Plaintiffs appeal from an order of the district court entering judgment for defendants in this class action lawsuit alleging intentional discrimination under 42 U.S.C. Sec. 1983, and municipal conduct causing a prohibited disparate impact under Title VII of the Civil Rights Act of 1964, 42 U.S.C. Sec. 2000e et seq. The suit, which was brought on behalf of entry-level female firefighters in the City of Cleveland, challenged the rank-order written and physical capabilities selection examination established by the city as perpetuating the exclusion of women from firefighting positions. The district court granted defendants' motion for a directed verdict on the charge of intentional discrimination and, at the close of all the evidence, found for defendants on the Title VII claim.

Citing Spurlock v. United Airlines, Inc., 475 F.2d 216, 219 (10th Cir.1972), the district court concluded that the possibility of hiring unqualified firefighters poses such significant risks to the public that it was appropriate to hold the city to a lighter burden of proving that its employment criteria were job-related than would be the case if the job at issue did not implicate public safety. The district court then concluded that the examination was job-related and validated according to content, construct and criterion-related validation studies and that plaintiffs failed to establish the existence of a less restrictive alternative. Finally, the district court held that there was a sufficient correlation between higher test scores and job performance to warrant a rank-order hiring procedure. Zamlen v. City of Cleveland, 686 F. Supp. 631 (N.D.Ohio 1988).

On appeal, the unsuccessful female applicants challenge the decision of the district court on five bases: (1) the section 1983 claim was improperly dismissed on a motion for a directed verdict because they presented a prima facie case of intentional discrimination; (2) the district court improperly excluded the testimony of women firefighters; (3) the district court improperly allocated the burdens and invoked the Spurlock doctrine; (4) the 1983 entry-level firefighter examination was not properly validated; and (5) plaintiffs demonstrated less restrictive alternatives to the challenged examination. For the reasons that follow, we affirm the district court.

Much has been written about historical discrimination based upon race, sex, and national origin in public service occupations such as law enforcement and firefighting.1 In order to address the problem, many municipalities have employed rank-order written and physical abilities tests which ostensibly allow employers to select candidates who possess the highest degree of skills required to perform the job.2 Although these tests have been designed to eliminate discriminatory hiring practices, some have been shown to have a disparate impact on women and, for that reason, are challenged.

Two claims are usually advanced in support of the argument that rank-order physical tests unfairly discriminate against women. First, it is said that these tests measure attributes in which men traditionally excel, such as speed and strength (anaerobic traits), while ignoring those in which women traditionally are said to excel, such as stamina and endurance (aerobic traits). Second, it is claimed that the tested attributes are not necessarily related to the skills which the specific job requires. The examination at issue was challenged for precisely these reasons.

As of 1977, the city had never hired any woman firefighters. In that year, it hired Dr. Norman Henderson, a tenured professor of psychology at Oberlin College who had significant experience developing tests for various municipalities throughout the country, to design, administer and score an entry-level firefighter examination. Dr. Henderson was again hired to develop and administer an entry-level exam in 1980.

No women were hired after the 1977 and 1980 examinations were scored and the applicants ranked. Out of 911 applicants who took both the written and physical portions of the examination in 1980, 18 were female. Only one female, or 5.6% of the women who took the exam, scored high enough to be placed on the eligibility list. With a ranking of 634, however, she was too far down the list to be hired. In contrast, 787 male applicants, or 88.2%, were placed on the eligibility list.

In 1983, Dr. Henderson was once again hired by the city to design and administer an exam for firefighters. In light of the female applicants' poor performance on the 1977 and 1980 examinations, and in order to minimize any disparate impact which his previous examinations may inadvertently have had on female applicants, Dr. Henderson prepared a new job analysis. The purpose of this new job analysis was to:

(1) establish a list of tasks required of entry-level firefighters;

(2) determine the frequency with which each task is performed and its importance to acceptable job performance;

(3) group tasks into broad job dimensions;

(4) assess the knowledge, skills and abilities required for learning and adequately performing critical and highly important job tasks;

(5) determine overall knowledge, skills and abilities required for entry-level firefighters with respect to the above tasks; and

(6) identify and define abilities and skills to be tested based upon the above data.

