Thomas v. City of Evanston

Full title: Brenda THOMAS, individually and on Behalf of a class, Plaintiffs, v. CITY…

Court: United States District Court, N.D. Illinois, E.D

Date published: Feb 26, 1985

610 F. Supp. 422 (N.D. Ill. 1985)

Opinion

No. 80 C 4803.

February 26, 1985.

Kenneth M. Flaxman, Chicago, Ill., for plaintiffs.

William Kurnik, Arlington Heights, Ill., Jack Spiegel, Chicago, Ill., for defendant.

MEMORANDUM OPINION AND ORDER

ASPEN, District Judge:

This is a Title VII class action in which the plaintiffs challenge the City of Evanston's ("the City") use of physical agility tests to screen job applicants for the Police Department ("the Department"). These tests allegedly worsened the female applicants' chances of being hired. Before being elevated to the Seventh Circuit, Judge Joel M. Flaum certified a class consisting of "all women who, from October 2, 1976 until the present, have been adversely affected by the physical agility test. . . ." The parties have now completed discovery and have filed cross-motions for summary judgment, relying mostly on a stipulation of uncontested facts. For the reasons stated below, we grant the plaintiffs' motion for summary judgment and deny the City's.

I.

This case centers around the hiring practices of the City's police department. Specifically, the plaintiffs challenge the Department's use of a physical agility test to cull out applicants for vacancies on the force. The parties have stipulated that the test administered in 1976 and 1979 had a disparate impact on female job applicants.

As of December 31, 1983, the Department employed 112 officers, only four of whom were women. These women perform their jobs satisfactorily.¹ This recent state of affairs represents little change from the time period relevant to this suit, the late 1970's. Employing only two women officers in 1975, the Department created an affirmative action plan which set a goal of hiring twenty women by 1980. However, over the next several years the Department hired very few women. In 1978, the Department had four female officers out of 106; in October 1980, three out of 121 officers were women.

1 This fact, like most of those written below, are taken from a stipulation filed by the parties. Unless otherwise indicated, all of the facts below are uncontested.

Before 1980, the City filled vacancies from a "list of eligibles" which ranked applicants according to their performance on a competitive examination.² The City gave these exams and drew up a list of eligibles in 1976, 1979 and 1982.³ The first phase of the exam was a "physical agility test," administered by the "Bureau of Testing Services" ("BOTS"). Those who failed this test were rejected automatically; those who passed took a written exam. To create the list of eligibles, the City combined the scores of those who had passed both tests and ranked the applicants by their performances on the exam. The total score was the sum of 30% of the physical test grade and 70% of the written test grade. Those with a resulting score below 70 did not make the final list. Finally, when vacancies arose in the Department, applicants on the list would receive a polygraph test, a psychological test, a background check and an interview. Those who passed these tests would be hired according to their rank on the list of eligibles.

2 After 1980, the City began giving preference to applicants with previous police experience. While this practice might have exerted a disparate impact on female applicants (since most experienced officers are men, we assume), it is not challenged in this suit.

3 The 1982 exam is not challenged here.

Both the 1976 and 1979 physical agility tests had an adverse impact on female applicants. In 1976, 293 applicants took the physical test, of whom 85.32% were men and 14.68% were women. 90.40% of the men passed, while 16.28% of the women passed. Thus, 226 out of 250 men were eligible for the rest of the application process, but only 7 out of 43 women so advanced. The City ultimately hired twenty of these men and none of the women. In 1979, 317 applicants took the test — 86.75% men and 13.25% women. The City had lowered the passing score from 70 to 60. 98.95% of the men passed, while 73.8% of the women passed. Thus, only 3 out of 275 men flunked, while 11 out of 42 women failed. The final list was made up of 114 (89.76%) men and 13 (10.24%) women. The City ultimately hired 14 men and one woman from this list.

The physical agility test itself contained several events, such as a stair-climb, 1/4-mile run and an obstacle course. Applicants would receive an integer score ranging from one through five for each event, depending on their performance as compared to a table of "norms." The City and BOTS assert that the test was scaled to create a passing score of 70, and that an applicant who scored below 70 could not perform physically as an officer. The City and BOTS also assert that the norms it used were based upon the performance on the test of a representative sample of incumbent police officers. Plaintiffs dispute both of these contentions. The parties do agree, however, that BOTS has no empirical evidence to support a contention that those who received high passing scores would be better police officers than those who received low passing scores.

Of those men and women who passed the exams, the men received a disproportionate number of the higher scores. For the 1979 test, 28 of the 273 men tested (10.25%) scored 79 or better, while 1 of 44 women tested (2.27%) did so. One woman ranked in the top 81 scores. 80 of 273 males (29.30%) scored better than 74. 7 of 44 women (15.90%) did so.

BOTS established its norms by computing the "mean" and "standard deviation"⁴ of actual scores on the events of the test. For the 1976 and 1979 tests, BOTS tested incumbent police officers to set these norms. An applicant who scored within one standard deviation of the mean set by these incumbents received a "3" for that event; one who scored more than one standard deviation from the mean received a "4." The test was thus scaled so that about 16% of incumbent officers would have received a "2" or less (a failing score for that event) had they taken the test as applicants. BOTS assumed that some 16% of incumbent officers were physically incapable of performing their jobs.

4 The "mean" is simply the arithmetic average of all scores. The "standard deviation" is a number which represents the variability or spread of scores with respect to the mean. For scores which are "normally distributed" on a bell curve, two-thirds fall within one standard deviation of the mean on either side. Thus, a large standard deviation generally tells one that the scores are spread. A small standard deviation means that the scores are clumped on either side of the average score.

BOTS developed the physical agility test following a "job analysis" of the type of work done by police officers in various Illinois cities. As part of the job analysis, BOTS surveyed police chiefs. The survey listed seven questions,⁵ and was completed and returned by three police chiefs. BOTS and the City have no other papers regarding this survey.

