OPINION OF THE COURT

A. LEON HIGGINBOTHAM, Jr., Circuit Judge.

This appeal and cross-appeal arise from the district court's decision in an employment discrimination class action that was brought under 42 U.S.C. §§ 2000e et. seq. (1982) ("Title VII"), and 42 U.S.C. § 1981 (1982) (" § 1981"). The named plaintiff-appellants, Elbert G. Green and Robert Danley, brought that action on behalf of the class of black persons ("the class"), who had unsuccessfully sought employment in the Production and Maintenance ("P M") Department of the Fairless Hills, Pennsylvania plant of the United States Steel Corporation ("USX").

The class appeals the district court's denial of recovery to its members under the disparate treatment theory of discrimination. Additionally, the class asserts that the measure of damages that the district court awarded was inadequate.

On its cross-appeal, USX contends that the district court erred by its determination that it was liable to the class under the disparate impact theory of discrimination. It also alleges error in the district court's certification of the plaintiff class, and in the ultimate relief that the district court granted.

We have considered each of these contentions, and will affirm the decisions of the district court in significant part. Specifically, we will affirm the district court's findings regarding USX's liability for unlawful discrimination under disparate impact analysis, in violation of Title VII. We will also affirm the district court's certification of the plaintiff class and its award of injunctive relief. Further, we conclude that the district court's decision included a finding of liability against USX regarding applicants for summer positions, and that this finding was not clearly erroneous. Therefore, we will affirm the district court's award of damages to summer employees.

The district court erred, however, in its conclusion that the class failed to establish a prima facie case of discrimination under Title VII. The district court misconstrued the applicable standard established by the Supreme Court regarding intent in this type of discrimination suit, and therefore its conclusion of law that the class failed to make a prima facie showing of disparate treatment was in error. Accordingly, we will reverse that part of the district court's decision that found in favor of USX on the issue of liability under disparate treatment analysis.

Our finding of liability on this theory does not affect the damage award made by the district court. In our view, the district court exercised sound judgment in its adjudication of principal damages that resulted from USX's discriminatory hiring during the class period, and its decision need not be disturbed on the basis of our finding that an additional theory supports USX's liability. We will reverse on different grounds, however, the conclusions that the district court reached regarding prejudgment interest and front pay. As to these issues, the rationale that the district court employed to support its findings was erroneous.

Accordingly, we will affirm the damage award in part, vacate in part and remand for proceedings consistent with this opinion.

I. FACTS

The factual basis giving rise to these appeals is essentially uncontroverted. Class member Elbert Green, a black man, unsuccessfully applied for employment as a laborer in the P M Department of USX on April 11, 1973. Following the rejection of his application, he timely filed a charge of race discrimination with the Equal Employment Opportunity Commission ("EEOC"). Upon receipt from the EEOC of a notice of the right to sue. Green initiated this action and alleged violations of both Title VII and § 1981.

A. The Hiring Process

The focus of the class's challenge was upon the manner by which new P M laborers were selected. The P M department was the largest at USX and accounted for almost the entire percentage of initial hiring into the USX Corporation. See Green v. United States Steel Corp., 570 F. Supp. 254, 257 (E.D.Pa. 1983) (Findings of Fact ¶ 15) (P M hiring constituted 95% of new hires). The stated criteria for employment in P M unskilled positions were minimal. USX required only that potential P M employees be eighteen years old, pass a physical examination and be sufficiently literate to read safety signs. Id. (Findings of Fact ¶ 17). USX's view of the requirements for P M jobs reflected its intent that the criteria not be highly selective. The district court found that USX, in newspaper advertisements for P M laborers, stated that "no experience was necessary" and that the only requirement was "common sense and a desire to work." Id. at 258 (Findings of Fact ¶ 18). Any training or instruction necessary to the performance of the jobs was provided by USX on the job.

Prospective P M laborers were required to complete written applications in which they listed basic background information regarding their educations, prior work experiences and physical conditions. Applicants were also requested to list the names of any relatives who were current or previous USX employees. USX evaluated each of these factors and used them in the decision whether to employ. USX collected these applications on a continuing basis and recorded them in a file known as the Job Application Log ("JAL"). Particular information from each applicant's file, including his or her name, race and sex was placed on the JAL. It appears that the applicants for summer employment were separated from other applicants and entered into a different JAL.