He then compiled an initial tasks list based upon a 1974 survey of 271 Cleveland firefighters. In the survey, firefighters rated 95 firefighting tasks in terms of frequency and importance. This initial list was then reviewed in conjunction with the Ohio Trade and Industrial Education Fire Service Training Manual and tasks lists from other cities. Dr. Henderson then produced a revised list consisting of 150 tasks and submitted this list to Chief William E. Lee, Director of the Cleveland Fire Training Academy. Chief Lee pared this list down to a final checklist of 135 tasks. Dr. Henderson also prepared a list of 12 intellectual and perceptual abilities relevant to effective firefighting.

Based upon this initial research, Dr. Henderson developed final written and physical components of the examination. The written component was designed to test reading comprehension, the ability to follow directions, mathematical skills, and other forms of cognitive reasoning. The physical component consisted of three events:

Event 1: Overhead Lift--using a 33 lb. barbell, candidates must lift the barbell overhead repeatedly for one minute or up to a maximum of 35 lifts.

Event 2: Fire Scene Set Up and Tower Climb--while wearing a custom-tailored self-contained breathing apparatus, candidates must drag two lengths of standard 2 1/2" hose 180 feet (90 feet one way, drop coupling, run to the other end of the hose, pick up and return 90 feet, drop coupling in designated area), run 75 feet to pumper, remove a one-person ladder (approximately 35 lbs.) from the side of the pumper, carry the ladder into the fire tower, place it against the back rail of the first landing and continue up the inside stairwell to the fifth floor where a monitor observes the candidates' arrival. Then, candidates return to the first landing, retrieve the ladder and place it on the pumper.

Event 3: Dummy Drag--still wearing their self-contained breathing apparatus, candidates must drag a 100 lb. bag 70 feet (40 of which includes low headroom), turn and, still dragging the bag, return to the starting point.

After the test was developed, but before it was administered, the city embarked on a program to recruit and train female firefighters. As part of its recruitment program, the city provided potential female recruits with a free twelve-week training program. This program, which included a two and one-half hour physical and cognitive portion, began on February 14, 1983. The written portion of the training program was based primarily upon the 1982 edition of the ARCO Civil Service Test training manual. The physical portion, which was based primarily on the content of previous examinations, included training in such activities as dummy lift and carry, dummy drag, hose drag, tower run, fence climb, ladder lift, balance beam walk, and hose coupling. The program did not include training in the use of barbells, nor did the director of the program recommend to any of the women that she work with barbells prior to the examination.

One week before the actual physical examination, the city notified all applicants of the content of the examination, including the barbell event. The training program obtained a set of barbells which was made available to the applicants.

On April 30, 1983, the city administered the written portion of the test and, on May 7 and 13, the physical portion of the test. There were 3,612 applicants initially, but only 2,212 took the written part and 1,233 the physical part. Each portion of the examination was worth a raw score of 50 points with a maxim achievable score of 100. The raw scores on the written portion were adjusted by capping the scores from different sections, by awarding five extra points to qualifying veterans, by awarding ten extra points to city residents, and by adding up to six points to the scores of minority candidates. The minority adjustment was undertaken as a means of complying with a consent decree entered against the city in a suit by minority candidates alleging bias in hiring. Only those applicants with an adjusted score of at least 35 were eligible to take the physical portion of the exam.

Of the 285 females who took the written portion, 122 passed; of the 1,927 males who took the written examination, 1,206 passed. After taking the physical portion of the examination, 29 females scored high enough on both portions of the exam to be placed on the eligibility list while 1,069 males were placed on the list. However, the woman with the highest score still only ranked 334 on the eligibility list--too low to be hired. The class of 35 firefighters, therefore, contained no women.