5 The survey was 1 1/2 pages long and was brief. It asked the chiefs to review the questions and check the responses "most related to the job" of police officers. These were the questions:

I. Sprint 50 yards ____ Sprint 100 yards ____ Sprint 200 yards ____ Does not apply ____

II. Endurance: Run 1/4 mile ____ Run 1/2 mile ____ Run 1 mile ____ Does not apply ____

III. Throwing tear gas cannister 25 ft. ____ Throwing tear gas cannister 50 ft. ____ Throwing tear gas cannister 100 ft. ____ Other ____ Does not apply ____

IV. Fit comfortably and safely in standardized squad car ____ Must be able to reach pedals and see properly ____ Other (concerned only with ability to safely drive vehicle) ____ Does not apply ____

V. Test of strength to determine physical ability to subdue or handcuff a 175# man ____ Does not apply ____

VI. Please list any other areas you feel are important. Please remember, they must be job related.

VII. What do you think the maximum and minimum height and weight should be? Why?

Space was provided for additional comments.

Another part of the "job analysis" involved observing the work actually performed by police officers. The only documented observations BOTS has consist of two reports written by a graduate student who went on a few "ride-alongs" with the Markham police department. BOTS and the City lack additional documentation of observation of work actually performed by police officers. Moreover, no documentation describes how the survey and observation were used to create the tasks used on the physical agility test.

The class representative, Brenda Thomas, applied with the Department on February 19, 1976. Her application contained a certificate from her doctor that she was physically capable of taking a strenuous physical aptitude test. Thomas took the 1976 test and flunked, thereby being rejected from further consideration. On February 4, 1977, she filed a timely charge with the Equal Employment Opportunity Commission ("EEOC"). The EEOC investigated her charge, and three years later, on February 19, 1980, found "reasonable cause" to believe that her charge was well founded because the test "has an adverse impact on the consideration of females for employment" and "has not been properly validated." The EEOC and the City entered into a conciliation agreement under which the City agreed to revise and rescore the exam. However, Thomas and the City refused to enter into a conciliation agreement, and on July 3, 1980, the EEOC issued a "right to sue" letter. Thomas brought this action pro se, and after retaining counsel, she amended the complaint to add the class claims.

On February 28, 1983, as noted earlier, Judge Flaum certified the class. On September 18, 1984, we approved a consent decree which settled the case in part. Under the decree, the City must hire at least four class members within the next year. Still unresolved however are the issues of liability and appropriate relief. On the basis of stipulated facts, the parties have submitted cross-motions for summary judgment with well-written supporting memoranda. We turn now to these motions.

II.

This is not the first challenge to a police department's use of a physical agility test. Several other courts have ruled that similar tests violated Title VII.⁶ In analyzing the test at issue here, we follow the traditional approach to disparate impact cases set forth in Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971), and its progeny. To establish a prima facie case of discrimination, "a plaintiff need only show that the facially neutral standards in question select applicants for hire in a significantly discriminatory pattern." Dothard v. Rawlinson, 433 U.S. 321, 329, 97 S.Ct. 2720, 2726-27, 53 L.Ed.2d 786 (1977). When a plaintiff makes this threshold showing of disparate impact, the employer bears the burden of showing that its discriminatory policy or practice is necessary and manifestly related to job performance. Id.; Griggs, 401 U.S. at 431-32, 91 S.Ct. at 853-54. If the employer satisfies its burden, the plaintiff may show that less discriminatory alternatives would also "serve the employer's legitimate interest in `efficient and trustworthy workmanship.'" Albemarle Paper Co. v. Moody, 422 U.S. 405, 425, 95 S.Ct. 2362, 2375, 45 L.Ed.2d 280 (1975), quoting, McDonnell Douglas Corp. v. Green, 411 U.S. 792, 801, 93 S.Ct. 1817, 1823, 36 L.Ed.2d 668 (1973); see Harless v. Duck, 619 F.2d 611, 616 n. 6 (6th Cir. 1980), cert. denied, 449 U.S. 872, 101 S.Ct. 212, 66 L.Ed.2d 92 (1980).

6 See Harless v. Duck, 619 F.2d 611, 616-17 (6th Cir. 1980), cert. denied, 449 U.S. 872, 101 S.Ct. 212, 66 L.Ed.2d 92 (1980); Blake v. City of Los Angeles, 595 F.2d 1367, 1381-83 (9th Cir. 1979), cert. denied, 446 U.S. 928, 100 S.Ct. 1865, 64 L.Ed.2d 281 (1980); Burney v. City of Pawtucket, 559 F. Supp. 1089 (D.R.I. 1983), appeal dismissed, 728 F.2d 547 (1st Cir. 1984); Officers for Justice v. Civil Service Commission of San Francisco, 395 F. Supp. 378, 381-86 (N.D.Cal. 1975); generally Annot. 53 A.L.R.Fed. 31 (1981); see also Berkman v. City of New York, 536 F. Supp. 177 (E.D.N.Y. 1982) (firefighter physical agility test violates Title VII), aff'd, 705 F.2d 584, 585 (2d Cir. 1983).

A.

The undisputed facts plainly show that plaintiffs have made out a prima facie case of disparate impact. The 1976 test eliminated about 85% of the women applicants, but only about 10% of the men, from further consideration. The 1979 test, which lowered the passing score, eliminated about one-fourth of the women but only about 1% of the men. Indeed, it can fairly be said that only a woman could have failed the 1979 test. This is clearly a "significantly discriminatory pattern" of selection.

This disparate impact is apparent from more than the relative pass rates. As the parties have agreed, the men received a disproportionate number of higher scores. Because the final list was drawn up in rank order, the men who passed fared better as a group than the women who passed. Only one woman in 1979 ranked in the top 81 scores. About 30% of men tested scored better than 74, while only about 16% of women tested did so.⁷

7 The City argues that the rank-ordering created no disparate impact. We disagree. It points to statistics that the final list of eligibles in 1979 (before applying "veterans points") showed a mean score for men of 73.34 with a standard deviation of 1.89. Women on the list averaged 73.27 with a standard deviation of 1.60. However, as the Seventh Circuit held in a related context, the City's argument

misses the point, for it is the impact of the examination on the racial composition of the first four hundred finishers, not the average impact, that is significant. The ultimate question is whether substantially fewer minorit[ies] . . . were promoted . . . and average scores are irrelevant in answering that question.

