Plaintiffs are forty-four African-American or Latino present or former sergeants of the Chicago Police Department. Plaintiffs failed to be promoted to lieutenant after taking the 1994 police lieutenant examination. Seven hundred sixty-five police sergeants took the examination of which 184 (24%) were African-American and 55 (7%) were Hispanic. The Police Department made 108 rank- order promotions based on the 1994 examination, granting promotions to those officers who obtained the highest 108 scores on the examination. Of the 108 officers promoted, five were African-American and one was Hispanic. Minority promotions, therefore, represented slightly less than 6% of the total number of promotions granted. It is undisputed that the 1994 examination had a disparate impact on minority candidates, and the parties have stipulated that this statistical evidence constitutes a prima facie case of discrimination.

In 1995, the plaintiffs filed a complaint against the City of Chicago (the "City"), alleging that the City deprived them of equal employment opportunities in violation of Title VII of the Civil Rights Act of 1964, as amended 42 U.S.C. sec. 2000e, et seq. No claim was made by plaintiffs, however, that the City intentionally discriminated against them because of their minority status. Plaintiffs sought to preliminarily enjoin the City from making any rank-order promotions based on the 1994 lieutenant examination. The preliminary injunction was denied by the district court on the basis that plaintiffs had failed to establish either irreparable harm or the lack of an adequate remedy at law. The district court also found the balance of harms weighed against the granting of the preliminary injunction.

In a Title VII disparate impact case, the plaintiff bears the initial burden of establishing a prima facie case by showing that the promotional method in question had an adverse impact on minorities. If the plaintiff makes this required initial showing, the burden then shifts to the employer who must prove that the evaluation method is valid by showing that it is "job related" and "consistent with business necessity." 42 U.S.C. sec. 2000e-2k(1)(A)(i). The evaluation method may be shown to be job related under any one of three tests: criterion related, content validity, or construct validity. Uniform Guidelines on Employee Selections Procedures, 29 C.F.R. sec. 1607.5B. If the employer succeeds in validating the evaluation method, the burden shifts back to the plaintiff to prove that there was another available method of evaluation which was equally valid and less discriminatory that the employer refused to use. 42 U.S.C. sec. 2000e-2k(1)(A)(ii); see also Albemarle Paper Co. v. Moody, 422 U.S. 405, 425 (1975).

This case was tried in a bench trial in March 1997. As previously mentioned, the City conceded an adverse impact on minority candidates because of their poor showing on the examination. In a thorough Memorandum Opinion and Order issued on June 30, 1998, the district court found that the 1994 examination was job related under the content validity approach, but agreed with the plaintiffs that the City had a less discriminatory but equally valid method of promotions available which it did not use, namely a combination of rank-order promotions with what are referred to as "merit promotions," a process which we shall examine in more detail shortly. When the court considered what relief was available under those circumstances, it found the record insufficient and set an additional hearing. That hearing resulted in a second Memorandum Opinion and Order issued on September 16, 1998. The court held that, in view of the equally valid but less discriminatory method which the City had not used, the court had the opportunity to broadly exercise its discretion in granting relief as outlined in 42 U.S.C. sec. 2000e-5(g). The court, therefore, awarded some relief to sergeants other than the named plaintiffs. The court found the City's failure directly impacted thirteen sergeants, minority and non-minority, who had taken the 1994 examination. These thirteen officers had been chosen for promotion to lieutenant under the merit promotion plan, but had not been promoted.1 Only one of these officers, Sergeant Raymond, is a plaintiff in the present case.

The court ordered the City to promote the thirteen directly injured sergeants to lieutenants and to award them differential back pay and other benefits from April 1995, the time when they had been passed over for promotion. Exceptions were made in individual cases where the candidate for some reason was no longer eligible for promotion.2 The district judge declined to try to identify and promote any additional sergeants, stating that to do so "would be conjectural and overly subjective," as well as disruptive, and instead ordered the City to pay the plaintiffs who were not promoted a minimal award reflecting their lost opportunity to be selected for a merit promotion. The district court added that, except for certain disputed issues, the City was doing all it could "to promote racial and ethnic diversity among the ranks of its lieutenants."

