Berkman v. City of New York

626 F. Supp. 591 | E.D.N.Y | 1985

626 F. Supp. 591 (1985)

Brenda BERKMAN, Plaintiff,
v.
The CITY OF NEW YORK et alia, Defendants.

No. CV-79-1813.

United States District Court, E.D. New York.

October 8, 1985.

Women's Rights Clinic, Washington Square Legal Services, by Laura Sager and Debevoise & Plimpton, by Robert L. King, Bart Schwartz, New York City, for plaintiff.

Booth Lipton & Lipton, by Charles J. Hynes, New York City, for plaintiff-intervenor Ahrens.

Frederick A.O. Schwarz, Corp. Counsel of the City of N.Y. by Norma Kerlin, Elizabeth Dale Kendrick, New York City, for City defendants.

Colleran O'Hara & Mills, P.C. by John F. Mills, Mineola, and Lipsig Sullivan & Liapakis, by Michael N. Block, New York City, for defendant-intervenor UFA Local 94.

Salles & Danzig, by Stuart Sales, New York City, for defendant-intervenor Firefighters Eligibles Ass'n List # 1162, Inc.

AMENDED MEMORANDUM AND ORDER

SIFTON, District Judge.

This action was commenced in 1979 to remedy alleged discrimination against women in the selection of New York City Firefighters in violation of Title VII of the Civil Rights Act of 1964, as amended, 42 U.S.C. § 2000e et seq.; the Civil Rights Act of 1871, 42 U.S.C. § 1983; the fourteenth amendment to the United States Constitution; and the New York Human Rights Law.

Plaintiff sued on her own behalf and as representative of a class of women who took and passed the written portion of the qualifying entry-level test for firefighters administered in 1977 ("Exam 3040") and, thereafter, either took and failed or were deterred from taking the physical portion of that test. On July 10, 1980, this Court determined that the action could be maintained by plaintiff on behalf of a class of women so described.

Trial of the action occurred in 1981, and on March 4, 1982, this Court determined that Exam 3040 violated Title VII. Berkman v. City of New York, 536 F. Supp. 177 (E.D.N.Y.1982), aff'd, 705 F.2d 584 (2d Cir. 1983). As a result, the Court directed defendants to hire up to 45 class members found to be qualified pursuant to a special qualifying examination to be agreed upon between the parties or, in the absence of agreement, to be determined by the Court. In addition, the Court directed defendants to commence forthwith preparation of a new qualifying exam that did not discriminate against women. On March 25, 1982, the Court entered a supplemental order directing the parties to cooperate in an effort to reach an accord as to the new selection procedure.

In August of 1982 the Court approved an agreement between the parties with respect to the qualifying test to be used as a basis for selecting the class members to be appointed to the fire department pursuant to the Court's prior decision. This test was administered to interested class members in September 1982, and from the class members passing the examination, some 38 women became firefighters.

Efforts to reach agreement with regard to the content of a new physical test for all entry-level firefighters were less successful. Accordingly, in October 1982, defendants presented a proposed new test to the Court for approval, and in January and February 1983 evidence was presented with respect to the validity of the new test. At the conclusion of seven days of testimony on the subject, the Court adjourned proceedings without a date to permit the parties to explore further the possibilities of arriving at an agreement as to the contents of the new test. By May 1983, with no agreement reached, defendants, out of concern for the age of the list of persons eligible for appointment created from the previous Exam 3040,^[1] proposed to proceed with the administration the new test (Exam 1162), which thereafter occurred over a period of several months.

During this same period of time, in September 1983, plaintiff and another class member, Zaida Gonzalez, were terminated at the conclusion of their probationary period ostensibly because of their poor performance as probationary firefighters. A motion made by plaintiff to compel both class members' reinstatement was heard by this Court in October 1983, which, on December 8, 1983, found that the two women had been subjected to intentional discrimination on account of their sex and directed their reinstatement. Berkman v. City of New York, 580 F. Supp. 226 (E.D.N.Y.1983).

Following the completion of the administration of the new test, plaintiff moved in June 1984 to enjoin the promulgation of a new eligibility list derived from it pending the completion of hearings with regard to the validity of the exam. On July 6, 1984, this Court entered an order restraining the promulgation of the eligibles list and appointments from it without prior court order, upon a showing of compelling necessity, pending determination of the validity of the exam. After further efforts to resolve the issues presented by the new test were aborted, hearings resumed and continued through the spring of 1985.