United States v. City of Chicago, 549 F.2d 415, 429 (7th Cir. 1977). Similarly, because of rank-ordering, what is important is the impact of the physical test on the top of the list, not on the average. As noted earlier, the City has conceded that men scored a disproportionate number of high scores. Simple arithmetic dictates that they would receive disproportionately better final scores when the physical test was combined with the written score.

It is also plain that the test significantly reduced the ultimate number of women hired. The City hired 34 men from the 1976 lists, and only one woman. Given the disproportionate number of women who failed the physical agility test, as well as the disproportionate number of women who received low passing scores, we find that the physical agility test was in large part responsible for the ultimate disparate impact in hiring.⁸ This is evident both from the face of the above facts and from an application of the so-called "80% rule."⁹

8 The City argues that the disparate impact in the physical agility test is not causally related to the disparate impact in hiring. The undisputed facts point to the opposite conclusion. In 1976, 85% of the women failed the test, while only 10% of the men failed. This could not lead to any result other than a disparate impact in hiring, unless women did exceptionally well in other parts of the process (in which case they still faced worse than they would have without the exam); or unless women would have done exceptionally poorly in other parts of the process (in which case the rest of the process was probably discriminatory). Similarly, given the rank-ordering process and the relatively poor showing of women who passed, the test could not help but significantly diminish their chances of being hired. In sum, the undisputed facts show that but for the physical agility exam, women applicants would not have fared as poorly in the hiring process.
In any event, to prove liability, plaintiffs need not show that the exam led to a proportional disparate impact in hiring. They need only show that the exam had a disparate impact in eliminating them from competition for a job; that is, Title VII "speaks, not in terms of jobs and promotions, but in terms of limitations and classifications that would deprive any individual of employment opportunities." Connecticut v. Teal, 457 U.S. 440, 448, 102 S.Ct. 2525, 2531, 73 L.Ed.2d 130 (1982) (emphasis in original). See also Costa v. Markey, 706 F.2d 1, 4-5 (1st Cir. 1982), cert. denied, ___ U.S. ___, 104 S.Ct. 547, 78 L.Ed.2d 722 (1983). The physical agility test undeniably limited the job opportunities of the women applicants, by flunking a disparate number or by giving them low passing scores.

9 Under EEOC guidelines, a finding of disparate impact may be appropriate where the success rate of the plaintiff class is less than 80% that of the majority group. See 29 C.F.R. § 1607.4D (1984); United States v. City of Chicago, 663 F.2d 1354, 1357 n. 8 (7th Cir. 1981) (en banc); Firefighters Institute v. City of St. Louis, 616 F.2d 350, 356-57 (8th Cir. 1980), cert. denied, 452 U.S. 938, 101 S.Ct. 3079, 69 L.Ed.2d 951 (1981). The relative success rate for the 1976 physical test was 18% (16.28% women passing 90.40% men passing). The relative success rate for the 1979 test, which had a lower passing score, was 74.6% (73.8% women passing/ 98.9% men passing). Of 525 men who applied in 1976 and 1979, 34 were hired, a success rate of 6.48%. Of 85 women who applied in those years, one was hired, a success rate of 1.18%. The relative success rate, then, was 18.2%. All of these figures easily satisfy the 80% rule.

Moreover, the City admitted in its answer, and again in its stipulated facts, that the physical agility test exerted a disparate impact on the female applicants. Such an admission, of course, is binding. The City argues, however, that it has not admitted that plaintiffs have made out a prima facie case, and, moreover, it argues that they have not in fact made out a prima facie case. It asserts first that it only admitted that the test had a disparate impact on the pool of female applicants, not on women generally. The 1976 and 1979 female applicants were not representative of women generally, argues the City. Essentially, the City contends that they were clumsier and in poorer shape than women generally.¹⁰ We reject this contention. We note first that the Supreme Court has rejected a similar argument:

10 The City relies on an EEOC guideline, which provides in part:

Greater differences in selection rate may not constitute adverse impact where the differences are based on small numbers and are not statistically significant, or where special recruiting or other programs cause a pool of minority or female candidates to be atypical of the normal pool of applicants from that group.

29 C.F.R. § 1607.4(D) (1984).

There is no requirement, however, that a statistical showing of disproportionate impact must always be based on analysis of the characteristics of actual applicants. See Griggs v. Duke Power Co., supra, 401 U.S. at 430, 91 S.Ct. at 853. The applicant process might itself not adequately reflect the actual potential applicant pool, since otherwise qualified people might be discouraged from applying because of a self-recognized inability to meet the very standards challenged as being discriminatory. See International Brotherhood of Teamsters v. United States, 431 U.S. 324, 365-67, 97 S.Ct. 1843 [ 52 L.Ed.2d 396].

Dothard v. Rawlinson, 433 U.S. at 330, 97 S.Ct. at 2727. While Dothard applied to height and weight standards, which might more easily encourage self-selection than an agility test, we think that the prospect of competing with men in a rigorous physical agility test might have discouraged otherwise qualified women from applying. See Blake v. City of Los Angeles, 595 F.2d 1367, 1375 (9th Cir. 1979) (applying above Dothard standard to police force which used both a physical test and height requirements), cert. denied, 446 U.S. 928, 100 S.Ct. 1865, 64 L.Ed.2d 281 (1980).

Secondly, the evidence simply does not suggest that the applicant pool was relatively clumsy. No evidence suggests that the City's affirmative action program recruited or attracted klutzes.¹¹ The pattern of disparate impact is not unique to Evanston, but rather appeared in the total of results from statewide administrations of the test. Only about 1/3 of women passed the test statewide between 1977 and 1980, while almost 4/5 of the men passed. Thus, it does not appear that the plaintiff class's poor showing on the test means that the women were sub-par.¹² In sum, then, plaintiffs have proved on the basis of undisputed facts that the physical agility test exerted a disparate impact on them.

11 "Klutz," originally a Yiddish word now in common usage in English, means "a clumsy, awkward person." The Oxford English Dictionary (1976 Supp.) at 525.