On June 14, 1983, plaintiffs filed this class action. The principal defense was that the selection procedure was job-related and properly validated. A second, similar suit was initiated by the United States Government, and the two cases were consolidated for trial.

On January 30, 1986, the class was certified and all individual defendants were dropped. The Title VII claims were tried to the court while the section 1983 and pendent state law claims (common law fraud and violation of administrative procedures) were tried to a jury. After the close of all the evidence, the city moved for directed verdicts on all counts. The district court granted the motion as to the section 1983 and state law claims. At the close of all the testimony on the Title VII claim, the district court found for the city.

In order to make out a prima facie claim under 42 U.S.C. Sec. 19833 for intentional discrimination, plaintiffs must demonstrate (1) a deprivation of rights protected under the federal constitution and (2) action under color of state law. Black v. City of Akron, Ohio, 831 F.2d 131 (6th Cir.1987) (citing Day v. Wayne County Board of Auditors, 749 F.2d 1199 (6th Cir.1984)). In order to survive a motion for a directed verdict pursuant to Fed.R.Civ.P. 50(a), plaintiffs must present sufficient evidence such that there is a controverted issue of fact upon which reasonable persons could differ. Hersch v. United States, 719 F.2d 873, 876-77 (6th Cir.1983). After taking the evidence in a light most favorable to the nonmoving party, a motion for directed verdict may only be granted if it is clear that reasonable people could come to but one conclusion from the evidence. Coffy v. Multi-County Narcotics Bureau, 600 F.2d 570, 579 (6th Cir.1979).

Plaintiffs' trial strategy consisted of introducing statistical evidence in the hope of establishing a prima facie case of discriminatory intent which would enable them to survive a motion for a directed verdict on the section 1983 claim. Relying upon the testimony of Martina Dusenberry, a mathematical statistician employed by the Employment Litigation Section, Civil Rights Division, of the United States Department of Justice, plaintiffs asserted that the disparate impact on women as a result of the 1983 exam was so statistically significant that this alone was sufficient to establish a prima facie case of discriminatory intent. At the close of the presentation of their statistical evidence, however, the district court, apparently unconvinced that Dusenberry's testimony alone had established a prima facie case of discriminatory intent, asked plaintiffs if they wished to present additional evidence in support of their prima facie case.

Although plaintiffs contended that they had additional evidence in support of their section 1983 claim, they declined the court's suggestion that it might be prudent to present more evidence at this stage, and elected to wait until the rebuttal phase of their Title VII action to proffer additional expert testimony regarding the construction, administration, and scoring of the test. None of this additional evidence was before the district court when it ruled on the motion for a directed verdict. Accordingly, the issue before us is whether the statistical evidence presented prior to defendants' motion for a directed verdict was sufficient to establish a prima facie case of intentional discrimination.

Both parties rely upon Black v. City of Akron, Ohio, 831 F.2d 131 (6th Cir.1987), in support of their respective positions. In Black, this court held that "statistical proof may be used in actions under 42 U.S.C. Sec. 1983" and that "allegations of statistical evidence of an adverse impact might be sufficient to survive summary judgment." Id. at 133 (emphasis added). Black does not say, as the city contends, that statistical evidence alone is insufficient to sustain an action under section 1983. Nor does it say, as plaintiffs contend, that statistical evidence is by itself always sufficient to create a prima facie case of intentional discrimination. Black stands only for the proposition that statistical evidence may be introduced to show a discriminatory purpose. The sufficiency of the statistical evidence proffered, however, still needs to be weighed by the district court in order to determine if it supports an inference of intentional discrimination.

Plaintiffs' statistical evidence, while not insignificant, simply was insufficient to demonstrate intentional discrimination. Having concluded that the evidence presented was, at best, amorphous, the district court was correct in granting defendants' motion for a directed verdict on the claim of intentional discrimination.