On appeal, plaintiffs contend that the City did not meet its burden of proving that the examination was content valid. If the examination is determined to be job related, however, then the plaintiffs argue that, based on its finding that the combination of rank-order and merit promotions represented an equally valid, less discriminatory alternative promotional method, the district court erred in failing to order additional merit promotions.

Initially recognizing the difficulties of developing and administering an acceptable process for the hiring and promotion of police officers in a large metropolitan area, the Mayor of Chicago appointed a "Blue Ribbon Committee" to submit recommendations about how to proceed. The Vice Chairman of that committee, James Holzhauer, for instance, called by plaintiffs at trial, testified that he was a partner in the firm of Mayer, Brown and Platt and specialized in handling employment law matters. He was also a part-time faculty member teaching labor and discrimination law at the University of Chicago Law School. Further, Holzhauer had represented the Fraternal Order of Police and other police unions in the area of discrimination. Earlier in his career he had been a city manager and for a short period of time also had been the civilian police commissioner for an upstate New York police department. He explained how the Mayor's Committee had functioned and that the members were an independent task force not paid by the City. He was obviously qualified for this committee assignment.

One of the recommendations of the Mayor's Committee was that outside consultants be retained to develop and to administer promotion examinations. Following that recommendation the City retained Barrett & Associates, Inc., of Akron, Ohio, described as a "Human Resource Consulting Firm" specializing in employee matters including promotion testing. The firm is headed by Dr. Gerald B. Barrett who holds a Ph.D. in psychology, as well as a law degree. He teaches testing and measurement, personnel selection, performance evaluation, and personnel psychology as well as law at the University of Akron. Dr. Barrett developed and administered the lieutenant examination challenged in this suit. Dr. Barrett and his firm had previously developed more than fifty examinations for police and fire departments including examinations for the cities of Cleveland and Akron, Ohio. Some of Dr. Barrett's work has been unsuccessfully challenged in court, including federal court. Along with Dr. Barrett and his firm, the City retained the Arthur Andersen company to aid in the grading of the examination.

Dr. Barrett was no stranger to the Chicago Police Department. In 1993, Dr. Barrett had developed an examination for promotion to Chicago Police Sergeant. In preparing that examination, Dr. Barrett conducted a job analysis of the sergeant position by interviewing approximately ninety Chicago sergeants along with twenty-eight lieutenants about their duties and responsibilities. For the lieutenant examination now in question, Dr. Barrett, following that same course, interviewed additional lieutenants, captains, and sergeants, including minorities. Dr. Barrett also toured the police districts, rode along with lieutenants on duty, observed the work of lieutenants, and reviewed applicable police documents, reports, and orders. Based on the data he gathered from his preliminary work, Dr. Barrett prepared a "Master Job Description" for the Chicago Police Lieutenant position. The Master Job Description identified what are referred to as "major work behaviors," including the associated tasks and responsibilities of lieutenants. In creating the Master Job Description, Dr. Barrett measured the importance and frequency of a lieutenant's tasks and responsibilities. Dr. Barrett also consulted certain source materials which contained information with which a lieutenant was expected to be familiar. These materials included Police Department policies and directives, certain sections of the Illinois Statutes and the Chicago Municipal Code, the collective bargaining agreement of the Union, and the Department's community policing strategy. A list of those source materials was made available to the candidates prior to the examination.

The examination developed by Dr. Barrett had three components. The first was a written job- knowledge test consisting of 150 multiple choice questions derived from the source materials. It was first pilot tested and then further refined before being given. The next component of the examination was referred to as the "In-Basket Exercise." This exercise was based on a hypothetical situation which a candidate might face in an emergency where he or she would have to assume the duties of a lieutenant who had become ill or incapacitated. In this exercise, the lieutenant candidates were each presented with a packet of information simulating a lieutenant's in-basket. The candidates were allowed two and one-half hours to study the materials before being given ninety minutes to answer sixty multiple choice questions. The answers to these sixty questions were contained in the materials provided to the candidates, and the candidates were allowed to refer to these materials during the examination. The in-basket materials had first been reviewed by Chicago Police Department subject matter experts, and the exercise was also pilot tested prior to implementation. The purpose of the in-basket exercise was to measure necessary skills and abilities of possible lieutenants, not to test job-knowledge as was intended by the first component. Those responsibilities, for instance, required knowledge of reports, personnel actions, and the assignment of tasks.