12 The City tries to bolster its argument that the applicant pool was sub-par by rescaling the 1976 and 1979 scores under new norms drawn up in 1983. According to defendant's expert, the women's rescaled scores were lower than expected, suggesting an inferior applicant pool. We agree with plaintiffs that this is comparing "apples and oranges." First, the equipment was changed on some of the events. Moreover, the 1983 norms are representative of women in 1983, not in 1976. It might be that the women who set the norms in 1983 were better than average, or that women are in general in better shape now than in the mid- and late 70's. Indeed, the average scores for men improved between 1977 and 1983. For example, men in 1983 ran the quarter-mile 15 seconds faster than in 1977. In sum, we do not think this rescaling of score creates a genuine issue of fact about the applicant pool. The overwhelming evidence supports plaintiffs' prima facie case of disparate impact.

B.

Because plaintiffs have conclusively proved their prima facie case, the City must now bear the burden of proving that the physical agility test is job-relatd and content valid. The City must show that "business necessity" justified using the physical agility test. Griggs, 401 U.S. at 431, 91 S.Ct. at 849; Blake v. City of Los Angeles, 595 F.2d at 1376. In other words, the City must prove that the test bears "a manifest relationship to the employment in question," Griggs, 401 U.S. at 432, 91 S.Ct. at 849, and that the discriminatory test is "necessary to safe and efficient job performance." Dothard, 433 U.S. at 332 n. 14, 97 S.Ct. at 2728; Liberles v. County of Cook, 709 F.2d 1122, 1132 (7th Cir. 1983). Specifically, the City must show that the test is "content valid."¹³ To be "content valid" the test must satisfy several attributes. First, the test-makers must have done a proper "job analysis," that is, a study of important work behaviors required for successful performance and their relative importance. Second, the test must be related to and representative of the content of the job. In other words, the test must measure ability to perform competently on the specific job. Third, the test must be scored so that it properly discriminates between those who can and cannot perform the job well. See Guardian Ass'n of New York City v. Civil Service Comm'n., 630 F.2d 79, 94-95 (2d Cir. 1980), cert. denied, 452 U.S. 940, 101 S.Ct. 3083, 69 L.Ed.2d 954 (1981); Harless v. Duck, 619 F.2d 611, 616 (6th Cir. 1980); United States v. City of Chicago, 573 F.2d 416, 425 (7th Cir. 1978); Berkman v. City of New York, 536 F. Supp. 177, 206-12 (E.D.N.Y. 1982), aff'd, 705 F.2d 584, 585 (2d Cir. 1983). We believe that the undisputed facts show that the City cannot meet its burden of proving content validity.

13 "Content validity" is one of three methods which the EEOC guidelines list as appropriate to validate an exam. See 29 C.F.R. § 1607.5B, 14 (1984); Guardians Ass'n of New York City v. Civil Service Commission, 630 F.2d 79, 91-92 (2d Cir. 1980), cert. denied, 452 U.S. 940, 101 S.Ct. 3083, 69 L.Ed.2d 954 (1981). The parties seem to agree that the "content validity" approach is proper in this case. We agree. See id. at 91-94 (content validity appropriate to exams which measure skills or abilities supposedly representative of job content).

1. The Job Analysis

The job analysis performed by the City was clearly insufficient. The record reveals that the "job analysis" consisted of two stages. First, BOTS, which designed the test, surveyed police chiefs, receiving only three responses. That survey contained just seven questions, see note 5 above, and the three chiefs who returned the survey did not answer all of the questions. Moreover, only four of the seven questions are relevant to the test later drawn up, and the answers to these questions were inconsistent.`¹⁴ This survey seems to have been useless and is hardly sufficient to lay a groundwork for a proper job analysis under Title VII. Compare Officers for Justice, 395 F. Supp. at 384 (finding 150 questionnaires to be insufficient), with Guardians Ass'n, 630 F.2d at 95 (holding job analysis sufficient where analysis consisted of five stages, including interviews with some 100 officers, review by supervisors, a questionnaire returned by 2,600 officers).

14 For example, in response to question two ( see n. 5 above), two chiefs felt that an officer should be able to run 1/2 mile. The third thought that 1/4 mile was enough. Two chiefs thought that an officer should be able to toss a tear gas cannister 50 feet, while the third thought 25 feet was enough. This so-called survey appears worthless, both because of the low number of responses and because of the low number of questions. The City contends that BOTS had more responses, but lost them. That does not help them, since they bear the burden of proving content validity. "Mere testimony that a test has been validated, without a record of validation, is insufficient to prove job relatedness." United States v. City of Chicago, 549 F.2d 415, 432 (7th Cir. 1977). Even if BOTS had 200 responses, this survey would have only slight worth. The seven questions are general and few, falling far short of what is needed to thoroughly assess the content of the job. See Guardians, 630 F.2d at 95.

The second stage of the "job analysis" was also not at all thorough. A series of 13 activities were drawn up following the above "survey." Then a graduate student went on "ride alongs" with eight police officers in the suburban village of Markham. The student rode with officers for about 30 hours, and based upon his observations, suggested some changes in the proposed draft of 13 activities. The record is bare of any other evidence of job analysis, and BOTS has no other documents showing how it constructed the exam.

The record as a whole falls far short of what is needed to document a valid job analysis. To be content valid, an exam must be based on a "thorough task analysis of the job to be performed." United States v. City of Chicago, 573 F.2d at 425. The test must "closely approximate tasks to be performed on the job." Id.; see also Guardians Ass'n., 630 F.2d at 95; 29 C.F.R. § 1607.14(C)(2) (1984).¹⁵ The above two-step analysis does not come close to being a "thorough job analysis."¹⁶ This holding alone means that the City has failed to show, as it must, that the physical agility test was "necessary to safe and efficient job performance." Dothard, 433 U.S. at 332 n. 14, 97 S.Ct. at 2728.

15 29 C.F.R. § 1607.14C(2) (1984) provides:

(2) Job analysis for content validity. There should be a job analysis which includes an analysis of the important work behavior(s) required for successful performance and their relative importance and, if the behavior results in work product(s), an analysis of the work product(s). Any job analysis should focus on the work behavior(s) and the tasks associated with them. If work behavior(s) are not observable, the job analysis should identify and analyze those aspects of the behavior(s) that can be observed and the observed work products. The work behavior(s) selected for measurement should be critical work behavior(s) and/or important work behavior(s) constituting most of the job.

16 While the City asserts that more was done, it lacks documentation to prove so. Since the City carries the burden of showing content validity, its near-total lack of substantiation means that it has not met its burden.