III. EXCLUSION OF WOMEN FIREFIGHTERS' TESTIMONY

Plaintiffs' second assignment of error is that the district court improperly excluded the testimony of women firefighters who had been hired pursuant to a 1985 court-approved agreement between the city and plaintiffs. Although the district court, in granting the city's motion in limine to exclude the testimony, did not explain the basis for this decision, the court did incorporate by reference the arguments the city made in its supporting memorandum.

The city argued, essentially, that the testimony would be irrelevant and, even if relevant, more prejudicial than probative. If the jury discovered that the city had already hired women firefighters outside its normal selection procedure, the city argued, "they would undoubtedly infer that the City or this Court believed that the City's selection method was faulty."

Plaintiffs' proffer suggests that the testimony would have been relevant to plaintiffs' case. According to the proffer, the female firefighters would have testified about:

1. Their experiences in the Cleveland Firefighters Training Academy, including information about the type of training they received and the type of exercises that were stressed.

2. Their relevant employment history.

3. Their conclusions that the events on the 1983 physical performance examination were not representative of their actual duties.

4. Their assignments since graduating from the firefighting academy.

Testimony of that nature could be relevant to a central issue in this case--the validity of the selection device. While the judge may have been motivated by a desire not to confuse or prejudice the jury, the decision to exclude all the proffered testimony was improper. The city was mainly concerned that the jury would conclude that its agreement in 1985 to hire female firefighters was tantamount to a concession of liability in this case. However, the female firefighters could have testified without alerting the jury to the manner in which they were hired.

In reviewing a trial court's decision to admit or exclude evidence on the basis of relevancy, unfair prejudice, or confusion of the issues, two factors must be considered. First, a reviewing court will not reverse such a decision on appeal absent a showing of a clear abuse of discretion. Conklin v. Lovely, 834 F.2d 543, 551 (6th Cir.1987). Second, even if the lower court's decision amounts to an abuse of discretion, it will not be disturbed on appeal if it did not result in a substantial injustice, as "no error in the admission or exclusion of evidence is ground for reversal or granting a new trial unless refusal to take such action appears to the court to be inconsistent with substantial justice." McGowan v. Cooper Industries, Inc., 863 F.2d 1266, 1271 (6th Cir.1988) (citing TCP Industries, Inc. v. Uniroyal, Inc., 661 F.2d 542, 550 (6th Cir.1981)).

While the district court perhaps was overzealous in its attempt to protect the jury from unfair prejudice, we are unable to conclude that its decision prejudiced plaintiffs' cause, resulting in substantial injustice. The evidence which plaintiffs wished to present through the testimony of the female firefighters was substantially the same as that of the male firefighters who were allowed to testify on their behalf. Similarly, they did present their own expert testimony and had ample opportunity to cross-examine the city's witnesses. While we view the decision of the district court as unfortunate, under these circumstances we are unable to conclude that the exclusion of the female firefighters' testimony was prejudicial to plaintiffs' cause.

The purpose of Title VII4 is to achieve equality of employment opportunities through the eradication of employment barriers which discriminate on the basis of race, gender, religion, and other protected classifications. Griggs v. Duke Power Co., 401 U.S. 424, 429-31, 91 S. Ct. 849, 852-53, 28 L. Ed. 2d 158 (1971). The Supreme Court has recognized that Title VII forbids not only overt discrimination "but also practices that are fair in form, but discriminatory in operation." Griggs, 401 U.S. at 431, 91 S.Ct. at 853. Employment practices which, although facially neutral, are discriminatory in practice have given rise to the theory of liability commonly known as the disparate impact theory. Under this basis for liability, a specific employment practice may be deemed violative of Title VII without proving the employer's subjective intent to discriminate. Wards Cove Packing v. Atonio, --- U.S. ----, 109 S. Ct. 2115, 2119, 104 L. Ed. 2d 733, 744 (1989). Here, plaintiffs proceeded under the disparate impact theory.