Based on the evidence adduced at the hearings with respect to the new exam, I conclude that it comports in general with the requirements of this Court's decision of March 1982 with the exception of its scoring which purports to distinguish between qualified candidates to a degree of exactness not consistent with the lack of precision inherent in the exam. Accordingly, defendants are authorized to promulgate an eligibles list created from Exam 1162 and to make appointments therefrom in the manner set forth below. What follows sets forth the findings of fact and conclusions of law on which these determinations are based as required by Rule 52(a) of the Federal Rules of Civil Procedure.

1. Standing. The first issue requiring consideration is the standing of plaintiff to seek a determination that Exam 1162 does not comply with this Court's orders directing the preparation and use of a new physical exam for entry level firefighters which does not discriminate against women. I conclude plaintiff has standing.

The decree which plaintiff seeks to enforce is one she obtained in protracted litigation in which she has consistently taken an active role both to secure her own rights and those of other class members. No suggestion has been made that she lacked standing to obtain the portions of the decree she now seeks to enforce, see Allen v. Wright, 468 U.S. 737, 104 S. Ct. 3315, 3331, 82 L. Ed. 2d 556 (1974); Gilmore v. City of Montgomery, 417 U.S. 556, 570 n. 10, 94 S. Ct. 2416, 2424 n. 10, 41 L. Ed. 2d 304 (1974). Nor has any argument been advanced that this Court lacks the jurisdiction retained in the original decree "to secure compliance with [the Court's original] order ... [and] for consideration of such further interim or remedial relief as may be necessary and appropriate." See, e.g., Green v. County School Board, 391 U.S. 430, 439, 88 S. Ct. 1689, 1695, 20 L. Ed. 2d 716 (1968); Guardians Ass'n of New York City Police, Inc. v. Civil Service Commission of New York, 630 F.2d 79, 109 (2d Cir.1980), cert. denied, 452 U.S. 940, 101 S. Ct. 3083, 69 L. Ed. 2d 954 (1981); Davis v. Board of Education, 674 F.2d 684, 680 (8th Cir.), cert. denied, 459 U.S. 881, 103 S. Ct. 178, 74 L. Ed. 2d 146 (1982).

Both objectively and subjectively, plaintiff has demonstrated a real and concrete interest in assuring that she and her fellow class members have a non-discriminatory work environment in which to carry out their functions as New York City firefighters, an environment which would be substantially undermined if the next class of women applicants for the position of entry-level firefighters are deprived of an opportunity to join the ranks because of their sex.

As already noted, there are now less than 40 women firefighters in New York City out of a total of approximately ten thousand. Because of the nature of the department's organization, no more than one woman firefighter has been assigned to any firehouse. As also previously noted, plaintiff herself and at least one other female firefighter were, after their appointment, subjected to intentional discrimination based on their sex in the all-male firehouses to which they were assigned. See Berkman v. City of New York, supra, 580 F. Supp. 226. Thus, plaintiff and the class she represents face the prospect of losing much if not all they gained if, as a result of discrimination in further hiring, they are left as an insular minority of women firefighters who "slipped through the [Department's] allegedly discriminatory screening practices." See Gray v. Greyhound Lines East, 545 F.2d 169, 173, 175-76 (D.C.Cir. 1976); Chicano Police Officers' Ass'n v. Stover, 526 F.2d 431, 437 (10th Cir.1975), vacated on other grounds, 426 U.S. 944, 96 S. Ct. 3161, 49 L. Ed. 2d 1181 (1976), standing aff'd, 522 F.2d 918, 921 (10th Cir.1977).

Contrary to the argument of defendants and intervenors, this is not a case like Allen v. Wright, supra, 104 S.Ct. at 3331, in which a plaintiff asserts no more than the stigmatic injury which she shares with any other woman seeking to work for the New York City fire department. Here, plaintiff has suffered injury not only in being denied appointment by virtue of a prior discriminatory entrance exam, but thereafter by discrimination in the essentially male-only workplace in which she functions, resulting in her termination. By her efforts to ensure that Exam 1162 is non-discriminatory, she seeks to ensure that she will preserve the gains she has won as a result of the earlier trial. The presence of increased numbers of qualified women in the workplace will demonstrate, in a fashion in which no judicial decree can be expected to, that hostility towards women firefighters is misplaced and furnish witnesses ready and willing to speak out against continued discrimination by defendants and their employees.

