This appeal by defendant-intervenor Uniformed Firefighters Association Local 94 ("UFA") questions the propriety of certain injunctive relief granted by the United States District Court for the Eastern District of New York, Charles P. Sifton, Judge, in favor of plaintiff Brenda Berkman and a class of women who sought to become firefighters in the New York City Fire Department, against defendants City of New York and certain individuals responsible for hiring firefighters for the City of New York (hereinafter collectively "the City"), to remedy unlawful employment discrimination on the basis of sex, in violation of Title VII of the Civil Rights Act of 1964 ("Title VII"), 42 U.S.C. Secs. 2000e to 2000e-17 (1976 & Supp. IV 1980). Berkman contended that the physical part of a 1978 Fire Department entrance examination called "Exam 3040" was not job-related and had a disparate impact on women. After a bench trial, the district court ruled, in an opinion reported at 536 F. Supp. 177 (1982), that the City's use of the physical portion of Exam 3040 violated Title VII. The court therefore enjoined further use of the eligibility list compiled pursuant to the exam ("Eligibility List 3040") "except upon a showing of compelling necessity," id. at 218, directed the City to develop a valid physical test, and ordered the City in the interim to appoint as entry-level firefighters up to 45 of those class members "who are found to be qualified for appointment and willing to be appointed," id. (footnote omitted). The court directed the parties to attempt to agree upon procedures for determining which of the class members were qualified for appointment. Id. Thereafter, the City and Berkman agreed on an interim qualifying physical test to be administered to the class members, and the district court approved their agreement.

The present appeal is taken by UFA, an employee organization representing current firefighters, which was allowed to intervene below in order to participate in the remedy phases of the proceedings. Pursuant to a preargument agreement among UFA, Berkman, and the City, and approved by the Court,1 the issues on this appeal are limited to the propriety of the remedial measures ordered by the district court.2 For the reasons below, we affirm.

The history of the City's use and development of tests for the selection of its firefighters is fully set out in the district court's comprehensive opinion, 536 F. Supp. at 180-204, familiarity with which is assumed. At issue in the present case was the physical portion of Exam 3040 given by the City in 1978. The written portion of Exam 3040 had been taken in December 1977 by 24,758 men and 410 women. Nearly 98% of the men (or 24,252), and 95% of the women (or 389), passed this portion, and all who had done so were eligible to take the physical portion. Of the men so eligible, 74%, or 18,060, took the physical test. Of the eligible women, only 23%, or 88, took the physical test. Of the men who took the physical exam, 16,925 completed it and 7,847, or 46% of those who completed it, passed. Of the 88 women who took the physical exam, 79 completed it and none passed.

The present action was commenced by Berkman in 1979. In 1980 the district court certified a plaintiff class consisting of women who took the written portion of Exam 3040 and either took and failed the physical portion of Exam 3040 or were deterred from taking it as a result of sex discrimination by the City. Berkman thereafter moved for preliminary injunctive relief prohibiting the City pendente lite from hiring additional firefighters from Eligibility List 3040; this motion was denied after the City agreed to reserve 50 entry-level firefighter positions for members of the plaintiff class in the event Berkman should prevail on the merits. In April 1981, UFA moved to intervene in the action, and its motion was granted to the extent of permitting it to participate at the remedy stage of the case, as well as in earlier proceedings having a bearing on the nature of the remedies to be granted.

After extensive discovery, a bench trial consuming several weeks was held between September and November 1981. The district court rendered its decision in an opinion dated March 4, 1982. See 536 F. Supp. 177. The court found, inter alia, that Berkman had established a prima facie case of discriminatory impact under both the Uniform Guidelines on Employee Selection Procedures ("Guidelines"), 29 C.F.R. Sec. 1607.4(D) (1981), promulgated by the Equal Employment Opportunity Commission ("EEOC"), and the statistical test stated by the Supreme Court in Castaneda v. Partida, 430 U.S. 482, 497 n. 17, 97 S. Ct. 1272, 1281 n. 17, 51 L. Ed. 2d 498 (1977). 536 F. Supp. at 205-06. The court found that the City had failed to rebut plaintiff's case since it did not establish that the physical portion of Exam 3040 was job-related. The City had failed to produce either "data showing that the content of the selection procedure [was] representative of important aspects of performance on the job for which the candidates [were] to be evaluated," 29 C.F.R. Sec. 1607.5(B) (sometimes referred to as "content validation"), 536 F. Supp. at 206-07, or "empirical data demonstrating that the selection procedure [was] predictive of or significantly correlated with important elements of job performance," 29 C.F.R. Sec. 1607.5(B) (called "criterion validation"). 536 F. Supp. at 208. On the basis of these findings the district judge concluded that the physical portion of Exam 3040 violated Title VII, and he granted the injunctive relief that is the subject of this appeal in orders dated March 25, 1982 ("March Order"), and August 3, 1982 ("August Order").