USX employment supervisors periodically reviewed new applications and denoted certain ones ( e.g., those in which the applicant had listed a relative at USX) for special consideration. Applications marked in this way were given preference over the other applications and moved immediately into the "ready file." Applicants in the ready file were scheduled for interviews. Green, 570 F. Supp. at 259 (Findings of Fact ¶ 27). Applications that were not immediately marked as ready were placed into a general application file, and the applicants were categorized according to factors such as race, sex and prior military experience. By this categorization, USX attempted to identify applicants who were in "protected classes" according to EEOC guidelines, for later reference when moving applications from the general file to the ready file. Clerks who maintained the JAL and ready files were instructed to "maintain a good mix" of applicants in the ready file. USX's personnel supervisor testified that he had instructed clerks that 20-30% of those persons ultimately interviewed should be minorities. Id. (Findings of Fact ¶ 31).

Hiring was prompted by requests for additional personnel submitted by USX's various department heads. When such a request was made, the personnel clerks selected a number of the applicants from the ready file that they believed would yield the requested number of new employees. These applicants were then invited to interview. The interview process was two-fold. First, the applicant was interviewed by someone from the personnel department, and then by the foreman of the particular department that had requested additional personnel. The department foreman, however, had authority automatically to reject or accept the applicant after the interview. USX policy required that records of these interviews be kept, and that each decision to reject an applicant be accompanied by an explication of the rationale for the decision. These requirements were largely ignored, however, and few records were actually made. Green, 570 F. Supp. at 261 (Findings of Fact ¶ 50). Moreover, the interview sheets that were completed for unsuccessful applicants were ultimately destroyed by USX. Id. (same).

None of USX's personnel or foremen interviewers were given any formal instruction regarding the evaluation of candidates or the implementation of employment criteria. Green, 570 F. Supp. at 260 (Findings of Fact ¶¶ 37, 39). Also, there was no minimum education or work experience requirement for persons hired to do personnel interviews. USX presented evidence that showed that it gave its interviewers only broad guidelines on which to base their determinations. On this point, the district court found that,

[w]hen making hiring decisions, . . . interviewers evaluated applicants using the following criteria; education; work experience; attitude; initiative; personality; ability to take directions; alertness; intelligence; physical fitness (that is, physical restrictions); ability to communicate; military experience or training; vocational training; ability to meet work schedules; prior criminal record; familiarity with industrial or factory work; interest; prior employment history at Fairless Works; personal references; citizenship or alien status; having a relative in the [USX] work force; and age.

Id. (Findings of Fact ¶ 39).

The district court noted that the majority of these criteria were evaluated in a wholly subjective fashion. Regarding those criteria that were susceptible to objective measurement, "such as education and work experience, [USX] employed no uniform system in assessing one applicant's qualifications vis-a-vis another's." Green, 570 F. Supp. at 260 (Findings of Fact ¶ 39). Significantly, at no point did USX ever validate the hiring criteria that it used, nor did it ever ascertain whether those criteria were job-related. The district court found that, "contrary to the requirements of its own personnel policies, [USX] maintained no system to determine whether its subjective hiring criteria were being fairly and nondiscriminatorily applied to all applicants." Id. (Findings of Fact ¶ 44).

B. Evidence Regarding The Disparity

The class argued that the hiring system utilized by USX had a demonstrably adverse impact on black applicants and, further, that USX was aware of this impact and intentionally refused to alter the system to avoid the discrimination that resulted from its use. The class alleged, therefore, that USX was liable to the class under the "disparate impact" theory of discrimination established by the Supreme Court in Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 848, 28 L.Ed.2d 158 (1971), Albemarle Paper Co. v. Moody, 422 U.S. 405, 95 S.Ct. 2362, 45 L.Ed.2d 280 (1975), and International Bhd. of Teamsters v. United States, 431 U.S. 324, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977) ("Teamsters"), and also under the "disparate treatment" theory as established by the Supreme Court in McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973).

