Court again emphasized Title VII's focus on equal employment opportunities:
Title VII guarantees these individual respondents the opportunity to compete equally with white workers on the basis of job-related criteria. Title VII strives to achieve equality of opportunity by rooting out `artificial, arbitrary and unnecessary' employer-created barriers to professional development that have a discriminatory impact upon individuals.
Id. 102 S.Ct. at 2533. (emphasis in original).
The artificial barrier to which the Court referred in Teal was a promotional test that was not job-related. The case before us involves two promotional tests. Neither test was validated nor was considered to be as job-related as the City would have liked. J.A. at 883, 897-98. However, the barrier to equal employment opportunities here is not so much the isolated tests as it is the overall process firefighters undergo to achieve promotion. The barrier encompasses the interrelationship among the racially biased method of assigning firefighters to administrative jobs, the advantage to promotional test-takers provided by the experience acquired on these jobs, jobholders' superior performance on the promotional exams and their disproportionately greater rates of promotion among white administrative jobholders over minorities.
Like Teal, this case also involves a governmental policy of affirmative action. However, evidence of the government's good faith effort to achieve a nondiscriminatory workforce was stronger in Teal where the government actually promoted minority employees in greater proportion than it did white employees. The government's policy in the instant case failed to yield significantly greater numbers of minority promotions than under previous policies. Moreover, the record in this case shows that the City's and Fire Department's general policy of racially equal employment opportunities was breached by Fire Department officers who excluded minority firefighters from job opportunities that could have contributed significantly to their professional development. We believe that the exclusion of minority firefighters from "administrative jobs" because of their race with its consequent detrimental effect on their promotional test scores is the kind of "`artificial, arbitrary and unnecessary' employer-created barriers to professional development" the Court found in Teal to be prohibited by Title VII. Connecticut v. Teal, 102 S.Ct. at 2533.
Plaintiffs' claim of intentional discrimination is directed at assignments to certain "administrative jobs." The method of selecting firefighters for these administrative positions was informal. Standards and qualifications for these positions were neither published nor known to firefighters. J.A. at 866, 954. Many of these appointments were the product of subjective preference of the officer under whom the firefighter served. Id. at 856-57, 866, 954. Plaintiffs argue that these administrative jobs gave assignees an unfair advantage on the 1978 and 1980 tests. The unfair advantage is the administrative experience these jobs provided.
For example, a Battalion Chief's aide works very closely with the Chief. Indeed, a Battalion Chief's aide is often described as a Chief's "right hand" and "an extension of the Chief himself." Id. at 812, 989. Because they accompany Battalion Chiefs to the scenes of fires, these aides receive a greater and a more sophisticated experience than the typical firefighter normally receives. Id. at 1008(1). This position provides on-the-job training in the Chief's firefighting techniques. Aides learn to anticipate the Chief's commands and to understand firefighting rationales. Id. at 990. In addition, this position requires aides to perform specific administrative jobs that expose them to more administrative procedures than the typical firefighter encounters. Id. at 989-90, 1003-04, 1006-07, 1010. These administrative tasks include preparing Battalion Chief alarm reports, correcting company officers' alarm reports, preparing memos and various kinds of reports, and adjusting the status board. Id. at 1003-04, 1010; 533 F. Supp. at 847.
The Office of Data and Statistics offers selected firefighters a more exclusive administrative experience than that available to the typical firefighter. Individuals assigned here gather, compile and record the information contained in fire alarm reports. Additionally, they maintain the fire alarm journal. J.A. at 876, 1009-10. Finally, they process attendance forms, sick leave forms and daily reports. Id. at 996-97.
Firefighters are assigned also to the Fire Marshal's office where they give weekly presentations to the public. This experience contributes to their public speaking skills and self confidence, Id. at 829, and provides them with an overview and better understanding of the inner workings of the Department and the interrelationship of various areas within the Fire Department. Id. at 1001-02. In addition, assignees to the Fire Marshal's office come into daily contact with Department officers resulting in an ease and familiarity with commanders. Id. at 830-31.