The third component of the examination was an oral briefing exercise intended to demonstrate a candidate's analytical abilities and oral communication skills. This exercise simulated a Chicago Police Lieutenant's responsibilities at roll call. Each candidate was given materials about Chicago gang activity and related Department directives. The candidates were allowed twenty-five minutes to review those materials and then they were required to give an oral briefing on the issue not to exceed ten minutes. During their oral presentations, candidates were allowed to refer to the materials and their own study notes. The oral presentations were recorded on audio tape for later review. A monitor sat in the room during the oral presentations, but was not permitted to communicate with the candidate. The oral presentation reviewers evaluated the recorded presentations without knowing the identity of the candidates. Three trained raters independently scored each presentation on an objective check list and then reached a conclusion about each of the candidates. This component had also been reviewed by Chicago Police Department subject matter experts and pilot tested.

Plaintiffs raise a number of arguments on appeal, but the core argument is that the district court committed reversible error when it accepted Dr. Barrett's testimony that the examination was content valid as sufficient evidence to rebut the plaintiffs' prima facie showing of disparate impact. Plaintiffs argue first that Dr. Barrett's testimony is inadmissible under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), and, secondly, that, even if admissible, Dr. Barrett's testimony is insufficient to justify the examination on which minorities fared so poorly. We address each of these arguments in turn.

Citing Daubert, plaintiffs characterize Dr. Barrett's testimony that the examination was content valid and that the final test scores could be used for rank-order promotions as nothing more than inadmissible conjecture, arguing that the testimony lacks "scientific validity." Daubert involved expert testimony in a personal injury suit involving a prescription drug taken by an expectant mother. The drug was alleged to have caused birth defects in plaintiff's children. The district court granted summary judgment in behalf of the drug company finding petitioner's scientific evidence and the principle upon which it was based were not "sufficiently established to have general acceptance in the field to which it belongs." Daubert, 509 U.S. at 583. Likewise, plaintiffs in the present case claim that Dr. Barrett's opinions are unsubstantiated and lack "scientific validity." His opinions, it is argued, were nevertheless admitted by the district court because of Dr. Barrett's "expertise." Appellants contend that the district court's decision was erroneous.

Under Daubert, the testimony of a scientific expert is admissible only if it is both relevant and reliable. Kumho Tire Co., Ltd. v. Carmichael, 119 S. Ct. 1167, 1171 (1999). In the present case, appellants challenge only the reliability of the admitted expert testimony. A district court enjoys broad latitude both in deciding how to determine reliability and in making the ultimate reliability determination. Id. It is clear from the record that the district court recognized the applicability of Daubert to Dr. Barrett's testimony. Furthermore, while appellants broadly assert that the district judge failed to consider Daubert in making his admissibility determination, their argument actually focuses on what they perceive to be the district court's improper application of the Daubert framework. Appellants contend that Dr. Barrett's testimony fails to meet the reliability prong of Daubert because there was no showing that his opinions were scientifically valid.

We review the district court's reliability determination for abuse of discretion, Kumho Tire Co., Ltd., 119 S.Ct at 1171, and affirm. The Daubert inquiry is "a flexible one" and is not designed to serve as a "definitive checklist or test," Daubert, 509 U.S. at 593-94, but rather to ensure "that an expert, whether basing testimony upon professional studies or personal experience, employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in the relevant field." Kumho Tire Co., Ltd., 119 S. Ct. at 1176. In the present case, it is clear that Dr. Barrett's testimony had "'a reliable basis in the knowledge and experience of [the relevant] discipline.'" Id. at 1175 (quoting Daubert, 509 U.S. at 592). Dr. Barrett has extensive academic and practical experience in designing employment evaluations. Furthermore, it is not accurate to claim that the district judge declined to conduct an inquiry into the scientific validity of Dr. Barrett's opinion. As the district court noted, Dr. Barrett based his opinions, at least in part, on the job analysis that Barrett & Associates meticulously formulated which detailed a relationship between the skills measured in the examination and an individual's effectiveness as a lieutenant. Furthermore, while plaintiffs contend that the "general scientific literature" in the area consists of a single unpublished study, it is undisputed that Dr. Barrett himself has authored approximately fifty articles dealing with employee selection and promotion testing for peer-reviewed journals. This is not a case in which the expert failed to conduct any studies or analysis to substantiate his opinion. See Deimer v. Cincinnati Sub-Zero Products, Inc., 58 F.3d 341, 344 (7th Cir. 1995). Given these facts, it is clear that the district judge's decision to admit Dr. Barrett's testimony was not manifestly erroneous. See id.