2. The Construction of the Test

In light of the above holding, we cannot analyze whether the content of the test is related to and representative of the content of the job. Without a thorough job analysis, one cannot assess whether a test relates to and fairly represents the job. Thus, we do not address this portion of the content validity analysis.

3. The Scoring System

Even if the City had performed a thorough job analysis and had constructed a test which fairly related to and represented the job of police officer, it could not meet its burden of showing that the test was scored properly. We conclude that the scoring system is deficient in two ways. First, the City concedes that no evidence supports the proposition that an officer who receives a high passing score will perform better on the job than one who receives a low passing score. This concession alone is enough to invalidate the rank-ordering system, which, as we held above, had a disparate impact on the members of the plaintiff class who passed. See Guardians, 630 F.2d at 100 (rank ordering should be used only if defendant can show that a higher score is likely to result in better job performance); Firefighters Institute v. City of St. Louis, 616 F.2d 350, 358 (8th Cir. 1980), cert. denied, 452 U.S. 938, 101 S.Ct. 3079, 69 L.Ed.2d 951 (1980).

Second, the evidence does not, as it must, show that a passing score of 70 or better validly predicts successful job performance. As noted earlier, BOTS established its norms by giving the test to incumbent police officers. It scaled the test so that 16% of incumbent officers would fail. The score on each subtest was scaled so that any person scoring within one standard deviation of the mean would receive a passing score of "3" for that test. An average score on each subtest would yield a final score of 70, the passing cutoff. The City has simply not presented any evidence that this scaling system validly predicts future job performance. First, no evidence supports what appears to be pure speculation that 16% of incumbent police officers would "fail" each subtest. That is, no evidence supports a conclusion that an applicant who scores in the lowest 16% of incumbent officers would perform incompetently. Because the City asserts that a person who flunks the test would be an incompetent police officer, one would expect evidence that at least some of the 16% of incumbents who "flunk" are not in fact performing their jobs satisfactorily. The City has presented no evidence to that effect. Indeed, the City's expert (BOTS employee Dr. Mendelsohn) admits that no empirical evidence supports the assumption that 16% of incumbent officers cannot capably perform their jobs; Dr. Mendelsohn chose that figure because it is "just a standard acceptable, traditional cut-off point in the psychological research." Mendelsohn Deposition at 24. Second, and as a corollary to the first point, no evidence supports a conclusion that a person who fails would be a poorer officer than one who passes.¹⁷

17 Plaintiffs argue that cutoff point is also invalid because the City lowered the passing score, thereby admitting that the scoring was not job-related. We have not considered this argument, because it appears that the City lowered the pass rate as part of conciliation talks with the EEOC. We think this change qualifies as a "subsequent remedial measure," and therefore renders that piece of evidence inadmissible. See Fed.R.Evid. 408.

Under the EEOC guidelines, a cutoff score "should normally be set so as to be reasonable and consistent with normal expectations of acceptable proficiency within the workforce." 29 C.F.R. § 1607.5(H) (1984); see Guardian, 630 F.2d at 105. The cutoff need not be precise. That is, the City need not show that one who scores a 70 would be a good officer and one who scores a 69 would be a bad one. But some evidence must support a principled decision that a cutoff figure really predicts job performance. "When a cutoff score unrelated to job performance produces disparate . . . results, Title VII is violated. Consequently, there should generally be some independent basis for choosing the cutoff." Id. As we have discussed, the cutoff point of 70 was based on a speculative estimate that 16% of incumbent officers cannot physically perform on the job. This unfounded estimate cannot support a valid cutoff line.¹⁸

18 We also think that the City's concession that the rank-ordering system is invalid suggests a conclusion that the cutoff point is invalid:

If it had been shown that the exam measures ability with sufficient differentiating power to justify rank-ordering, it would have been valid to set the cutoff score at the point where rank-ordering filled the City's needs. The justification would be that each incremental change in score represents an incremental change in job-related ability, so that, for any given cutoff (even one determined solely by hiring needs), those who passed would likely perform the job better than those who failed. But the City can make no such claim, since it never established a valid basis for rank-ordering.

Guardians, 630 F.2d at 105. This is not to say that an exam must necessarily be valid for rank-ordering purposes in order to be valid for cutoff purposes. But the concession that rank-order. ing was invalid means that the scoring of the exam is weakly, if at all, correlated with successful job performance.

C.

In sum, we conclude that the stipulated facts establish as a matter of law that plaintiffs have made out a prima facie case of disparate impact, and that the City cannot document that the physical agility test was content valid.¹⁹ In so concluding, we would like to emphasize that by this opinion the Court does not presume to exercise the prerogatives of police superintendent or to second-guess the police superintendent's exercise of his professional discretion. It is well within the police executive's authority to devise a physical agility test which would be justifiable as a device to screen police applicants. Certainly the job demands some minimum level of coordination and strength. However, Title VII requires that a test that is discriminatory be necessary to the job and carefully validated. See Harless, 619 F.2d at 616. Too often tests which on the surface appear objective and scientific turn out to be based on ingrained stereotypes and speculative assumptions about what is "necessary" to the job. Thus, tests which discriminate against protected groups must be thoroughly documented and validated in order to minimize the risk of unwarranted discrimination against groups which have been traditionally frozen out of the work force. While some physical agility test may properly satisfy Title VII's demands, the City has fallen far short of what is needed to document and thereby justify its use of the 1976 and 1979 BOTS physical agility tests. It is therefore appropriate to enter summary judgment in favor of the plaintiffs on the issue of liability.

III.

Having determined the liability issue, we turn to the question of appropriate relief. Plaintiffs seek back pay for the class and retroactive seniority for those to be hired under the consent decree. The statute permits, but does not demand, such relief, as it provides that the court may "order such affirmative action as may be appropriate, which may include, but is not limited to, reinstatement or hiring of employees, with or without back pay . . . or any other equitable relief as the court deems appropriate." 42 U.S.C. § 2000e-5(g). While backpay is discretionary, a strong presumption, "which can seldom be overcome," favors backpay. Liberles v. County of Cook, 709 F.2d 1122, 1136 (7th Cir. 1983). We may deny backpay relief in this case "only for reasons which, if applied generally, would not frustrate the central statutory purpose of eradicating discrimination throughout the economy and making persons whole for injuries suffered through past discrimination." Albemarle Paper Co. v. Moody, 422 U.S. 405, 421, 95 S.Ct. 2362, 2373, 45 L.Ed.2d 280 (1975). Despite this strong presumption favoring backpay, the City makes several arguments for reducing the total amount of backpay awarded.