2. Disparate Impact. This Court's earlier decree required defendants to prepare a new physical exam for entry-level firefighters which did not discriminate against women. Accordingly, it is appropriate to consider whether the results of the exam present a prima facie case of sex discrimination and, if so, whether that prima facie appearance of discrimination may be explained by the requirements of the job. See Griggs v. Duke Power Co., 401 U.S. 424, 432, 91 S. Ct. 849, 854, 28 L. Ed. 2d 158 (1971); Albarmarle Paper Co. v. Moody, 422 U.S. 405, 425, 95 S. Ct. 2362, 2375, 45 L. Ed. 2d 280 (1975); Guardians Ass'n, supra, 630 F.2d at 88.

Here, the undisputed facts concerning the hiring that may be expected to result from the administration of Exam 1162 demonstrate a disparate impact on women directly attributable to the physical portion of the facially neutral test instrument because of the statistically disproportionate results for men and women.

The written portion of Exam 1162 was administered on September 11, 1982, to 31,421 candidates of whom 28,946 identified themselves as males and 566 identified themselves as females. 30,474 candidates or 98.58 percent passed the test, representing 98.65 percent of the male applicants (28,559) and 97.8 percent of the female applicants (554). In this connection, it bears noting that, despite a determination that the cognitive abilities tested by the written portion of the exam should be weighted equally with the results of the physical exam in determining which applicants were most qualified to be firefighters, the written exam failed lamentably to distinguish between the varying cognitive abilities of the applicants passing the exam. 83.21 percent of the candidates received a score of 90 or higher; 66.68 percent received a score of 94 or higher; the median score was between 96 and 97. 9,788 candidates received a score of 98 or above.

The physical portion of Exam 1162 was administered over several months to those candidates passing the written test with all female applicants being tested on two days, September 12 and 13, 1983. Unfortunately, the testing of the women applicants coincided with extensive publicity concerning discrimination on the job against plaintiff and her fellow class member Gonzalez, which eventually resulted in what this Court found, after hearings, to be retaliatory discharges on September 21, 1983 of the two women.

As occurred with respect to the prior physical exam, 3040, a surprisingly small proportion of the women passing the written exam presented themselves to take the physical exam. Thus, while 22,255 men, or 77.93 percent of those eligible, took the physical exam, only 165 women, or 29.78 percent of the eligible women, did so. Had women candidates eligible to take the exam presented themselves in the same proportion as did their male counterparts, 432 women, rather than 165, would have taken the test.

Out of the total of 21,313 candidates, both men and women, taking the physical exam, 21,236 men and 77 women passed it. The pass rate for men was 95.42 percent; the pass rate for women was 46.67 percent. Thus, the pass rate for women on the physical exam was less than 80 percent, i.e., less than four-fifths of the pass rate for men, and the pass rates for the two groups were separated by at least 27 standard deviations. Accordingly, it is clear that the physical exam had a disparate impact on women. See Castenada v. Partida, 430 U.S. 482, 497 n. 17, 97 S. Ct. 1272, 1281-82 n. 17, 51 L. Ed. 2d 498 (1977); EEOC Uniform Guidelines on Employees Selection Procedures (the "Guidelines"), 29 C.F.R. § 1607.4(D).