The last position under the informal appointment structure is the Temporary Acting Lieutenant ("TAL"). A TAL is designated to replace the commander of an engine company who is absent because of sickness or vacation. However, these appointments are erratic and of brief duration, usually lasting for only a day or two. Id. at 828(1). Consequently, a TAL does not possess the stature or receive the experiences that are attached to the other positions discussed above. Id. at 828(1), 872-73. TAL was the only administrative position to which minorities were assigned.
One other administrative position offered firefighters valuable experience, but these appointments were the product of a more structured process of selection. Designated Acting Lieutenant ("DAL") was a position created because of the allegedly rigged 1974 exam. The City agreed to appoint DALs to serve as company commanders pending final resolution of the conflict surrounding the 1974 Test. When the final settlement was reached in 1978, the City appointed a new group of DALs based upon scores achieved on portions of the 1974 Test. 533 F. Supp. at 848. DALs acted as full-time Lieutenants. They performed all the firefighting and administrative duties and responsibilities of Lieutenants for extended periods of time. J.A. at 877, 1007.
The record shows, therefore, that serving as Battalion Chief aides, DALs, and in the office of Data and Statistics provided selected firefighters with experience that helped considerably in the 1978 and 1980 promotion examinations. They were better prepared for the administrative portions of the examinations because these positions provided an opportunity to work in the administrative process on which they were tested. Firefighters were required to work with forms they were required to know in the ATA portion of the 1980 exam. Id. at 826, 1005-07. Officers testified that these positions gave firefighters an advantage in the promotional tests. Id. at 831, 877-78. Administrative experience aided selected firefighters in three primary ways. First, these positions provided them with a practical understanding of fire department administration; Second, these positions contributed to their ease and self-confidence in interacting with superiors which was of assistance in the oral portion of the exams; finally, these positions provided firefighters with experiences that helped them in dealing with the kinds of problems and situations with which they were confronted in promotional exams. Id. at 829-31, 999, 1005.
Statistical evidence compiled by Dr. Rives supports the contention that this administrative experience benefitted firefighters in taking the promotional tests. Comparing the test scores of the 1978 and 1980 examinees with administrative experience to those without, Dr. Rives concluded that administrative experience affected scores on the tests. Thus, on the 1978 Promotion Test, firefighters with administrative experience scored 80.16, or 6.67 promotion points better than those without administrative experience who scored 73.59. Id. at 787-88. The probability of such a result if administrative experience made no difference is .1% which is well within the 5% standard. Even when DALs are excluded from test results, those with administrative experience still finished 4.66 points higher 79.27 to 74.61. Again, the probability of this result if administrative experience had no effect on performance is only 2.3%. Id. at 788-89. Thus, Dr. Rives concluded that these statistics clearly indicate that administrative experience was a factor in separating the test points in the 1978 exam. Id. at 789.
The 1980 test was even more sensitive to administrative experience than the 1978 test. Id. at 788-89, 792-93. Thus, firefighters with administrative experience outscored those without such experience 10.9 to 9.93 or a difference of 1.039. Id. at 792. The probability of this result if administrative experience had no effect on the test is only .3%. If DALs are excluded, then the results are 10.89 to 9.99 or a difference of .90. Id. at 793. The probability of such a result if administrative experience had no effect is 1.2%. Id. Thus, Dr. Rives concluded: "There is a clear evidence in this case, based on the data, that administrative experience plays a role in promotion as far as promotion points go." Id.
The district court found "that assignment to some `administrative' positions in the Fire Department has depended, in part, upon race." 533 F. Supp. at 855. It concluded that "conscious and unconscious racial bias . . . has clearly effected [sic] the selection of personnel for these positions in the past." Id. The court noted that all witnesses testified that minority firefighters were qualified to serve in the designated administrative positions, but that only whites were selected to fill these positions. Id. at 856. In light of the racial composition of the firefighter force, the probability that only whites would be selected was only 3.96%. Id. It concluded that "race has played a role in the selection of BCAs [Battalion Chief's Aides], D S [Data and Statistics] staff, and FMO [Fire Marshal's Office] staff. . . ." Id. at n. 30.