Plaintiffs argue in the alternative that Dr. Barrett's testimony, if properly admitted, was insufficient to support a finding that the examination was job related. They contend that the 1994 test was not job related because it did not approximate the work situation. Plaintiffs cite Griggs v. Duke Power Co., 401 U.S. 424 (1971), in which the Court held that, under Title VII, employment tests are forbidden which produce a disparate impact "unless they are demonstrably a reasonable measure of job performance." Id. at 436. Additionally, plaintiffs note that, in 1972, Congress realized that equal employment had been thwarted when employment was based on "criteria unrelated to job performance and on discriminatory supervisory ratings," citing Connecticut v. Teal, 457 U.S. 440, 449 n.10 (1982). From these two cases it can be seen that, if a facially neutral employment practice has a significant discriminatory impact, the employer bears the burden of demonstrating that any requirement of employment imposed has a manifest relationship to the particular employment. Griggs makes clear, however, that even employment tests with a disparate impact are acceptable if "they are demonstrably a reasonable measure of job performance." Griggs, 401 U.S. at 436. It would be unrealistic to require more than a reasonable measure of job performance. It therefore is a matter of reasonableness, except in cases in which the plaintiff can show that the employer was using the practice as a mere pretext for discrimination. There is no claim, however, of employer pretext in the present case.

As previously noted, an evaluation method may be shown to be job related under any one of three tests: criterion related, content validity, or construct validity. Gillespie v. State of Wisconsin, 771 F.2d 1035, 1040 (7th Cir. 1985). In the present case, the district court found that the test was content valid. In evaluating content validity, a court must consider

(1) the degree to which the nature of the examination procedure approximates the job conditions; (2) whether the test measures abstract or concrete qualitites; and (3) the combination of these factors, i.e. [sic] whether the test attempts to measure an abstract trait with a test that fails to closely approximate the working situation. Id. at 1043.

In the present case, the district court recognized the correct standard for determining content validity. After considering all of the evidence, the court determined that the 1994 examination measured a significant portion of the knowledge, skills, and abilities necessary for a police lieutenant and, therefore, was content valid. Because this is a factual finding, we will affirm the decision of the district court unless it is clearly erroneous. Gillespie, 771 F.2d at 1042.

The factual details we have already related about the development of the lieutenant test are enough to refute plaintiffs' arguments. The record shows not only the knowledge, expertise, and experience of those involved in the test development, but also the preliminary use of peer review and pilot testing of each of its three parts. Officers of various police ranks and experience, including minorities, were consulted during the development of the process. It would be totally unjustified to fail to take note of those preliminary efforts and, considering our standard of review, to reverse the trial judge who had carefully considered and weighed all the evidence before coming to the conclusion of test validity. The standard to be applied is not simply whether minorities do well or not on a test. That is only the beginning. It is obviously not impossible to develop a useful and nondiscriminatory test.

Plaintiffs also raise an issue about the City's use of the examination scores to make promotions in rank order, citing Gillespie, 771 F.2d 1035, for the proposition that the use of rank-ordering must be independently justified when the scoring system results in a disparate impact. In Gillespie, this court relied on the Second Circuit's decision in Guardians Ass'n of New York City v. Civil Service Commission, 630 F.2d 79 (2d Cir. 1980). The Guardians court recognized that an employer who wants to use rank-order scores for hiring decisions must demonstrate that rank- ordering is sufficiently justified. Id. at 103. That "task is by no means impossible," even without resort to a criterion related study of the issue. Id. Under Guardians, rank-order promotions can be validated by a substantial showing that (1) the test is job related and representative and (2) the test maker achieved "an adequate degree of reliability."