3. Relation to the Job. As was the case in the trial of the underlying action, the parties have not attempted to prove all that the job of firefighting in New York City consists of so as to compare it with the test to establish whether the test is, in fact, job related. Instead, the parties have concentrated their efforts on the issue of test preparation in order to argue whether the methods of test preparation engaged in by the parties were or were not of the sort from which a job-related test could be reasonably expected to result. The distinction is significant because in the case of Exam 1162, unlike Exam 3040, all parties to a greater or lesser extent participated in the test preparation and did so with substantial amount of job knowledge and sophistication about job testing, acquired as a result of months of involvement in this litigation. Plaintiff's efforts to measure the validity of Exam 1162 solely by reference to defendants' compliance with a variety of standards for the preparation of neutral tests, including the EEOC Guidelines, the American Psychological Association's Standards for Education and Psychological Tests (1974) (the "APA Standards") and its Principles for the Validation and Use of Personnel Selection Procedures (2d ed. 1980) (the so-called "Division 14 Principles") ignores the very substantial role she and her experts and attorneys played in the preparation of Exam 1162. And, while plaintiff points to the drawbacks of preparing Exam 1162 in the context of ongoing litigation in terms of the antagonism she and women applicants for employment faced as a result of her role in bringing to light sex discrimination in the fire department, she fails to place proper emphasis on the opportunities she enjoyed as a result of the adversarial process to prevent discriminatory impulses from realizing themselves in any blatant way in the process of test preparation.

As plaintiff concedes the qualifying test pursuant to which she and a sizeable portion of the class interested in becoming firefighters in fact did so "established the framework for Exam 1162." Given the remarkable degree of success with which the class members performed on the qualifying test and the opportunities afforded plaintiff to improve upon it in developing a new entrance exam, it is difficult to accept plaintiff's position that so much of Exam 1162 as derives from the qualifying test is not job-related and valid.

The qualifying test proposed by the City to this Court on May 25, 1982, consisted of two parts, a simulation of engine company tasks and a simulation of ladder company tasks, separated by a 5-minute rest interval. The engine company simulation consisted of two tasks: dragging a length of 3½ inch hose and lifting a roll of 2½ inch hose from the ground and carrying it upstairs. The ladder company simulation consisted of five tasks: raising a portable ladder, climbing a preset portable ladder with a hook, entering a window and climbing stairs to the fifth floor, hitting a 80 pound rolled hose with an 8 pound maul down the length of a 6 foot table, and dragging a 145 pound dummy around a marked course. The City proposed a cutoff time of 4 minutes, 9 seconds to complete the test.

There can be little doubt that the qualifying test was developed under the pressures of litigation without great attention to the guidelines, standards and principles available to test preparers proceeding on a more leisurely basis, and, yet, there is little reason to think that the test proposed did not, in the main, adequately test for the important behavioral requirements of the job.

For one thing, the qualifying test, unlike Exam 3040, did not purport to test for the abstract physical endowments necessary to do firefighting work. Instead, it was founded on tasks in many important respects like those engaged in by firefighters on the job.^[2] Whatever the merits of a test of this sort in situations in which only one or none of the parties has previously studied the important work behaviors representative of the job, in a situation such as that presented here in which all involved had been studying that subject exhaustively as a result of ongoing litigation, a jobsample test had the advantages of both facial validity (securing acceptance of test results both from those who take it and those for whose benefit it is taken) and recognized content validity for those who have become job experts in the course of ongoing litigation.

In fact, the City's proposed qualifying examination was speedily content-validated in this sense shortly after its initial presentation. Plaintiff's expert, Dr. Raymond Mendel, after a week of revisiting City firehouses and observing job behavior with particular attention to the tasks identified as important in the City's proposed test, criticized several aspects of the City's proposal including the forcible-entry simulation, the rolled-hose carry, and the pace at which the test was to be performed. As a result of these criticisms, the parties agreed to modifications of the test by having the candidates lift the 2½ inch rolled hose from shoulder height, rather than from the ground; reducing the weight of the rolled hose in the forcible-entry simulation from 80 to 60 pounds; and increasing the length of the rest period from 5 to 7½ minutes.

The qualifying test became Exam 1162 with two additions and two modifications. A wall vault and hose pull were added. In addition, the 7-minute rest interval was converted to a 2-minute rest interval. Finally, instead of being a qualifying test rated on a pass/fail basis, Exam 1162 became a competitive event, rated on the basis of a candidate's time to complete the series of evolutions.