A.

The City argues that it is entitled to a defense to backpay because it relied in good faith on an Illinois statute which required rank-ordering of applicants. The City relies on LeBeau v. Libbey-Owens Ford Co., 727 F.2d 141 (7th Cir. 1984), where the Court held that good faith reliance on a state statute may sometimes justify denial of backpay relief. We cannot accept the City's good faith defense.

The Illinois statute in question appears to have required some form of rank-ordering by the police department. See Ill.Rev. Stat. ch. 24, § 10-1-12 (1979).²⁰ In addition, Illinois law mandated some kind of test "of physical qualifications." Ill.Rev.Stat. ch. 24, § 10-1-7 (1979).²¹ We note at the outset that these statutes are relevant only to the City's rank-ordering of applicants. That is, the City's reliance on these statutes, if any, only extended to its practice of rank-ordering. The good faith defense does not extend to those women who failed the exam. The statutes certainly did not require that the City eliminate applicants via a physical exam which exerted an unjustified disparate impact on female applicants.

20 Ill.Rev.Stat. ch. 14, § 10-1-12 (1979) provided:

From the returns or reports of the examiners, or from the examinations made by the Commission, the Commission shall prepare a register for each grade or class of positions in the classified service of such municipality of the persons whose general average standing upon examination for such grade or class is not less than the minimum fixed by the rules of such Commission, and who are otherwise eligible. Such persons shall take rank upon the register as candidates in the order of their relative excellence as determined by examination, without reference to priority of time examination. (Emphasis added).

21 That section stated: "Such examinations . . . shall include tests of physical qualifications and health."

At present, a genuine issue of fact exists as to whether the City actually relied on these statutes in deciding to rank-order. Brooke Koons, the City's former Director of Personnel, swears in an affidavit that he believed that the above statutes required rank-ordering based on a physical agility test. However, Judith Witt, another City employee involved in the testing process, testified at her deposition that, although State law required a rank-order list, she did not believe that State law required that scores on a physical agility test be used to rank-order applicants. Witt Deposition at 112. Although there is a dispute about whether the City subjectively relied on these statutes, we will assume that it did so rely. Even under this assumption, the City is not entitled to a good faith defense.

In LeBeau, the Court reaffirmed the rule that the presumption favoring backpay can be overcome by "special circumstances," which include reliance on state laws. 727 F.2d at 149. "Special factors . . . usually include only circumstances where state legislation is in conflict with Title VII.'" Id. at 150, quoting, Stewart v. General Motors Corp., 542 F.2d 445, 451 (7th Cir. 1976), cert. denied, 433 U.S. 919, 97 S.Ct. 2995, 53 L.Ed.2d 1105 (1977). In considering these factors, the district court is to "balance the merits of the [Title VII] claim and its support in public policy against the hardship on an employer acting in good faith." Id.

We believe that "special circumstances" do not exist in this case because there is no conflict between the state statutes and Title VII. Even if we assume that the statute required rank-ordering based on a physical agility test, the statute does not conflict with Title VII. The statute does not require that the test exert an unlawful disparate impact on women applicants. Put another way, the statute cannot be read to relieve the City of the burden of constructing a test which is content valid under Title VII principles. The statute may have demanded a test, and it may have demanded rank-ordering, but it did not demand that the City shirk its duty of rank-ordering on the basis of a content-valid test. Thus, even if the City relied on this statute, it could not have relied on the statute to fail to use a valid test.

Because the Illinois statutes and Title VII did not conflict in this case, LeBeau is not controlling. In LeBeau, an Illinois statute explicitly limited the amount of overtime that women could perform. This statute on its face conflicted directly with Title VII, thus placing the employer in that case on the "horns of a dilemma." 727 F.2d at 147. The Seventh Circuit affirmed the district court's discretionary decision to immunize the employer from backpay liability resulting from its good faith reliance on the state statute. Because there was no dilemma in this case, as the City was not required to use a discriminatory test, we see no reason to depart from the strong presumption favoring backpay for Title VII violations.

B.

The City also argues that any award of backpay must be shaved to reflect the time that Thomas' EEOC complaint lay dormant in the Commission. She filed her charge with the EEOC on February 4, 1977, but the EEOC did not issue her "right to sue" letter until July 3, 1980, almost 3 1/2 years later. Under EEOC regulations, Thomas could have requested a right to sue letter 180 days after she filed her charge. See 29 C.F.R. § 1601.28(d) (1983).²² Relying on Kamberos v. GTE Automatic Elec., Inc., 603 F.2d 598, 603 (7th Cir. 1979), the City argues that it should not be ordered to pay back wages covering the period during which Thomas could have sued, because that would prejudice it.

22 Arguably, Thomas had no such right to request a notice. Between 1977 and 1980, the EEOC regulation which was at issue in Kamberos, infra, applied only to charges against private employers. See 29 C.F.R. § 1601.28(a) (1980). No right to request a notice was mentioned in connection with suits against governmental respondents. See 29 C.F.R. § 1601.28(d) (1980). The regulations have since been amended and now appear to allow a complainant to request a notice, although the Attorney General may now have to issue the notice. See 45 Fed. Reg. 73037 (Nov. 4, 1980), now codified at 29 C.F.R. § 1601.28(d). In our discussion below, we have assumed that Thomas could have requested a right to sue letter even though it is unclear whether she in fact had that right.

In Kamberos, the Seventh Circuit vacated a district court award of backpay to the extent it was based on an over four-year period during which the plaintiff's EEOC charge had languished in the agency. The Court followed the approach of the Ninth Circuit in Lynn v. Western Gillette, Inc., 564 F.2d 1282 (9th Cir. 1977), in which that court held that the "`complainant should not be permitted to prejudice the employer by taking advantage of the [EEOC's] slowness in processing claims . . . particularly where the aggrieved party has consulted counsel and is aware of this right.'" Kamberos, 603 F.2d at 603, quoting, Lynn, 564 F.2d at 1287. The plaintiff in Kamberos was herself a lawyer and had retained counsel. "Under these circumstances," id. at 603, the Court held that the district court should have reduced the plaintiff's backpay award.