Id. at 104.

As previously discussed, the 1994 lieutenant examination was based on a detailed job analysis and was constructed in adherence to the Uniform Guidelines. See Guardians Ass'n of New York City, 630 F.2d at 104. The City has made a substantial showing of job relatedness sufficient to satisfy the first prong of the Guardians test. The reliability prong is also met. Barrett & Associates used a number of methods, including pre-testing, to ensure the reliability of the 1994 examination. See id. Furthermore, we agree with the Second Circuit's holding that when an examination measures ability with sufficient differentiating power to justify rank-ordering, it is permissible for the City to set a cut-off score at the point where the rank-ordering provides the number of promotions necessary to fill the City's available openings. See id. at 105. In the present case, the City's use of rank- ordering is valid, and the City was justified in setting a cut-off score which resulted in the necessary number of promotions.

The majority of plaintiff's remaining objections are de minimus and require no analysis. We affirm the trial judge's admission of Dr. Barrett's testimony and the court's conclusion about the validity of the tests, as well as the court's exercise of its discretion in appointing some additional sergeants to lieutenants and no more. See EEOC v. Laborers' Int'l Union, 49 F.3d 304, 307 (7th Cir. 1995) ("We review the decision to grant an equitable remedy under an abuse of discretion standard."). There is one remaining issue.

On the merits, the district court held that the City's promotion test and scoring system were valid under Title VII. However, the court further held that plaintiffs had shown the existence of an equally valid, less discriminatory alternative to rank-order promotions, the merit promotion method. The City likewise believed the use of this method was an equally valid, less discriminatory method, but argued that it was unavailable because the state court had enjoined its use. The district court found, however, that federal law prevailed over the state court action, relying on 42 U.S.C. sec. 2000e-7. The City does not appeal this finding, so we need not pass on the validity of the merit promotion method, but we note that the parties agree as to the value of the merit method despite the fact that it includes a subjective element which minorities often find objectionable.

Following the district court's decision on the merits, plaintiffs applied for attorney's fees and costs under 42 U.S.C. sec. 2000e-5(k). The district court denied plaintiffs' request in part on the basis that the issues on which plaintiffs did not prevail, the validity of the 1994 test and its scoring system, were unrelated to the issue regarding the availability of an alternative promotion method on which plaintiffs did prevail. The district court held that the two claims were distinct and independent and could have been pursued separately. The district court concluded that more than 90% of the time expended related exclusively to the issues of test validity on which the plaintiffs did not prevail. Nevertheless, the court compensated plaintiffs for 20% of their time spent for the entire litigation, describing that as "more than three times the percentage of the total time and expenses devoted to the issue on which plaintiffs prevailed."

The court adopted the City's suggested fee award of $134,699.88 for the litigation of the merits. The court then added $28,284.00 for the time expended in the remedy phase and $2,629.50 for time expended reviewing materials in a companion case. The court also awarded plaintiffs $10,915.00, the total amount of fees sought in connection with the presentation of the fee petition. This added up to fees in the amount of $176,528.38, considerably less than the requested lodestar fee amount of $518,445.85. Plaintiffs also sought $15,085.30 in statutory costs under 28 U.S.C. sec. 1920 and $37,257.00 in expert fees and non-taxable costs.3 The court allowed no specific costs related to the unsuccessful challenge to the validity of the examination, fixing total costs in the amount of $11,441.55.

Plaintiffs are not satisfied with the district court's substantial reduction in the amounts requested and assert that the district court committed legal error in holding that the validity of the 1994 examination was a separate claim from the existence of an equally valid, less discriminatory alternative promotion method. In response, the City argues that in a "mixed result" case as this one is, the district court may exercise broad discretion, and therefore, the fees award as limited by the district court should be affirmed.