In support of its disparate treatment claim, the class presented statistical evidence of the disparity between whites and blacks in the USX hiring practice that was generated by a study completed by statistics expert, Dr. Samuel Litwin. Using information derived from USX's applicant and employment records, and from the Affirmative Action Quarterly Reports ("AAQR's") filed by USX. Dr. Litwin determined that, between 1971 and 1982, 49,585 applications for P M positions were received by USX. Of those applicants, 12,857 (25.9%) were black. Green, 570 F. Supp. at 262 (Findings of Fact ¶ 57). This data was analyzed for each year of the class period:YEAR APPLICATIONS APPLICATIONS % BLACK

C. The District Court's Decision

The district court considered the rebuttal evidence offered by USX and concluded that USX had failed adequately to explain the shortfalls in the hiring of blacks. Green, 570 F. Supp. at 272 (Findings of Fact ¶ 132). After making extensive findings of facts, the district court concluded that USX's "hiring practices had a statistically significant disparate and adverse impact on blacks," id. (Conclusions of Law ¶ 6), and, therefore, that USX was liable to the class under the disparate impact theory. Id. at 275 (Conclusions of Law ¶ 7). The district court concluded, however, that USX had not engaged in intentional discrimination and, accordingly, found in the company's behalf on the disparate treatment claim.

As the result of its conclusions, the district court ordered a second trial to determine the appropriate measure of damages. Following more than two years of discovery, that trial took place in April, 1986. At its conclusion, the district court awarded damages to the class in the amount of $12,397,928, which represented mitigated back pay and fringe benefits only. Green v. United States Steel Corp., 640 F. Supp. 1521, 1548 (E.D.Pa. 1986) ("Damages Opinion") (¶ 268), reprinted in Joint Appendix ("Jt.App.") at 179. The district court also awarded "prejudgment" interest dating from the 1983 liability judgment. Id. at 184 (Damages Opinion ¶¶ 294-95). Additionally, the district court permanently enjoined USX from further discrimination, and ordered the company to follow specific administrative and reporting guidelines to insure compliance with the injunction. Id. at 188 (Damages Opinion ¶ 317).

II. THE APPEALS

The class appeals from the decision of the district court on three grounds. First, it argues that the district court erred in its determination of liability by failing to find USX liable under the disparate treatment theory. Second, the class argues that the district court erred by failing to award front pay to compensate the class members for the period following the determination of liability and prior to employment by USX, during which the class members continued to suffer from the harm that resulted from USX's discriminatory acts. Finally, the class asserts that the district court erred by denying its request for a damage award that includes prejudgment interest dating from the discriminatory acts to the time of final judgment.

USX cross-appeals from the district court's decision. It argues that the district court erred by applying the disparate impact theory to this case in which a multi-component hiring system consisting of subjective elements was challenged. USX, together with amicus curiae The Equal Employment Advisory Council ("EEAC"), asserts that application of disparate impact theory is proper only in cases in which a single, objective hiring criteria has been identified by the plaintiff as having an adverse impact upon a protected class. USX and EEAC argue that the use of disparate impact analysis to attack USX's overall system of hiring, without the identification of any particular aspect of that system as discriminatory, placed an unfair burden upon USX that required it to demonstrate the legitimate business necessity of its entire hiring system. Moreover, EEAC argues that the application of disparate impact test to cases such as the present one, for which it is ill-suited, undermines the intention of Congress that disparate treatment analysis be employed in the disposition of these cases.

USX's cross-appeal raises several additional points. It contends that the district court erred by certifying the class, and also by subsequently failing to de-certify the class, pursuant to Fed.R.Civ.P. 27, because the named parties did not adequately represent the class for whom relief was sought. USX further contends that the district court erred by awarding damages to applicants for summer employment for whom no liability award had been entered, and by awarding damages for lost social security, pension and other workers' compensation benefits.

Because USX's contentions regarding disparate impact present significant issues of first impression to this Court, we will address them first in this discussion. We will then address the issues raised by the class on the principal appeal, and conclude by discussing the remaining issues raised by USX.