Kamberos does not demand a similar result in this case. Unlike Ms. Kamberos, Ms. Thomas is not a lawyer and was unrepresented while her charge was before the EEOC. She retained counsel only after filing this suit pro se. No evidence suggests that she knew of her right to request a right to sue letter. We do not think she and the class should be penalized for not knowing about the right to request the letter. See Wangsness v. Watertown School Dist. No. 14-4, 541 F. Supp. 332, 340-41 (D.S.D. 1982) (distinguishing Kamberos and Lynn where plaintiff was not a lawyer and had been unrepresented before EEOC). The language of Lynn supports this result. The Court held that a district court may properly exercise its equitable discretion and reduce a backpay award where a complainant takes advantage of the Commission's slowness or procrastinates while knowing that the Commission intends to take no further action. 564 F.2d at 1287-88. Lynn and Kamberos are best understood as requiring reduction of backpay where a complainant has somehow knowingly benefitted from the EEOC's delay. No evidence in the record of this case suggests such procrastination or unfair play warranting an equitable reduction of backpay. Indeed, the parties' stipulations suggest otherwise. On February 19, 1980, the Commission found "reasonable cause" to believe that Thomas' charge was true. She filed her suit later that year. Thus, it cannot be said that Thomas procrastinated after the EEOC had wrapped up its investigation. Moreover, the EEOC and the City entered into a conciliation agreement in 1980, under which the City agreed to rescore the exams. While the conciliation process did not begin until that year because of the EEOC's delay, had Thomas requested a right to sue letter, a suit might have frustrated these eventual conciliation efforts and prevented the settlement. It is clear that Title VII favors administrative resolution of disputes. Thomas and the class should not be prejudiced for having waited out the eventual conciliation process, especially absent evidence that she knew of her rights to do otherwise. In sum, Kamberos and Lynn do not warrant a reduction of backpay.

C.

The City also argues that 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), forbids the class-wide make-whole relief plaintiffs seek. The City incorrectly exaggerates the reach of Stotts.

In Stotts, the Supreme Court vacated an injunction which had prevented the City of Memphis from following its seniority system in laying off city workers. The district court had held that the seniority based layoffs would have had a racially discriminatory impact, and it entered the injunction in order to either enforce or modify a consent decree from an earlier Title VII case. The Supreme Court, among other things, held that the district court lacked power to enter the injunction because there had been no finding that any of the black beneficiaries of the injunction had been an actual victim of the past discrimination. 104 S.Ct. at 2588-89. The remedy of competitive seniority is proper under Title VII "only when the beneficiary . . . has actually been a victim of illegal discrimination." Id. at 2588.

The City argues that under Stotts class-wide backpay relief is inappropriate in this case. That is plainly incorrect. We have held above that the members of the class were actual victims of discrimination. The women who failed the test, including Thomas, were rejected on the basis of a sexually discriminatory criteria which had not been properly validated. Similarly, the women who passed were ranked relatively low because of the discriminatory rank-ordering system. Any award of backpay or competitive seniority will be going to class members who took the 1976 and 1979 tests, who were thus actual victims of discrimination. No relief is going to women who did not participate in those tests. In sum, then, Stotts has no bearing on this case.

D.

In light of the preceding sections, we conclude that it is appropriate to follow the general presumption of awarding backpay to the class. Backpay is necessary to fulfill one of Title VII's central purposes, that is, "`to make persons whole for injuries suffered on account of unlawful employment discrimination.'" Franks v. Bowman Transportation Co., 424 U.S. 747, 763, 96 S.Ct. 1251, 1263, 47 L.Ed.2d 444 (1976), quoting, Albemarle Paper Co., 422 U.S. at 418, 95 S.Ct. at 2372. We cannot say with certainty which of the individual class members would have been hired but for the illegal discrimination. However, in spite of our ignorance of which individuals would have been hired, a general award of backpay to the class may properly be awarded, since some class members doubtless would have been hired. See Liberles, 709 F.2d at 1136; Hameed v. International Ass'n, 637 F.2d 506, 520 (8th Cir. 1980) (where identifying which class members would have been hired is complex and uncertain, a class-wide backpay remedy is appropriate); Stewart v. General Motors Corp., 542 F.2d 445, 452-53 (7th Cir. 1976), cert. denied, 433 U.S. 919, 97 S.Ct. 2995, 53 L.Ed.2d 1105 (1977).

To determine the class of actual persons discriminated against, we may estimate the number of women that would have been hired but for the discriminatory test; mathematical precision is not necessary. See, e.g., Hameed, 637 F.2d at 520-21; Stewart, 542 F.2d at 452. The City hired 20 males from the 1976 list. Had the City's selection policy been non-discriminatory, the percentage of women hired would have roughly equaled the percentage who applied.²³ See Hameed, 637 F.2d at 520. The parties agreed that 14.68% of the applicants who took the physical agility test were women. It also is undisputed that the 20 males hired from the 1976 list earned $1,281,058.90 through 1982. Thus, the class members who took the 1976 test are entitled to ($1,281,058.90) X (.1468), or $188,059.44.

23 The City argues that any attempt to estimate the number hired from the number who took the test would be speculative, since the application process included many other steps, including a written test, interview, background check, etc. This is merely a repeat of the causation argument, which we dealt with in note 7 above by holding that the physical agility test did in fact cause a disparate impact in hiring. Moreover, at this "remedial" stage of the case, the City bears the burden of showing that the class-wide results would have been the same even absent discrimination. Cf. Caviate v. State of Wisc. Dept. of Health Social Services, 744 F.2d 1289, 1296 (7th Cir. 1984). The City has made no such showing.

Moving to the 1979 test, the parties agree that the City hired 14 men and one woman. Absent discrimination, about 13.25%, of those hired would have been women, since that was the percentage of female applicants. It is uncontested that the 14 males hired off the 1979 list earned $759,270.19 through 1982, while the one woman earned $13,470.77. Thus, the class members who took the 1979 test are entitled to $759,270.19 X .1325, or $100,603.30, less the $13,470.77 earned by the one woman, or a total of $87,132.53.²⁴

24 If some women took both the 1976 and 1979 tests, this figure should be adjusted to eliminate double-counting of back wages.