The March Order granted immediate, long-term, and interim relief. Effective immediately, the court permanently enjoined further use of Eligibility List 3040 except upon a showing of compelling necessity. As long-term relief, the court directed the City to begin preparation of new, properly validated selection procedures that would have the least adverse impact on women. For the interim, the court ordered the City to reserve entry-level firefighter positions for the hiring of up to 45 members of the plaintiff class who remained interested and were adequately qualified.

The court arrived at the number 45 with respect to interim hiring as follows. It found that some women who had passed the written portion of Exam 3040 had been deterred from taking the physical portion by pre-test publicity suggesting that no woman could pass the physical test.3 536 F.Supp. at 217. The district court correctly concluded that a "sex-neutral exam would have been followed by an equal decline (26%) in interest between both groups." Id. The court reasoned that absent the actual and anticipated disparate impact of the physical test, the decline in interest among women who had passed the written portion could reasonably have been expected to be proportionately no greater than the decline in interest among the men who had passed the written portion. Since 74% of the men who had passed the written portion of Exam 3040 took the physical portion, the court estimated that 74% of the 389 women who passed the written portion, or 288, would have taken a sex-neutral physical test. Then, noting that 16% of the men who had taken the physical test had been called up for appointment,4 and taking 16% of 288, the court arrived at "a figure of close to 45 women who one would expect to have been called up under a sex-neutral exam." Id.

The March Order directed the parties to attempt to agree upon a procedure for determining which members, "if any," of the plaintiff class were adequately qualified to become entry-level firefighters and hence to be appointed to the reserved positions. Paragraph 4 of the order provided, in pertinent part, as follows:

4. In order to determine which, if any, members of the plaintiff class are entitled to further remedial relief, the parties shall forthwith determine which members of the plaintiff class (a) continue to be interested in the position of firefighter, and (b) are adequately qualified for that position. In this connection plaintiff shall, on or before April 2, 1982, serve and file on 3 days' notice to all parties its proposed form of notice to the class, directing members to notify plaintiff's counsel of their intent to qualify for appointment within a reasonable period of time to be specified in the notice. In the event that the parties are unable to agree expeditiously upon procedures for the determination of which class members responding to such notice are adequately qualified, then the procedures for determination of whether interested class members are adequately qualified for the position shall be, on notice to the parties, established by the Court. In the event that more than 45 members of the plaintiff class are found to be interested in and adequately qualified for appointment to the position of entry-level firefighter, notices of appointment shall be sent to 45 of those qualifying, selected by lot. In the event that 45 or fewer members of the plaintiff class are found to be interested in and adequately qualified for appointment, notices of appointment shall be sent to all those qualifying.

The court refused to order that the City affirmatively recruit women to fill the 45 reserved places, finding that Berkman had failed to prove either a pattern of significant prior discrimination or that the City's discrimination against women had been intentional. 536 F. Supp. at 217.

The August Order approved a stipulation between Berkman and the City agreeing on an interim physical test to be used to determine which members of the plaintiff class were "adequately qualified" to be firefighters and hence could be appointed to the reserved positions. The events leading to the August Order began in the wake of the March Order. Berkman and the City had different ideas as to the most appropriate procedure for determining which members of the plaintiff class were adequately qualified to become firefighters. The City favored development of a new physical exam, whereas Berkman sought a training program in which interested class members would participate and have their fitness to be firefighters evaluated solely on the basis of their performance in that program. Each side proceeded to develop its own proposal.