Despite its finding "that assignment to some `administrative' positions in the Fire Department has depended in part on race," and despite the overwhelming evidence demonstrating that administrative experience benefitted selected firefighters, the district court found that this administrative experience that gave an advantage to whites alone did not in and of itself account for the difference in scores in the 1978 test between those who had such experience and those who did not. Id. at 856. Rather, it asserted that "study and memorization were obviously the keys to success." Id. Consequently, the court held that the test did not treat "the `experienced' differently from the `inexperienced.'" Id.
As to the 1980 test, the court stated that "[t]he correlation between the `administrative' experience and scores on the ATA portion of the 1980 test appears to be very close." Id. at 857. It accepted plaintiffs' contention that administrative experience "was of some benefit to BCA's, DAL's, and firefighters assigned to D S and FM." Id. It, nevertheless, concluded that this evidence failed to show that the administrative experience accounted for the promotions because minorities did better than whites on the oral portion of the exam by 7.08 to 6.37. Id. "This difference," the court held, "offset any disparate impact resulting from ATA." Id. Even though firefighters with administrative experience from which minorities were excluded did better than firefighters without such experience, minorities did better on average than whites by 10.3 to 10.2, Id. at 854, because of their superior performance on the oral portion of the exam. Id. at 857. Therefore, the court concluded that the administrative experience from which minorities were excluded did not cause disparate impact upon the results of the examination as a whole. Id.
The court adopted an alternative explanation for the apparent racially disproportionate promotion results. It found
that the process for selecting people to serve in these jobs, even if tainted by racial bias, has selected a group of people who, on average, are likely to perform better on a lieutenant's promotional exam than the group of test-takers as a whole, on average.
Id. at 858. The reasons for this conclusion, the court explained, is
that to the extent the administrative job selection process identifies the same abilities selected by the promotional exams, those chosen for administrative jobs can be expected to perform better on promotional tests on average without regard to the value of the experience of service in those jobs.
Id. In short, the court found that firefighters who had administrative experience performed better on the promotional tests and were promoted because they were superior, not because they enjoyed an unfair advantage.
Having concluded that the racially biased selection process by which white firefighters were exclusively given the advantage of administrative experience did not affect the performance on promotional exams, the court found that the promotion procedures did not discriminate against racial minorities. Id. The court reached this legal conclusion because the mean score of minorities was higher than that of whites on the 1980 promotional exam and because it found that administrative experience did not affect test performance. By focusing on the results of this exam, rendering irrelevant to these results the racially discriminatory process of selection to administrative positions, the court concluded that 1978 and 1980 promotional procedures did not have a disparate impact on minorities. Id. This conclusion of law and its reasoning must be rejected.
The district court's decision rests on a fatal error in reasoning and a clearly erroneous factual finding. That the 1978 and 1980 tests were not validated and were not proven to be job-related precludes the court's finding that the abilities tested on the exams and those used in the administrative job selection process are identical. Moreover, the record contains evidence, and the court found, Id. at 856, that minority firefighters were qualified for the administrative jobs that were reserved to white firefighters. Id.; J.A. at 811, 993, 1008, 1017-18. It also contains evidence that individuals were selected for those jobs because of the personal preferences of the superior under whom they served, not for any particular administrative abilities. In light of this record, there is no reason to assume that the qualities of those selected were so superior as to explain their better test performance. The evidence does not support the court's finding that firefighters who served in administrative jobs possessed superior administrative abilities. Moreover, plaintiffs' statistical evidence at J.A. 793, 795-97 establishes the opposite conclusion — that administrative experience affected the outcome of the promotional tests. Therefore, the district court's finding on this point is clearly erroneous.
Despite this statistical correlation between administrative experience and performance on the promotional exam, the court concluded that administrative experience did not affect the test scores because minorities did better than whites on the oral part of the 1980 exam. 533 F. Supp. at 854. This reasoning penalizes minorities for doing well on one part of the exam and overlooks how much better their overall test results would have been if they had not been so handicapped on that part of the exam that was affected by administrative experience. Therefore, the district court erred in holding that plaintiffs failed to show that the promotional process caused the racially discriminatory result.
We will therefore reverse the decision below and remand these proceedings to the district court with the direction that it grant appellants' petition for declaratory relief, injunctive relief and damages.