The wall event originally proposed by the City was the 5 foot wall climb which formed the crucial part of Exam 3040 and had been the subject of extensive criticism in connection with the proceedings leading to the invalidation of that exam. Accordingly, the defendants not surprisingly acceded to plaintiff's demands that the height of the wall be reduced to 4½ feet in order to conform more closely to the tasks of firefighters in climbing over obstructions in the course of responding to a fire. At the same time, defendants also acceded to plaintiff's demands that a somewhat duplicative walk through or over tires, initially proposed as an additional test, be eliminated, in part at least because of plaintiff's argument that such a test, drawing on the experience of men in the military and in training for competitive athletic contests, would adversely impact women who lacked such a background.

The hose pull was added because it had been part of the set of job performance evaluators used in Philadelphia from which other essential elements of the qualifying test had been drawn, which had been eliminated only because of a misunderstanding on the part of a representative of the City's Department of Personnel as to the manner in which a hose pull was performed by firefighters in New York City. According to credible testimony of Dr. Phillip Siegel of the Department of Personnel, the hose pull was left out of the qualifying exam because of his understanding that, when the task of pulling an uncharged length of hose to an upper story of a building over a hose roller attached to a window or roof was performed in New York City, it was done by two or more firefighters working together. When this misunderstanding was corrected, the hose pull using a hose roller was restored to the series of evolutions used in Philadelphia to measure job performance. While plaintiff correctly faults defendants for the somewhat ad hoc manner in which the hose roller subtest was added and argues that its adverse impact on women was not adequately studied, she does not offer any persuasive evidence to answer defendants' proof that the task does, in fact, reflect an important firefighting activity and calls upon physical capacities, namely, upper body strength, required by a number of other firefighting tasks not otherwise tested in the battery of physical tests developed.^[3] Moreover, plaintiff's complaints concerning the inadequacy of the job analysis performed by the City as a basis for adding the hose pull are considerably undercut by her acknowledgement that the subtest had its origin in the same series of job evaluators from which the other subtests reflecting a representative group of important job tasks were selected. Nor is the subtest as actually administered, after a familiarization program, of such complexity or difficulty as to justify fears that it tested learned behavior as opposed to a candidate's capacities to be trained to be a New York City firefighter.

The principal ground for contention between the parties relates to their differing approaches to an issue which first arose in connection with Exam 3040, namely, the issue as to the role of speed versus pacing in the performance of firefighting or, to put it another way, the need for anaerobic as opposed to aerobic energy in performing important firefighting tasks. On this issue, this Court had previously rejected the City's position that firefighting is an all-out sprint event performed at maximal heart rate, calling on reserves of aerobic energy in connection with Exam 3040,^[4] and plaintiff's arguments that defendants failed to develop an adequate job analysis with respect to this issue carry considerable force.

In its opinion invalidating Exam 3040, this Court found:

"Few jobs making large physical demands, least of all firefighting, are properly performed at maximum speed or at the limits of one's strength or endurance. As [plaintiff's] witnesses testified, what must be identified are not those who are strongest or fastest but, instead, those who, with the benefit of training in pacing or because of their natural capacities of endurance, can perform the punishing tasks of firefighting as they are actually required to be performed. According to these witnesses, firefighting takes its toll, not as a result of failures of maximum strength or speed, even at critical moments, but rather through the physical demands extending over long periods of time which necessitate paced performance at less-than-maximum levels. This explanation makes sense of a number of otherwise puzzling features of firefighting, namely, the ability of firemen to continue to do their work competently over an entire working career and the results of physical testing of incumbent male firemen which show that on a variety of measures of maximum physical capacity firemen rate no better than the average American male. It also makes sense in terms of the recognized dangers of firefighting and their unpredictability which, as numerous witnesses testified, make hazardous in the extreme performance at top speed or at the limits of strength or capacity. Not only does maximal performance compromise the firefighter's ability to pace performance over the long periods of time during which physical demands are being made on the body, in addition, the very unpredictability of fire and the instability of burning structures call upon qualities of foresight, endurance, and pacing not examined by tests of maximum physical strength."

Berkman v. City of New York, supra, 536 F.Supp. at 212 (footnotes omitted). Despite these findings and this Court's stated readiness to hear additional proof on the issue, defendants failed lamentably to establish a basis for the emphasis placed on maximal strength and speed in Exam 1162.