The City's Department of Personnel undertook a review of the literature regarding firefighter testing, existing analyses of the New York City firefighter's job, and the district court's opinion in the present case. Eventually, after consultation with members of the Fire Department, UFA, and outside consultants, the City devised a two-part physical exam, as described in the margin.5 This exam was itself tested on a group composed of 37 incumbent firefighters, selected to be representative of members of the Fire Department with respect to assignment, age, and years on the force. The testing was supervised by 20 superior officers, who were instructed as to what constituted acceptable and unacceptable performance on each subtest. A score of 4:09 minutes for performance of all the tests was eventually set as the cutoff passing score.6 This was derived from the distribution of the test scores of the incumbent test-testers, the recommendations of the Fire Department officers who had supervised the incumbents in the test, and the recommendations of a psychometrician retained as a consultant by the Personnel Department.

In the meantime, Berkman, who knew the City was working on an interim test but did not know its contents, continued to press for a physical training program. The City met with Berkman's counsel several times to discuss the nature and cost of Berkman's proposal, but did not disclose any information regarding its interim test prior to May 21, when the test was finalized. It did not invite Berkman to participate in the test development.

On May 21, in an effort to have interim hiring qualification procedures commenced without further delay, Berkman moved to have the district court approve her proposed physical training program. The City quickly cross-moved to have its proposed interim test approved. On May 26, the district court held a hearing on the applications, at which Berkman, the City, and UFA were represented.

The City urged approval of its proposed interim test, arguing that it was job-related and would permit the City to select persons who would be able to perform firefighting duties in New York City. Berkman adhered to her preference for a training program, citing her lack of participation in the development of the City's proposed test, her consequent uncertainty as to the validity of the test, and her desire to end delay in gaining the interim relief granted by the March Order.

The court expressed concern over the secrecy with which the City's test had been prepared, stated its preference for a qualifying procedure based on "observations over a substantial training period" (May 26, 1982 Tr. at 32), and questioned the fairness of asking class members to take a physical test on short notice and without training procedures. The court reserved decision as to each side's proposal and encouraged the parties to attempt a compromise.

At the end of this hearing, UFA, which had offered no proposal of its own and had submitted no papers, testimony, or argument either for or against either proposed qualifying procedure, appeared to favor the City's proposed test. When the court indicated that it would reserve decision and asked counsel to return the next day, counsel for UFA suggested that perhaps a decision by the court could be obviated if Berkman's experts were to review the City's proposed exam and find it acceptable:

MR. TEITLER [counsel for UFA]: Your Honor, before a determination is made by the bench, the plaintiffs have indicated that they haven't had their experts have time to go over the exam broached by the City.

Wouldn't it be more advantageous to see if between us, if their experts review the exam proposed by the City and find it acceptable that it might give a little more leeway as to what the result might be?

THE COURT: I don't have any problem at all with you discussing it amongst yourselves and coming up with any resolution you can, I would encourage it ....

... If you can take the problem out of my hands, I would be happy to have you do it.

(Id. at 60-61.)

Compromise was not quickly forthcoming, however, and the parties so notified the court. Accordingly, on June 3, 1982, the court ruled orally, in essence, that it would approve the City's proposed interim test if it could be validated.7 It ordered the City (1) to prepare to administer the proposed exam to interested members of the plaintiff class in September, (2) to appear in court in early August for a hearing on the validity of the proposed exam, (3) to disclose to the plaintiff class all information on the validity, fairness, and adverse impact of the exam, and (4) to conduct, beginning on June 7, familiarization and training programs for members of the plaintiff class who wished to take the interim exam.8 Berkman and the City were directed to cooperate in developing the familiarization and training programs. This order was embodied in a written order filed on June 9 ("June 9 Order"). UFA immediately appealed.9

Berkman and the City nevertheless continued to negotiate, and eventually they agreed upon a qualifying physical test that differed only in minor respects from the interim test proposed by the City in May. The agreement was embodied in a stipulation dated July 12, 1982, which described the physical test as set forth in the margin.10

UFA refused to join in the stipulation, however, and the City moved to have the district court compel UFA to sign. At a hearing on August 3, the district court stated that it had no authority to order a party to sign a stipulation but that it would construe the City's application as a request that UFA "show cause why, if there is any reason, why this qualifying test shouldn't be approved." (Aug. 3, 1982 Tr. at 3.) The court then put that question to UFA's counsel:

THE COURT: ...

....

Mr. Teitler[,] [p]erhaps you could tell me what your position is?

MR. TEITLER: My position is very simple, your Honor.