The above calculations should be updated through the date of the final determination of damages. This figure will then have to be diminished to account for the actual earnings of class members during the relevant periods. According to plaintiffs, more discovery is needed to determine these interim earnings. One way to make the calculations is to compile figures from forms submitted by individual class members; once the figures are compiled, the class-wide award could be diminished on a class-wide basis.²⁵ See Hameed, 637 F.2d at 521. We will enter an order referring this case to a Magistrate for a recommendation to the Court as to the proper amount of mitigation, as well as to whether prejudgment interest should be added to the final amount of damages.²⁶ The Magistrate shall have authority to allow limited discovery for these purposes and to conduct any necessary hearings. The Magistrate's report and recommendation, with a proposed final draft order, shall be submitted within 90 days.

25 According to the City, plaintiffs' counsel represents a similar class which alleges discrimination on the written test. The City is concerned about duplicative damages from the two suits. It is unclear to what extent the classes overlap. This is not relevant now. In any event, any problems of duplication can be handled by the judge presiding over the other case. As of now, it is speculative whether plaintiffs in the other suit will recover anything.

26 Plaintiff's claim that under Heiar v. Crawford County, 746 F.2d 1190, 1202-03 (7th Cir. 1984), they are entitled to prejudgment interest. We express no opinion on the issue now. It is clear that under Heiar prejudgment interest is neither to be granted nor denied automatically. In Heiar the district court had denied prejudgment interest because "the benefits of a later lump sum settlement which may be received and which can be invested offsets the diminished value of those dollars which may have been paid at an earlier date." Id. at 1202. The Seventh Circuit vacated this order, holding that the district court had not fully exercised its discretion on the issue. The Court noted that, in general, the value of money diminishes overtime, but that equitable concerns or other factors in a given case may override an award of interest. In sum, Heiar holds that the question of interest must be resolved on a case-by-case basis. At present, the Court has no record upon which it can base a ruling on this issue. Accordingly, this issue too is referred to the Magistrate for briefing and/or a hearing on whether prejudgment interest is proper in this case.

We agree with the plaintiffs that under the "make whole" purposes of Title VII, an award of retroactive seniority to the class members hired under the consent decree is appropriate. See Franks, 424 U.S. at 763-67, 96 S.Ct. at 1364-65; Romasanta v. United Air Lines, Inc., 717 F.2d 1140, 1147 (7th Cir. 1983), cert. denied, ___ U.S. ___, 104 S.Ct. 1928, 80 L.Ed.2d 474 (1984). Under Franks, we must presume that retroactive seniority is appropriate. 424 U.S. at 729 n. 41, 96 S.Ct. at 1271. We may deny or limit seniority if the City shows that such relief would unusually harm the interests of other employees. Id.; Romasanta, 717 F.2d at 1148-1150. The City has made no such showing in this case. We conclude that full retroactive seniority — both "benefit" and "competitive" — is appropriate to put those hired under the consent decree in the position they would have been in had they been hired when they applied. Those class members who unsuccessfully applied in 1976 and are hired under the decree should receive a seniority date of November 1, 1977, which is when the City first hired men from the 1976 list. Those who unsuccessfully applied in 1979 should, after hired under the consent decree, receive a seniority date of April 14, 1980, as that is when the City first hired officers from the 1979 list.

E.

The last two defenses raised by the City can be readily disposed of. First, the City argues that the challenge to the rank-ordering is not "like or reasonably related to" Thomas' EEOC charge, and therefore must be dismissed under Jenkins v. Blue Cross Mutual Hospital, Inc., 538 F.2d 164, 167 (7th Cir. 1976) (en banc), cert. denied, 429 U.S. 986, 97 S.Ct. 506, 50 L.Ed.2d 598 (1976). This argument is plainly wrong. The Jenkins test is broad, as it grows out of the policy of affording relief to the pro se Title VII complainant. See, e.g., Oxman v. WLS-TV, 595 F. Supp. 557, 559-60 (N.D.Ill. 1984). In this case, Thomas' EEOC charge clearly challenged the use of the physical agility examination. Although she did not mention the practice of rank-ordering, that aspect of the exam is "like or reasonably related to" the aspect complained of in the charge: that the exam failed Thomas unlawfully. In any event, we think the City has waived the Jenkins argument, as it waited four years before raising it, following all of the discovery and briefing about the rank-ordering process.

Second, the City makes a related argument that Thomas has no standing to challenge the rank-ordering because she failed, and thus was not directly affected by the rank-ordering. The City relies on General Telephone Co. of Southwest v. Falcon, 457 U.S. 147, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982), in which the Supreme Court held that class certification was improper where a named plaintiff was harmed by an employer's promotion practices, but the class was allegedly harmed by the employer's hiring practices. Falcon does not render Thomas an improper class representative. Both she and her class were harmed by the same discriminatory test, albeit in slightly different ways. Her interests are consistent with those women who passed, but who suffered from the rank-ordering process. Indeed, Thomas too was "ranked," but in a way which eliminated rather than diminished her chances of being hired. The Falcon court recognized that class relief is proper where a biased testing procedure affected class members in slightly different ways. 457 U.S. at 159 n. 15, 102 S.Ct. at 2371. In sum, Thomas meets Falcon's requirement that her claim "fairly encompass" the class claims. Id. at 158, 102 S.Ct. at 2370-71. Moreover, Judge Flaum already considered and rejected the City's arguments in certifying the class. Memorandum Opinion at 5-8 (Feb. 28, 1983). His holding is the law of the case, and the City has given us no reason to depart from it.

IV.

We find that the City's use of a physical agility test violated Title VII. We therefore grant the plaintiffs' motion for summary judgment and deny the City's cross-motion. The sub-class which took the 1976 test is entitled to $188,059.44; those who took the 1979 test are entitled to $87,132.53. Both figures must be updated to this date and then adjusted to reflect mitigation earnings. The case will be referred to a Magistrate, as provided for above. It is so ordered.