Three separate aspects of the physical portion of Exam 1162 are principally responsible for the test's emphasis on speed and anaerobic energy as the distinguishing characteristics of applicants meriting selection as a New York City firefighter: first, the notice to applicants that performance would be measured on time to complete the physical tests; second, the brevity of the test; and, third, the relatively narrow banding of the ranks from which appointments will be made. With respect to these three hotly contested aspects of the exam, defendants performed an inadequate job analysis.

The principal basis for the exam's emphasis on speed was a questionnaire distributed to a small group of job experts on October 18, 1982. Since defendants had already, in June 1982, distributed a job-analysis questionnaire to approximately 2,000 firefighters which could have but did not address the key issue of pacing, and since the distribution of the October questionnaire followed rather than preceded the defendants' application to this Court in October 1982 for approval of Exam 1162 on a rank ordered time to complete basis, there is considerable substance to plaintiff's contention that the questionnaire was conceived of as an effort to rationalize a decision to emphasize speed and anaerobic energy, rather than as a neutral inquiry to determine the role of pacing in the performance of actual firefighting tasks.

However, quite apart from the circumstances giving rise to the distribution of the questionnaire, its content hardly permits a conclusion that it constituted a neutral inquiry into the question of the pace at which firefighting is actually performed. Instead, the questions appear designed to lead respondents to answers that are supportive of defendants' oft-expressed contention that firefighting tasks are performed at top speed and at maximal energy levels. Thus, rather than ask the respondents to describe the pace (or even speed) at which various firefighting tasks are performed, the questionnaire describes the proposed new physical test for entry-level firefighters and then baldly asks whether firefighters performing those tasks should complete them in the "fastest possible time," assuming only that they are not hindered by smoke or obstructions. Quite apart from the dangers of alerting respondents to the significance of their answers in terms of constructing a rating system for a new physical exam and the ambiguity of the phrase "fastest possible time" in terms of what factors, if any, other than smoke and obstructions should be taken into account in framing an answer, the leading nature of the question quite obviously suggests its answer, and not surprisingly the largest number of respondents answered "yes." Nevertheless, the obvious qualitative common sense of a number of the comments offered by those able to resist the invitation to respond affirmatively should have indicated to a neutral observer both the ambiguity of the question and the need for further inquiry. Thus, among the responses from incumbent firefighters (on a questionnaire which did not even leave space for comments) were the following:

"What is considered maximum possible speed? On the fire scene most is done walking, not running, for safety reasons"; "Hose stretching at fire operations is never performed at maximum speed (or to point of exhaustion) because the truly arduous task begins at the apartment door"; "If operating at maximum speed one would tend to get emotional and emotions could interfere with clear thought"; "[F]ire operations require speed, tempered with safety and corrections of actions. The terms fastest possible time and maximum possible speed seem to represent an overemphasis on speed."

Essentially the questionnaire responses which form the basis for the exam's emphasis on speed and anaerobic activity do no more than reveal the obvious, that all else being equal everyone would like fires to be extinguished and victims extricated in the shortest time possible given the constraints of the job. However, it is precisely the question as to what the constraints of the job are that remained unexplored, with the exception of smoke and obstructions on stairs, together with the effect of those constraints on the speed with which firefighters actually do their work. The defendants' failure to explore in a neutral and above-board fashion this well-identified issue can only confirm this Court's earlier conclusion that a test that places undue emphasis on anaerobic energy and speed invidiously discriminates against women.

It is also the case that defendants failed to pay reasonable attention in constructing the test to the degree of emphasis placed on anaerobic energy by the manner in which the test was administered. Essentially, the physical test consisted of two bouts of violent physical exercise which applicants were encouraged to perform at maximum speed (and which, if performed at a pace meriting selection according to the City's rating scheme, had to be performed in approximately 90 seconds) separated by an enforced 1 minute 40 second recuperation period. As plaintiff points out, the effect of inserting the rest period between two episodes of anaerobic exercise converts the test into the equivalent of "wind sprints" in which persons who exhaust themselves by a maximum expenditure of anaerobic energy on the engine company simulation are given a recovery period permitting them to rely again principally on anaerobic energy to complete the ladder company portion of the physical exam. Defendants appear never to have considered that a test taking the same time to administer, but filled completely with strenuous physical activity, might not only resemble more closely the level of physical activity actually required of firefighters at fires, but also might enforce pacing on the candidates, giving some reading of the candidate's aerobic energy. In structuring the test, defendants appear to have ignored not only this Court's prior findings concerning the role of aerobic energy in performing firefighting functions but also the recommendations of the same expert whose Philadelphia evaluators appear to have been the principal source of both the qualifying exam and Exam 1162 that the test last 5 to 10 minutes without a recovery period.^[5]