III. DISPARATE IMPACT

USX's contention that disparate impact analysis is inapplicable to cases in which a multicomponent hiring system that utilizes subjective criteria is challenged, brings to this Court for the first time issues that are the subject of a significant conflict between the various Courts of Appeals. Compare, e.g., Pouncey v. Prudential Ins. Co., 668 F.2d 795 (5th Cir. 1982) (use of disparate impact theory in cases involving challenge to multicomponent, subjective criteria hiring system is improper because it unjustly burdens employers); Watson v. Fort Worth Bank Trust, 798 F.2d 791 (5th Cir. 1986) (same), cert. granted, ___ U.S. ___, 107 S.Ct. 3227, 97 L.Ed.2d 734 (1987), argued Jan. 20, 1988, 56 U.S.L.W. 3513 (U.S. Feb. 2, 1988) and E.E.O.C. v. Federal Reserve Bank, 698 F.2d 633, 639 (4th Cir. 1983) (challenge that is not directed to "`specific procedure . . . [that] has a discriminatory impact on blacks' . . . does not fit within the model disparate impact claim") (quoting Pouncey, 668 F.2d at 800), rev'd on other grounds sub nom. Cooper v. Federal Reserve Bank, 467 U.S. 867, 104 S.Ct. 2794, 81 L.Ed.2d 718 (1984) with Atonio v. Wards Cove Parking Co., 810 F.2d 1477 (9th Cir. 1987) ( en banc) (holding that disparate impact analysis may be applied to subjective employment practices in employment discrimination actions) petition for cert. filed, 56 U.S.L.W. 3670 (U.S. March 29, 1988) (No. 87-1387, filed Feb. 9, 1988), and petition for cert. filed, 56 U.S.L.W. 3670 (U.S. March 29, 1988) (No. 87-1388, filed Feb. 10, 1988); Griffin v. Carlin, 755 F.2d 1516 (11th Cir. 1985) (concluding that subjective multicomponent hiring practices are subject to disparate impact analysis) and Segar v. Smith, 738 F.2d 1249 (D.C. Cir. 1984) (expressly rejecting Pouncey and holding that the application of disparate impact analysis to such cases is more consistent with the purposes of Title VII), cert. denied sub nom. Meese v. Segar, 471 U.S. 1115, 105 S.Ct. 2357, 86 L.Ed.2d 258 (1985). Although the propriety of applying disparate impact analysis to these cases is often discussed without distinguishing the problems presented by a "multicomponent" challenge from those presented by a "subjective criteria" challenge, the issues are distinct. We have examined the tensions presented by these issues individually and conclude nonetheless that the application of disparate impact analysis to this case is proper.

A. Multicomponent Criteria

In large part, USX's argument on this issue is predicated upon the rationale announced by Pouncey. In Pouncey, the Court of Appeals for the Fifth Circuit reviewed the district court's denial of an insurance company employee's claim that the overall system of promotion utilized by his employer had the effect of discriminating against him and the class of black employees that he represented. The court held that "the discriminatory impact model of proof . . . [was] not . . . the appropriate vehicle from which to launch a wide ranging attack on the cumulative effect of a company's employment practices." Pouncey, 668 F.2d at 800. The court determined that that model applies "only when an employer has instituted a specific procedure, usually a selection criteria [sic] for employment, that can be shown to have a causal connection to a class based imbalance in the work force." Id.

USX urges this Court to apply the rationale of Pouncey for two reasons: first, it asserts that the class has failed to identify sufficiently one aspect of USX's P M hiring practice on which to predicate its claim of discrimination. Because the class did not isolate a single criterion of the hiring practice that was discriminatory, USX asserts that it has failed to meet its burden of proving direct causation between some criterion and the disparate result. Second, USX argues that the application of impact theory to this case unfairly deprives it of the ability to assert the defense of "business necessity." That defense was established by the Supreme Court in Griggs as an employer's only justifying response to an employee's prima facie showing of discrimination under disparate impact analysis. See 401 U.S. at 431, 91 S.Ct. at 853; accord Albemarle Paper Co., 422 U.S. at 425, 95 S.Ct. at 2375 (holding that in order for an employer to meet the "business necessity" test, it must prove that the practice is significantly correlated with relevant work behavior.) USX argues that the class's claim effectively challenges the entire range of USX's procedures for hiring P M employees and, therefore, that USX must bear the "insurmountable burden" of demonstrating the business necessity of each of the criteria that it uses. See Brief for Appellee/Cross-Appellant at 13-14. We do not find any of these arguments persuasive.