We are in the midst of appealing Court Order dated June 9th, and we feel if we sign any stipulation, it would prejudice our rights under the appeal.

THE COURT: All right. Well, I'll note your opposition, your note that you don't consent to the relief that's being requested and otherwise approve the qualifying test ....

(Id. (emphasis added).) UFA's counsel proffered no other objection and made no other statements.

After further discussion of minor details, the court so-ordered the stipulation signed by Berkman and the City. This was the August Order that is one of the subjects of this appeal.11

After filing its notice of appeal from the August Order, UFA applied to the district court for a stay of the holding of the interim physical test pending the outcome of its appeal. In support of its application, UFA submitted affidavits by its counsel and its president asserting, inter alia, that the interim physical test approved by the court on August 3 was inadequate to establish whether candidates were qualified to be firefighters.12 At a hearing on September 2, the court denied UFA's application, stating as follows:

THE COURT: You are raising questions about [this] qualifying examination, which I asked someone from your office whether they intended to raise back, as I recall, in the beginning of this month [sic: last month?], at a time when I said I wanted to hear those, so that these people who are preparing to take this test would not spend a month o[r] more preparing to take the test which was subject to complaints, and the only answer I got from your representative there was that simply because they felt it would be inconsistent with the position that they were taking on the [appeal] that they were not going to consent to the test, test form. I didn't get a pray [sic] of an explanation of any deficiencies that you found in this test. As a result people have gone out and literally, I assume, sweated through a month in preparation, and I really wonder, that you come in on the eve of the time when these people are going to go out and match their physical capacities with a test of this sort, and start to raise questions about the adequacy of the exam....

....

... I'm going to deny the request for a stay pending the appeal. I find I can only conclude that it's being brought for the purposes of delay.

(Sept. 2, 1982 Tr. at 11, 13 (emphasis added).) A similar motion to this Court was likewise denied.

UFA challenges both the March Order and the August Order on several grounds. It challenges the March Order's requirement that places be reserved for up to 45 women who may be hired as firefighters on the grounds (1) that the order granted "affirmative" relief, which was improper in the absence of a finding of intentional discrimination or an egregious past history of discrimination, (2) that this relief discriminated against men who failed the physical portion of Exam 3040, and (3) that the number 45 was too high. UFA contends that the August Order should not have approved the stipulated interim physical test because that test (1) improperly set a passing grade by reference to the performances of persons too old to qualify as entry-level firefighters, (2) did not provide an adequate test of upper body strength, and (3) lowered the standards, as reflected in Exam 3040, for becoming a New York City firefighter, thereby endangering public safety. We find no merit in any of UFA's contentions.A. The Propriety of Reserving 45 Places for Interim Hiring

Once a violation of Title VII has been established, the district court has broad, although not unlimited, power to fashion the relief it believes appropriate. The bounds of the court's discretion are set by the purposes of Title VII, which are to prevent discrimination and achieve equal employment opportunity in the future, see International Brotherhood of Teamsters v. United States, 431 U.S. 324, 364, 97 S. Ct. 1843, 1869, 52 L. Ed. 2d 396 (1977); Albemarle Paper Co. v. Moody, 422 U.S. 405, 417, 95 S. Ct. 2362, 2371, 45 L. Ed. 2d 280 (1975); Griggs v. Duke Power Co., 401 U.S. 424, 429-30, 91 S. Ct. 849, 852-53, 28 L. Ed. 2d 158 (1971), and to make whole the victims of past discriminatory practices, see, e.g., International Brotherhood of Teamsters v. United States, supra; Franks v. Bowman Transportation Co., 424 U.S. 747, 764, 96 S. Ct. 1251, 1264, 47 L. Ed. 2d 444 (1976); Albemarle Paper Co. v. Moody, supra, 422 U.S. at 418, 95 S. Ct. at 2372. Our inquiry with respect to the March Order is whether, in requiring the City to hire up to 45 members of the plaintiff class who are found to be qualified, the court abused its discretion. Id. at 421-22, 95 S.Ct. at 2373; Association Against Discrimination in Employment v. City of Bridgeport, 647 F.2d 256, 279 (2d Cir.1981), cert. denied, 455 U.S. 988, 102 S. Ct. 1611, 71 L. Ed. 2d 847 (1982) ("ADE v. City of Bridgeport"); cf. Kirkland v. New York State Department of Correctional Services, 520 F.2d 420, 429-30 (2d Cir.1975) (employment discrimination action brought under the Fifth and Fourteenth Amendments to the Constitution and 42 U.S.C. Secs. 1981 and 1983), cert. denied, 429 U.S. 823, 97 S. Ct. 73, 50 L. Ed. 2d 84 (1976). We conclude that it did not.