What has been said concerning the demands of firefighting should not be taken as a finding that firefighting is entirely an aerobic activity, making no demands for the expenditure from time to time of anaerobic energy at maximal oxygen levels, but only that defendants' exam, including in particular their proposed rating system, places too great an emphasis on the anaerobic capacities of candidates with too little regard for those aspects of firefighting that call upon prudent, paced performance. Ironically, this finding is confirmed by defendants' criterion study which shows a lower correlation between raw scores achieved on the physical test and the performance of persons taking the test as evaluated by defendants' criterion measures than when a banding system was used, a subject discussed more at length below.

Finally, defendants appear to have paid singularly little attention to the emphasis placed on anaerobic energy by the narrow thirty-second banding of the scoring system selected and its extraordinarily adverse impact on women. Indeed, the scoring system selected by the City, given the realities of the hiring process, appears to have virtually assured that no women would be appointed from the eligibility list created for Exam 1162, despite the substantial number of women who passed the exam.

Based on past experience with other eligibility lists, defendants predict that they can reasonably expect to hire only some 2,800 firefighters from the total of 21,313 candidates who passed the exam. Again, based on past experience, the City predicts that offers of employment may be expected to be extended to approximately 6,500 applicants on the list in order to find the 2,800 candidates needed to fill the fire department's needs over the expected life of the eligibility list. To reach these top 6,500 candidates using the rank order proposed by the City, the City would get no farther down the list than to those candidates with a combined score on the written and physical tests of 94.5 or above. There are, however, 7,344 persons on the current list with a score of 94.5 or above, and, of this group, only two are women. Accordingly, the City's rank ordering decision appears of crucial significance in determining the adverse impact of the physical exam on women.

This Court's earlier decision directed defendants in their preparation of a new exam to validate the exam in general accord with the Uniform Guidelines including the Guideline's requirement that alternative measures with equal validity but with less adverse impact on the protected group be explored and used where available. 29 C.F.R. § 1607.14(B)(9). Accordingly, it is of particular significance that the alternative three-band scoring system proposed by plaintiff shows greater validity than the seven-band rating system proposed by the City. Use of the three-band rating system accordingly appears required by defendants' own criterion measures^[6] as well as by the failure of defendants in their job analysis and test preparation to give due consideration to the demands made on aerobic energy in performing firefighting tasks. Finally, the selection of firefighters from among broader rating bands will serve to redress the undue emphasis placed on the physical portion of Exam 1162 as a result of the extraordinarily indiscriminate written portion of the exam.

There remains for consideration what steps, if any, should be taken because of the extraordinary fall off of interest on the part of women candidates who passed the written exam when called on to present themselves for physical testing. Had women passing the written exam presented themselves for the physical exam in the same proportion as their male counterparts who passed the written test, some 432 women would have taken the physical exam as opposed to the 165 who actually did so. Here, even more obviously than was the case with respect to Exam 3040, it may be inferred that continued discrimination within the fire department in the form of highly publicized events leading to the termination of plaintiff and class member Gonzalez served to discourage women applicants from joining the fire department ranks. For this consequence of the fire department's continued discrimination against women some remedy appears appropriate.

While counsel for defendants has proposed offering the women deterred from taking exam 1162 a new chance to do so, a number of considerations weigh against this course of action. First, a substantial time lapse has occurred between the time the physical portion of Exam 1162 was administered to women and now, so that age and lack of conditioning will undoubtedly continue to deter a substantial number of women once interested in taking the test from taking it at this time. Equally important, nothing defendants have done in the interim since September 1983 has served to allay the fears generated at the time that the fire department has failed to correct the intentional discrimination against women in its ranks which this Court found existed in its decision restoring plaintiff and class member Gonzalez to the ranks. Finally, no women who passed the written exam but did not take the physical have come forward to express an interest in this alternative. Accordingly, it seems more appropriate to consider the interests of plaintiff and the class she represents, namely, the group of approximately 38 who are currently women firefighters, in fashioning a remedy that affords them colleagues who are committed to becoming firefighters and not easily deterred by the department's failure to eliminate discrimination in the ranks.