USX asks us to disregard the fact that a statistically significant discriminatory result occurred as the result of its hiring practices, and focus only upon the fact that the class is unable to isolate a single offensive component. We cannot conclude that this is a proper balance of the burdens between employees and employers in such cases, or that it is the result contemplated by Congress, or by the Supreme Court in its interpretation of Title VII.

In Griggs, the Supreme Court established disparate impact analysis as a means of effectuating Congress's objective in the enactment of Title VII. That objective, the Court noted, "was to achieve equality of employment opportunities and remove barriers that have operated in the past to favor an identifiable group of white employees over other employees." Griggs, 401 U.S. at 429-30, 91 S.Ct. at 853. The Court determined that pursuant to Title VII, "practices, procedures or tests neutral on their face, and even neutral in terms of intent," could not be maintained if their result was to "`freeze' the status quo of prior discriminatory employment practices." Id. at 430, 91 S.Ct. at 853. The Court was no more specific in its description of instances to which disparate impact analysis was applicable than by its reference to "practices, procedures, or tests." In our view, however, it is a mistake to construe this language as somehow evincing the Court's intent to circumscribe the use of disparate impact analysis in a manner that leaves a broad range of offensive discriminatory practices unassailable. We can too easily imagine the instance in which an employer, who without any discernible discriminatory intent, devises a scheme the aggregate components of which cause disproportionate hiring. Under the test urged upon this Court by USX, such a scheme would be immune from challenge. As the Court of Appeals for the Eleventh Circuit noted in Griffin,

limiting the disparate impact model to situations in which a single component of the process results in an adverse impact completely exempts the situation in which an adverse impact is cause by the interaction of two or more components.

755 F.2d at 1525 (citing Gilbert v. City of Little Rock, 722 F.2d 1390 (8th Cir. 1983), cert. denied, 466 U.S. 972, 104 S.Ct. 2347, 80 L.Ed.2d 820 (1984)).

The correctness of our view is supported by the definition given to criteria susceptible to disparate impact analysis set forth in federal regulations promulgated jointly by the EEOC, the Civil Service Commission, the Department of Labor and the Department of Justice. 29 C.F.R. §§ 1607.1-.17 (1987). Those regulations, entitled "Uniform Guidelines on Employee Selection Procedures (1978)," were designed "to assist employers, labor organizations, employment agencies, and licensing and certification boards to comply with requirements of Federal law prohibiting employment practices which discriminate on grounds of race, color, religion, sex and national origin." 29 C.F.R. § 1607.1(B). Significantly, these regulations define those procedures that are subject to disparate impact review as

[a]ny measure, combination of measures, or procedure used as a basis for an employment. Selection procedures include the full range of assessment techniques from traditional paper and pencil tests, performance tests, training programs, or probationary periods and physical, educational, and work experience requirements through informal or casual interviews and unscored application forms.

29 C.F.R. § 1607.16(Q) (emphases added).

Disparate impact analysis was intended to provide a means to an end: eradicating the dilatory effects of racial discrimination in hiring. The analysis is not an end in itself, and should not be scrutinized as such. It was precisely this "excessive preoccupation with the various formulae used in . . . employment discrimination case[s]" that we admonished against in Goodman v. Lukens Steel Co., 777 F.2d 113, 129-30 (3d Cir. 1985), aff'd on other grounds, ___ U.S. ___, 107 S.Ct. 2617, 96 L.Ed.2d 572 (1987). In that case, we "observed that [these formulae] are simply tools designed to aid in the analysis of evidence. The ultimate question remains whether the defendant has discriminated." 777 F.2d at 130.