We begin by considering the propriety of the court's selection of the number 45 for interim hiring, since that question has relevance not only to UFA's contention that that number was improperly inflated by the court's consideration of women who did not take the discriminatory physical test, but also to UFA's contention that this provision improperly granted affirmative relief in the nature of a quota.

We see no impropriety in the court's selection of the number 45. The March Order was linked precisely to the disparate impact that the court had found in the physical portion of Exam 3040. The court found not only that women had failed because the test was discriminatory, but also that women had been deterred from even taking the test because of the advance publicity predicting that no woman could pass. Those who have been deterred by a discriminatory practice from applying for employment are as much victims of discrimination as are actual applicants whom the practice has caused to be rejected. E.g., Dothard v. Rawlinson, 433 U.S. 321, 330, 97 S. Ct. 2720, 2727, 53 L. Ed. 2d 786 (1977); International Brotherhood of Teamsters v. United States, supra, 431 U.S. at 365-66, 97 S. Ct. at 1869-70 ("When a person's desire for a job is not translated into a formal application solely because of his unwillingness to engage in a futile gesture he is as much a victim of discrimination as is he who goes through the motions of submitting an application."); ADE v. City of Bridgeport, supra, 647 F.2d at 281-82 & n. 23. It was thus well within the court's discretion to fashion interim relief that took into account the entire discernible impact of the City's discriminatory test, and not just its most obvious effect, i.e., that all women who actually took the test failed.

It was also reasonable for the court to estimate the number of women deterred from taking the discriminatory physical test by referring to the percentage of eligible men who took the test, rather than seeking an actual deterrence census of the 301 eligible women who elected not to take the physical portion of Exam 3040. Accordingly, we find no abuse of discretion in the district court's selection of the number 45 with respect to the interim hiring requirement imposed on the City.13

Title VII remedies ordinarily may be described as compliance relief, compensatory relief, or affirmative relief. ADE v. City of Bridgeport, supra, 647 F.2d at 278; Guardians Association of the New York City Police Department v. Civil Service Commission, 630 F.2d 79, 108 (2d Cir.1980), cert. denied, 452 U.S. 940, 101 S. Ct. 3083, 69 L. Ed. 2d 954 (1981) ("Guardians IV"). The three types of relief differ to some extent in purpose, effect, and justification.

Compliance relief is designed to erase the discriminatory effect of the challenged practice and to assure compliance with Title VII in the future. Such relief may include prohibiting the use of an invalid test or criterion, restricting appointments from an eligibility list compiled by reference to the results of an invalid test, ordering that new and valid selection procedures be adopted, and authorizing interim hiring that does not have a disparate impact on any group protected by Title VII. An interim hiring order does not have a disparate impact on any such group when (a) it orders the hiring of members of the plaintiff class whom the court has found to be victims of the defendant's discrimination, and (b) it calculates the number of victims to be so hired--in relation to the total number of interim hirees--by reference to the percentage the victims constituted of the total applicant pool. Such an order does not go beyond the simple elimination of the disparate impact of the practice found to be discriminatory and is properly regarded as compliance relief. Compliance relief is appropriate whenever a Title VII violation has been found, irrespective of any history of prior discriminatory practices or the intent of the defendant. See ADE v. City of Bridgeport, supra, 647 F.2d at 278; Guardians IV, supra, 630 F.2d at 108 & n. 25.

Compensatory relief is designed to "make whole" the victims of the defendant's discrimination. To the extent that an order requires the hiring of a member of the plaintiff class--i.e., a victim of the discrimination--it constitutes both compliance relief and compensatory relief. Compensatory relief may also be granted in the form of backpay, payment of the value of past fringe benefits, and retroactive seniority. These forms of relief are generally appropriate under the same circumstances as compliance relief. See ADE v. City of Bridgeport, supra, 647 F.2d at 278-80.