In Albemarle Paper Co. v. Moody, 422 U.S. 405, 95 S. Ct. 2362, 45 L. Ed. 2d 280 (1975), the Supreme Court outlined the burdens of the parties in disparate impact cases:

Title VII forbids the use of employment tests that are discriminatory in effect unless the employer meets "the burden of showing that any given requirement [has] ... a manifest relationship to the employment in question." [Griggs v. Duke Power Co., 401 U.S. at 432, 91 S.Ct. at 854.] This burden arises, of course, only after the complaining party or class has made out a prima facie case of discrimination, i.e., has shown in effect that the tests in question select applicants for hire or promotion in a racial pattern significantly different from that of the pool of applicants. See McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802 [93 S. Ct. 1817, 1824, 36 L. Ed. 2d 668] (1973). If an employer does then meet the burden of proving that its tests are "job related," it remains open to the complaining party to show that other tests or selection devices, without a similarly undesirable racial effect, would also serve the employer's legitimate interest in "efficient and trustworthy workmanship." Id. at 801 [93 S.Ct. at 1823].

Albemarle, 422 U.S. at 425, 95 S.Ct. at 2375.

In Wards Cove, the Supreme Court clarified prior case law by pointing out that once a plaintiff has made out a prima facie case,5 it is not the burden of persuasion which shifts to the employer; instead, at that point, the employer shoulders the burden of production, of producing evidence of a business justification: "The ultimate burden of proving that discrimination against a protected group has been caused by a specific employment practice remains with the plaintiff at all times." Wards Cove, --- U.S. ----, 109 S. Ct. at 2126, 104 L.Ed.2d at 753 (citing Watson v. Fort Worth Bank & Trust Co., 487 U.S. 977, 108 S. Ct. 2777, 101 L. Ed. 2d 827 (1988)).

Prior to Wards Cove, this circuit adopted the doctrine announced in Spurlock v. United Airlines, Inc., which allows an employer to meet a lighter standard of proof of job-relatedness where the job at issue clearly requires a high degree of skill and the economic and human risks involved in hiring an unqualified applicant are great. Chrisner v. Complete Auto Transit, Inc., 645 F.2d 1251 (6th Cir.1981).

The district court specifically held that firefighting poses sufficient risks to the public's safety to invoke the Spurlock doctrine: "The risks to the public in terms of life, limb and property as a consequence of hiring an unqualified firefighter are great, resulting in a less heavy burden for demonstrating the job relatedness of the employment criteria." Zamlen, 686 F. Supp. at 654. On appeal, plaintiffs contend that, since the Spurlock doctrine was intended to apply to jobs which require special skills or abilities "which are extremely difficult to define with significant precision and even harder--to the extent so identified--to test for or measure," Davis v. City of Dallas, 777 F.2d 205, 215 (5th Cir.1985), cert. denied, 476 U.S. 1116, 106 S. Ct. 1972, 90 L. Ed. 2d 656 (1986), and since the 1983 examination at issue was designed to measure tangible, objective cognitive abilities and concrete physical capabilities, the district court improperly invoked this doctrine.

Actually, we need not decide whether the Spurlock doctrine was properly invoked since, in view of the Supreme Court's intervening opinion in Wards Cove, whether the district court lowered the city's burden of persuasion is no longer at issue. The city had only the burden of producing evidence to justify the use of its selection device. Since the city was found to have satisfied the Spurlock burden of persuasion, and that burden is still more onerous than the burden of production which the city actually is obligated to bear under Wards Cove, the city, having met the heavier burden, necessarily satisfied the lighter burden. While one may produce without persuading, one cannot persuade without first producing. Accordingly, plaintiffs were not harmed by the district court's allocation of burdens.