Disparate impact analysis is best understood as a burden-shifting device that fairly and reasonably apportions the burdens of proof between parties in a manner that is best suited to achieve the desired goals of Title VII — whether the case presents a challenge to a single criterion or to several. In this case, we are persuaded that the class has sufficiently identified the interview process as one component of the USX hiring system that resulted in the disparate hiring results. We note, however, that the arguments raised in the class's principal brief to this Court indict each of the facets of the USX hiring policy. More significantly, as we noted above, the district court reached its conclusion with the view that the class's challenge was to the entire hiring practice. The district court explicitly rejected USX's argument that impact analysis was improper to challenge a multicomponent hiring system. Green, 570 F. Supp. at 273 (Conclusions of Law ¶ 6). It found that the class's challenge to USX's hiring system was sufficient to make a prima facie showing of discrimination, and noted that

[f]urther refinement of [the class's] definition of the hiring selection system [it] attack[s] was made impossible by [USX's] own refusal to be pinned down as to the selection criteria it employs, beyond listing twenty subjective criteria . . . and stating that these criteria were applied as an "amalgam."

Id. at 274 (Conclusions of Law ¶ 6). We hold that on this record, the district court's validation of the class's multicomponent challenge to USX's hiring system under the disparate impact theory of analysis was proper.

The argument against the application of disparate impact analysis to cases in which several aspects of a hiring system are challenged, appears predicated upon an inference drawn from the Supreme Court's holding in Connecticut v. Teal, 457 U.S. 440, 102 S.Ct. 2525, 73 L.Ed.2d 130 (1982). In Teal, the Court held that an employer could not avoid liability under impact analysis when one of the components in its hiring scheme was discriminatory, by demonstrating that its "bottom line" reflected an overall acceptable hiring of blacks. Id. at 452, 102 S.Ct. at 2533. Proponents of the view that disparate impact should not be applied in cases where multiple criteria are challenged argue that a mutuality principle should apply. If the employer cannot advert to the "bottom line" result and use the overall statistics of the system to validate the system notwithstanding the presence of a discriminatory component, then the employee should also be precluded from using "system statistics" to invalidate the entire hiring system without pinpointing a specific component. See Brief for Appellee/Cross Appellant at 25-26; see also Wilborn, The Disparate Impact Model of Discrimination: Theory and Limits, 34 Am.U.L.Rev 799, 830 (1985). This argument, however, fundamentally misperceives the rationale of Teal. Teal expanded the scope of impact analysis so that it could be used in a greater number of instances in which discrimination was present. The Supreme Court noted that Title VII was enacted with the broad intention of removing "`artificial, arbitrary and unnecessary barriers to employment,'" Teal, 457 U.S. 447, 102 S.Ct. 2531 (quoting Griggs, 401 U.S. at 431, 91 S.Ct. at 853), and that to meet that goal, the statute sought to address any " limitations and classifications that would deprive any individual of employment opportunities." Id. at 448, 102 S.Ct. at 2531 (emphases in original). The Court recognized that in some instances, discriminatory criteria would have an effect that would not be apparent in the "bottom line" number. It instructed, however, that the fact that the discrimination was not evident in the "bottom line" did not satisfy entirely the purpose of Title VII disparate impact analysis. It noted that

[w]hen an employer uses a nonjob-related barrier in order to deny a minority or woman applicant employment or promotion, and that barrier has a significant `adverse effect on minorities or women, then the applicant has been deprived of an employment opportunity. . . . In other words [Title VII] . . . prohibits discriminatory "artificial, arbitrary, and unnecessary barriers to employment . . ." that "limit . . . or classify . . . applicants for employment in any way which would deprive or tend to deprive any individual of employment opportunities."