This circuit has considered similar problems on a number of occasions attempting to apply the principles set forth by the Supreme Court in Hensley v. Eckerhart, 461 U.S. 424 (1983). See, e.g., Jaffee v. Redmond, 142 F.3d 409 (7th Cir. 1998); Kurowski v. Krajewski, 848 F.2d 767 (7th Cir. 1988). Our case, Spanish Action Comm. v. City of Chicago, 811 F.2d 1129 (7th Cir. 1987), interprets Hensley in helpful language which we set forth in some length as that language cannot be substantially improved.

In Hensley v. Eckerhart, 461 U.S. 424, 103 S. Ct. 1933, 76 L. Ed. 2d 40 (1983), the Supreme Court set out guidelines for calculating the proper amount of an attorney's fee award in cases where the plaintiff only partially prevails on his claims. The Court divided these partial recovery cases into two categories. The first category involves cases where the plaintiff presents distinctly different claims for relief that are based on different facts and legal theories. A plaintiff may not recover attorney's fees for time expended on an unsuccessful claim if that claim is "distinct in all respects from his successful claims." Id. at 440, 103 S. Ct. at 1943. Unrelated claims must be treated "as if they had been raised in separate lawsuits." Id. at 435, 103 S. Ct. at 1940.

. . . .

The second category of partial recovery cases, into which this action does fall, includes those cases in which the plaintiff's claims for relief involve a common core of facts or are based on related legal theories. Because the majority of counsel's time will be devoted to the litigation as a whole, as opposed to any one specific claim, this type of lawsuit cannot be viewed as a series of discrete claims. As a result, time spent on related claims that ultimately prove unsuccessful should not be automatically excluded from the attorney's fee calculation. Instead, the focus in arriving at the appropriate fee award should be on "the significance of the overall relief obtained by the plaintiff in relation to the hours reasonably expended on the litigation." Hensley, 461 U.S. at 435, 103 S. Ct. at 1940.

. . . .

Where the plaintiff fails to obtain all that he reasonably could have asked for and achieves only partial or limited success, the lodestar amount-- the product of the number of attorney's hours reasonably expended on the litigation as a whole times a reasonable hourly rate--is likely to be excessive. The Supreme Court therefore provided: "A reduced fee award is appropriate if the relief, however significant, is limited in comparison to the scope of the litigation as a whole." Hensley, 461 U.S. at 440, 103 S. Ct. at 1943. The Court, however, articulated no precise rule or formula to be followed in making such a reduction, instead choosing to leave this determination to the discretion of the district court in view of its greater familiarity with the litigation. Id. at 436-37, 103 S. Ct. at 1941. The Court did indicate that in reducing a fee award to reflect the plaintiff's limited success, a district court may attempt to identify specific hours that should be eliminated, or it may simply reduce the award across the board to account for the limited success. Id.

Spanish Action Comm., 811 F.2d at 1133.

In the present case, the district judge recognized that the relief obtained was very limited in relationship to the total relief sought. As Hensley points out, in such cases, a reduced fee amount is appropriate. Hensley, 461 U.S. at 440. As we have already noted, Hensley does not require the application of a precise rule or formula, leaving fee reduction to the discretion of the trial court. Spanish Action Comm., 811 F.2d at 1133. In reducing a fee award, a district court may attempt to identify specific hours to be eliminated or it may "simply reduce the award across the board to account for the limited success." Id. (citing Hensley, 461 U.S. at 436-37).

The district court determined that more than 90% of time expended by plaintiffs related exclusively to plaintiffs' "main goal" of having the 1994 examination declared invalid. That left less than 10% of plaintiffs' time which could be applied to their successful claim. The court did not automatically limit plaintiffs' fee award to 10% however, but went on to allow plaintiffs' fees for 20% of their time spent on the entire litigation, noting that this amount represented more than three times the percentage of the total time and expense devoted to the issues on which plaintiffs prevailed. The district court also allowed fees for 20% of the time plaintiffs' counsel spent reviewing materials in a companion case. In addition, plaintiffs' attorneys were awarded 100% of their fees in the remedy phase of the case and in prosecuting their fee petition. These fee and cost allowances we view as not only fair and reasonable, but practical. It, therefore, makes no difference that the court originally viewed each of the claims as separate as the applicable fee criteria was fully satisfied.

The district court is affirmed in all respects. The parties shall bear their own costs.