V. ADEQUACY OF VALIDATION PROCEDURES AND FAILURE TO MEASURE

The standard of review in Title VII cases is strict. As this court recently recognized,

[w]hen reviewing the district court's factual findings in a Title VII discrimination case, we may not reverse unless we find that the district court committed clear error. If the district court's account of the evidence is plausible ... the court of appeals may not reverse it even though ... it would have weighed the evidence differently.

Wrenn v. Gould, 808 F.2d 493, 499 (6th Cir.1987) (citing Anderson v. Bessemer City, 470 U.S. 564, 573, 105 S. Ct. 1504, 1511, 84 L. Ed. 2d 518 (1985)).

The Equal Employment Opportunity Commission ("EEOC") has developed guidelines "to assist employers ... to comply with requirements of Federal law prohibiting employment practices which discriminate on grounds of race, color, religion, sex and national origin ... [and] to provide a framework for determining the proper use of tests and other selection procedures." 29 C.F.R. Sec. 1607.1(B). Under the guidelines, employers may use three types of studies to validate an employee selection procedure: content, construct, or criterion-related validity studies. The particular device that is chosen depends upon the nature of the job, the way in which the test will be interpreted, and the type of data that is available.

A content validity study is appropriate when test items directly measure abilities that are prerequisites to entry-level job performance (for example, a shorthand test for a secretarial position). Content validity studies must include a thorough job analysis identifying the most important knowledge, skills, and abilities necessary to successful job performance. Guardians Association v. Civil Service Comm'n of New York, 633 F.2d 232, 242 (2d Cir.1980). An employer who seeks to validate a physical examination on the basis of a content validity study must therefore focus on the job's actual physical requirements. In addition, an employer must avoid testing for those skills that can readily be learned on the job.6

A construct validity study tests for abstract qualities that are difficult to test but that are, nonetheless, important characteristics for proper job performance. Such studies are appropriate where the necessary qualities, such as creativity, cannot be measured directly. Construct validity studies are usually not appropriate for validating physical exams because these qualities are readily observable and quantifiable.

Finally, a criterion-related study, which also makes use of empirical data, indirectly tests for those skills necessary for successful job performance. According to the guidelines, a criterion-related study "should consist of empirical data demonstrating that the selection procedure is predictive of or significantly correlated with important elements of job performance." 29 C.F.R. Sec. 1607.5(B).

The district court found that the 1983 examination was properly validated according to content, construct, and criterion validation principles. Nonetheless, plaintiffs attack the district court's finding on the ground that the test that Dr. Henderson developed was not properly validated and, in addition, did not test for all of the attributes he identified in his job analysis as important to an effective firefighter. Although Dr. Henderson recognized that successful firefighters must possess a high level of aerobic capacity and aerobic fitness, as well as muscle strength, muscular endurance, flexibility, coordination, muscle balance, and speed, the physical portion of the 1983 examination measured anaerobic--maximal speed and strength--but not aerobic--stamina or paced performance--traits.

We first note that there was substantial evidence before the district court supporting the conclusion that the examination was properly validated. Each event in the physical examination was designed to test a representative firefighting task. The barbell lift was designed to simulate the use of a pike pole to tear out ceilings. The fire scene set-up and tower climb event was intended to duplicate critical firefighting tasks performed where speed is the most critical factor, such as setting up ladders and climbing stairs. The dummy drag simulated the rescue of a disabled person under circumstances where heat and smoke make it difficult to stand upright. Although plaintiffs attack the content validity of these events as only superficially replicating an actual sequence of job tasks, an expert testified that these events did, in fact, simulate actual firefighting tasks. Plaintiffs may find fault with the way in which the district court came to certain conclusions but there is an insufficient basis for us to conclude that these conclusions are clearly erroneous.