Affirmative relief is that designed principally to remedy the effects of discrimination that may not be cured by the granting of compliance or compensatory relief. It may include the setting of long-term hiring targets or the imposition of a requirement that the defendant actively recruit or train members of the Title VII-protected group. Such relief may be required where, for example, the defendant has intentionally or egregiously engaged in a practice of discrimination that is likely to have discouraged members of the protected group from becoming members of the applicant pool at any stage. Affirmative relief may also include interim hiring relief that is extended to persons other than members of the plaintiff class and in proportions exceeding the ratio of plaintiff class members to the total applicant pool. See id. at 278, 282; Guardians IV, supra, 630 F.2d at 108-09.

Affirmative relief is normally justified only if the defendant's discrimination has been intentional, see, e.g., id. at 109, or there has been a long-continued pattern of egregious discrimination, see id.; ADE v. City of Bridgeport, supra, 647 F.2d at 286; Rios v. Enterprise Association Steamfitters Local 638, 501 F.2d 622, 631 (2d Cir.1974). We have indicated, however, that affirmative relief for Title VII-type violations may be upheld even in the absence of these factors if it is limited to an interim stage and does not require the hiring of persons who are not members of the plaintiff class. Thus, in Kirkland v. New York State Department of Correctional Services, supra, in which the members of the plaintiff class comprised approximately 10% of the applicants who took a discriminatory promotional exam, we reviewed a provision that forbade interim promotions without court approval and required that at least one out of every four persons so promoted be a member of the plaintiff class. Although we struck down other provisions of the court's order that imposed permanent quota restrictions, we upheld this interim provision, stating as follows:

Since this portion of the decree is interim in nature, does not mandate the making of any promotions, does not disregard an existing civil service eligibility list, and since its benefits are limited to the members of the plaintiff class, we affirm it as not being an abuse of the District Court's discretion.

520 F.2d at 429-30; see also Guardians IV, supra, 630 F.2d at 109 (hiring order to assure Title VII compliance on interim basis may properly "select[ ] from among adequately qualified applicants ... according to some appropriately noncompensatory ratio, see, e.g., Kirkland, supra, 520 F.2d at 429-30").

Within this framework we think it plain that in the present case the court's order that up to 45 members of the plaintiff class be hired, if found qualified, constituted compliance relief rather than affirmative relief.14 As discussed in the previous section, the number 45 was determined strictly by the court's estimate of the approximate number of women who had been the victims of the City's discriminatory use of Exam 3040. Since only women who had passed the written portion of Exam 3040 were made eligible to take the interim test, the relief focused solely on the pool of applicants as it existed just prior to the discriminatory phase of Exam 3040.

The March Order neither imposed a quota--i.e., the relatively permanent use of a specified hiring ratio--nor established a hiring goal for the City. The order imposed no numerical long-range hiring requirements whatever; the only long-range relief ordered was the development of properly validated selection procedures that would not have a disparate impact on women. Further, the interim number 45 represented simply the maximum number of the plaintiff class members that the City would be required to hire. If more than 45 class members took and passed the interim test, only 45 were to be hired. Nor was the City ordered to take any affirmative steps to attract women to the Fire Department. If fewer than 45 members of the class remained interested in taking the interim physical test, the City was not required to seek out others. If fewer than 45 passed the interim test, the City was to hire only that lesser number. If no women passed the interim test, the order did not require that any be hired.

Thus, the interim hiring provisions of the March Order simply provided an opportunity for the victims of the discrimination15 to qualify by means of a sex-neutral interim physical test. The district court properly regarded this as compliance relief, 536 F. Supp. at 216-17, which was justified by its finding of a Title VII violation, and we find in it no abuse of the court's discretion.

In enacting Title VII, Congress expressed a preference for achieving compliance by voluntary means. See, e.g., Alexander v. Gardner-Denver Co., 415 U.S. 36, 44, 94 S. Ct. 1011, 1017, 39 L. Ed. 2d 147 (1974); United States v. City of Alexandria, 614 F.2d 1358, 1362 (5th Cir.1980); Patterson v. Newspaper & Mail Deliverers' Union, 514 F.2d 767, 771 (2d Cir.1975), cert. denied, 427 U.S. 911, 96 S. Ct. 3198, 49 L. Ed. 2d 1203 (1976). In general, therefore, when the parties to the litigation have agreed on a means of achieving compliance, the district court should normally approve the settlement unless it contains provisions that are unreasonable, unlawful, or against public policy. On appeal, the district court's approval of a settlement should be upheld unless it constituted an abuse of discretion.