With these considerations in mind, it seems appropriate that, in the random selection process within each of the ranks to be established pursuant to this decision with respect to the physical portion of Exam 1162, each women who passed the exam will be afforded a somewhat greater chance of selection than her male counterparts. This will reflect the likelihood that, had 432 women presented themselves to take the exam, the number of women passing the exam would presumably have increased correspondingly from 77 (out of 165) to 201 (out of 432). Assuming that the 124 additional women would have been spread over the ranks in a pattern corresponding to the 77 women who actually took the exam and passed, each woman on the eligibility list should be afforded an increased opportunity over an equally ranked male to be selected in a ratio of 2.62 to 1.

The parties have apparently not explored the impact on women of using a three-band system with respect to the physical portion of Exam 1162 in combination with the results of the extraordinarily indiscriminate written exam. Since the written exam fails to make meaningful distinctions between various candidates' cognitive ability, it may be that rating the written exam on a pass/fail basis or in bands will result in a test with equal validity but less adverse impact on women. The parties are accordingly directed to address themselves to this issue and to settle an order on ten days notice together with supporting briefs and, if necessary, additional evidence that is consistent with this opinion.

In summary, defendants have constructed and administered a physical examination that tests for the important physical capacities necessary to be a successful firefighter. However, the scoring system used by defendants places undue emphasis on certain physical capacities not required for the job. Accordingly, defendants are directed to develop a new scoring system, using the three-band division of the raw scores on the physical portion of Exam 1162 proposed by plaintiff in combination with the results of the written portion of the exam scored on a basis that has equal or greater validity than the present scoring system and the least adverse impact on women. Selection of women passing Exam 1162 using this scoring system shall compensate for the impact of defendants' discriminatory conduct in deterring qualified women candidates in the manner outlined above.

SO ORDERED.

NOTES

[1] Pending the creation of a new exam, appointments of male firefighters continued to be made from the eligibility list created from Exam 3040, with corresponding increases in the number of positions open to females who passed the qualifying exam.

[2] In fact, the principal source of the City's proposed test appears to have been a criterion test used by the City of Philadelphia to measure performance by incumbent firefighters. While there are, of course, dangers in using a job performance measure to choose between applicants for entry-level employment, those dangers appear to have been in the main eliminated here because of the simplicity of the test as modified and the familiarization program afforded applicants prior to the test's administration.

[3] For example, it appears undisputed that overhauling and pulling of ceilings form an important part of firefighting in New York City and call upon reserves of upper body strength not otherwise tested in Exam 1162.

[4] As previously noted, the facts that the qualifying exam included a 7-minute rest period and was rated on a pass/fail basis militate against any finding that the successful performance of plaintiff and other class members on the qualifying exam demonstrates that all-out speed as opposed to pacing is of the essence in firefighting.

[5] Significantly, Chief Homer Bishop testified that firefighters need at least 5 to 10 minutes to get a one-room fire "under control," not counting additional time needed for overhauling and take-up operations. Other evidence established that the average time spent by firefighters at a "good job" is 40 to 60 minutes with a range from 25 minutes to 2 hours or, in unusual cases, 7 hours.

[6] Plaintiff launches a considerable attack on the defendants' criterion measures, relying principally on her concerns that the "highly polarized" atmosphere prevailing within the department at the time the criterion studies were performed and the small sample of women available within the department to be evaluated render the results of the criterion studies suspect or worse. In ideal circumstances, certainly such concerns are well-founded: a more neutral atmosphere and a larger sample of female firefighters would, of course, be desirable to ensure reliability of the City's criterion studies. However, test preparation and the application of the Guidelines must go forward in the real world including situations such as that presented here where a charged and contentious atmosphere results from ongoing litigation and the limited number of female incumbents results from the relatively recent interest of women in firefighting work. What the Guidelines enjoin are reasonably competent efforts to prepare a non-discriminatory exam, not perfection. Plaintiff's arguments that the criterion studies were deliberately rigged to validate Exam 1162 are not borne out by the evidence of record.