Similarly, plaintiffs argue that there was insufficient evidence supporting the district court's conclusion that the examination was construct and criterion valid. However, plaintiffs ignore the substantial evidence before the district court, including Dr. Henderson's technical report correlating higher test scores with higher supervisor ratings as well as Dr. Henderson's further analyses, supporting the conclusion that the examination was construct and criterion valid.

Plaintiffs' most forceful argument is that the 1983 examination failed to measure attributes which are concededly important to effective firefighting, attributes in which it is often argued that women traditionally excel, such as stamina and endurance. By failing to test for these aerobic qualities, plaintiffs argue, Dr. Henderson devised an examination hopelessly biased in favor of male applicants who will almost always score higher than female applicants on tests which solely measure anaerobic qualities such as strength and speed.

A similar case was decided recently in the Second Circuit. In Berkman v. City of New York, 812 F.2d 52 (2d Cir.1987) ("Berkman IV "), plaintiffs argued that a firefighter exam used by the City of New York to select entry-level firefighters should be invalidated because it failed to measure aerobic capacity and placed undue emphasis on anaerobic performance. Recognizing that aerobic attributes are an important component of effective firefighting, the court nonetheless held that the city's failure to include events that test for such qualities did not invalidate the examination. While the court implied that an examination that included events which tested an applicant's aerobic energy system would be preferable, "[i]t does not follow, however, that a physical test of the ability to perform simulated job tasks of firefighters, without a specific measurement of stamina, lacks validity to a degree that renders it vulnerable to a Title VII challenge." Berkman, 812 F.2d at 59.

While aerobic ability enables firefighters to sustain a consistent level of energy over a long period of time, speed and strength are critical at the initial stages of a fire where matters of life and death are most acute. A firefighter who tires may be replaced by a fresh recruit but, as the Berkman court recognized, "if the first firefighters on the scene are deficient in the speed and strength necessary to handle their tasks, those in need of immediate rescue will not be comforted by the fact that those first on the scene might be able to sustain their modest energy levels for a prolonged period of time." Berkman, 812 F.2d at 60. Here, the district court concluded that anaerobic qualities are more important. Certainly, we are unable to say a fire department is not entitled to select firefighters whose abilities enable them to act more effectively in the first moments of a fire. Accordingly, although a simulated firefighting examination that does not test for stamina in addition to anaerobic capacity may be a less effective barometer of firefighting abilities than one that does include an aerobic component, the deficiencies of this examination are not of the magnitude to render it defective, and vulnerable to a Title VII challenge. Berkman, 812 F.2d at 59.

Finally, plaintiffs contend that a different scoring system--one which would eliminate the addition of variable numbers of minority points, the use of the capping system and the addition of veterans' points--would have raised the rank-order of women on the eligiblity list and, thus, constitutes a less restrictive alternative. Under Wards Cove, once the plaintiffs have shown that other nondiscriminatory means exist which could effectively serve the employer's legitimate interests, the employer must adopt them or risk the inference that the preferred test is a pretext for discriminatory conduct. Wards Cove, --- U.S. at ----, 109 S. Ct. at 2126-27, 104 L.Ed.2d at 753-54.

The district court concluded that plaintiffs failed to demonstrate a less restrictive alternative, a conclusion with which we agree. Although the use of a different scoring system might raise the rank-order of women on the eligibility list, given the fact that the woman with the highest test score still only ranked 334 on the eligibility list, and that the city only hired approximately forty firefighters each year, it is doubtful that any alternative scoring system would have had less of a disparate impact on women. The evidence suggests that, at best, an alternative scoring system would result in female applicants ranking higher on the eligibility list, but still too low to actually be hired. Since rescoring the examination is unlikely to result in higher numbers of successful female applicants, it is an insufficient reason to invalidate an otherwise lawful examination.

Because the examination did parallel the actual tasks which firefighters perform on the job, and the city did demonstrate a direct correlation between higher test scores and better job performance, the examination withstands plaintiffs' challenge. Accordingly, the judgment of the district court is affirmed.