We find no abuse of discretion by the district court in entering the August Order approving the agreement between Berkman and the City to the terms of the interim qualifying test. We note, to begin with, that when the City commenced, following the March Order, to design an interim test, it appears to have done so with a proper concern for public safety. The record indicates that the City sought expert advice throughout, consulted the Fire Department repeatedly, and received and implemented advice from UFA.

Moreover, prior to entry of the August Order, UFA never suggested that either the agreed-upon test or its City-prepared predecessor was in any way inadequate. UFA plainly had no quarrel with the adequacy of the proposed interim test developed by the City prior to May 26, 1982. UFA came forward with no criticism of that test, and indeed, indicated its own approval by suggesting that if Berkman's experts would review and could approve it, a court decision would be unnecessary. There appears to be little difference between that initially proposed test and the one that was eventually agreed on by Berkman and the City, compare note 5 supra with note 10 supra; and when the court asked UFA why the agreed-upon test should not be approved, UFA gave no indication that it thought the test was inadequate. It stated only that it opposed the test because it had an appeal pending from the court's June 9 Order. The court stated that it took this to mean that UFA challenged the nature of relief that had been granted but that it "otherwise approve[d] the qualifying test." (Aug. 3, 1982 Tr. at 3.) UFA gave no indication that this was not an accurate interpretation of its position. Thus, the district court was not presented with any basis for believing that it should not approve the interim test on grounds of inadequacy.

Nor do the particular criticisms of the interim test voiced by UFA on appeal suggest to us that it would have been unreasonable for the district court to approve the test if UFA had timely voiced these criticisms below. We are unpersuaded by the complaint that incumbent firefighters over the age of 29--the maximum age for entrance into the Fire Department--were used to test the test and that their scores were used to derive the passing grade of 4:09 minutes. To be sure, the performance times of the incumbent test-testers indicated that those over the age of 29 generally performed less swiftly than those who were younger. Two countervailing considerations, however, suggest that this pattern may have no substantial significance. First, the trial testimony suggested that fighting real fires requires pacing and caution rather than continuing exertion at maximum physical capacity, 536 F. Supp. at 212 & nn. 23-24, and it may be that the relatively slower times of those older than 29 reflected to some extent their greater experience. Second, as suggested by the district court in expressing its reservations as to whether a test tested by incumbents might not be unfair to the plaintiff class, see note 7 supra, it may be that as a result of the incumbents' firefighting experience, the performance times of all of the test-testers were faster, even without maximum exertion, than would be the times of persons of comparable ages without any experience. In any event, it does not seem to us unreasonable for the City to have sought to determine the average performance of actual job incumbents, and we note that the EEOC Guidelines expressly recommend that a cutoff score "should normally be set so as to be reasonable and consistent with normal expectations of acceptable proficiency within the work force." 29 C.F.R. Sec. 1607.5(H).

We are similarly unmoved by UFA's claim that the interim test does not adequately measure upper body strength. We note, preliminarily, that the evidence suggested, and the court found, that the most important physical attribute for a firefighter is not strength but stamina. 536 F. Supp. at 188, 191 n. 6, 204, 212. Moreover, it is plain that the interim test does require upper body strength. It requires, for example, that candidates raise a 58-pound, 20-foot ladder from the ground to a vertical position, lift and stretch an 80-pound hose, and drag a 145-pound dummy for some distance. The 145-pound dummy is 82 pounds heavier than the 63 pounds that the record indicated was the average weight that firefighters were most often required to drag and, indeed, is 20 pounds heavier than the average extreme weight that firefighters were only occasionally required to drag. See id. at 191. The 80-pound hose is 17 pounds heavier than the weights the record indicated firefighters were most often required to lift. Id. Thus, the interim test required, during a short amount of time (not more than 4:09 minutes), the use of greater upper body strength than is required on the average in fighting real fires.

Finally, we need not dwell on UFA's contention that the interim test is inadequate because it sets lower standards than did Exam 3040. Exam 3040 was found to be discriminatory and not job-related. It cannot rationally be used as a benchmark against which other tests may be judged.

The orders appealed from are affirmed.