Id. at 448, 102 S.Ct. at 2531 (quoting Griggs, 401 U.S. at 429-31, 91 S.Ct. at 852-53) (emphases in original). Rather than advancing these goals, however, the mutuality theory that USX urges would provide a significant obstacle to their realization. It would require that statistically significant disparate effects that are evident in the "bottom line" be ignored until and unless the employee can identify with particularity one offensive component. In too many cases, this view would place upon plaintiffs the onerous burden of sifting through layers of bureaucracy in order to determine precisely the component that results in the discriminatory result. Moreover, it would encourage employers to structure hiring and promotion systems to contain enough subjective criteria to avoid the necessity of responding to complaints of discrimination, or vindicating disparate results. This notion is contrary to Teal's expansion of disparate impact analysis to reach those cases in which it is apparent that "arbitrary and unnecessary barriers . . . limit or classify . . . applicants . . . [in ways that] would deprive or tend to deprive . . . individual[s] of employment opportunities." Id. This broad view of disparate impact analysis is consistent with Congress's intent by its enactment of Title VII, and is reflected by the legislative history of the 1972 Amendments to Title VII in which both the House and the Senate explicitly endorsed Griggs. That legislative history also reveals that Congress understood that in order to address adequately those cases involving discrimination that pervades an entire hiring system, an approach that scrutinizes for more than manifest signs of discriminatory intent is necessary. In one Senate Report, the Congress noted that "[e]mployment discrimination as viewed today is a . . . complex and pervasive phenomenon. Experts familiar with the subject now generally describe the problem in terms of `systems' and `effects' rather than simply intentional wrongs." Id. (quoting S.Rep. No. 415, 92nd Cong., 1st Sess. 5 (1971)) (emphases supplied).

We are unpersuaded by USX's contention that the requirement that it defend a challenge to its hiring practices is unreasonable. Applying disparate impact analysis to this employer's hiring "system" and measuring the disproportionate "effects" on minority hiring that result may impose a difficult burden upon the employer, but not an unfair one. In these cases, the employer has better access and opportunity than the plaintiffs to evaluate critically the interrelationship of the criteria that it uses in its hiring practice, and to determine which aspects actually result in discrimination. This burden is not triggered merely because an employer's hiring practice fails to result in actual hiring that reflects statistically expected percentages of whites and blacks. Rather, it is imposed only in those cases in which the plaintiffs have made a prima facie showing that the disparity would not have occurred except for the employer's hiring practices. Cf. Segar, 738 F.2d at 1271 ("[a]n employer will face the justificatory burden only after a plaintiff class has shown a disparity in the positions of members of the class and the majority group who appear to be comparably qualified"). Moreover, the employer retains the opportunity to demonstrate that each of the criteria that it uses are necessary. We are convinced that "[a] rule placing this justificatory burden on the employer advances the purpose of Title VII far better than would the contrary result." Id. We also believe that the imposition of this obligation will not result in an unfair burden upon employers. For these reasons, we sanction the use of disparate impact theory in these cases.

B. Subjective Criteria

Apart from the multifaceted character of its employment practices, USX argues that disparate impact analysis was improperly applied to this case because USX's hiring procedure included subjective criteria, particularly the individual interviews conducted by the foremen. USX asserts that these procedures are not susceptible to impact analysis, but instead are only properly reviewable under disparate treatment theory. We reject this contention.

Although we recognize the reasonableness of an employer's use of subjective criteria to help make hiring or promotion decisions, nothing in the Supreme Court's analysis in Griggs, or that we can find in Title VII, suggests that these criteria should be insulated from scrutiny under impact analysis. As we noted above, the objective of impact analysis is to effect Title VII's goal of eliminating "artifical, arbitrary and unnecessary barriers to employment." Griggs, 401 U.S. at 431, 91 S.Ct. at 853. We are not persuaded that there is a well-reasoned distinction between objective criteria that present barriers to employment opportunities for blacks, and subjective criteria that lead to the same results, that requires us to circumscribe the use of the tool that the Supreme Court has provided to combat such discrimination. In our view, subjective practices such as interviews are just as capable of operating to hinder the employment opportunities of blacks as are tests or other presumptively objective criteria. See Griffin, 755 F.2d at 1525; Rowe v. General Motors Corp., 457 F.2d 348, 359 (5th Cir. 1972). They are also just as likely to be "engendered by a totally benign purpose, or carried on as a matter of routine adherence to past practices whose original purposes are undiscoverable." Atonio, 810 F.2d at 1487. For these reasons, we cannot conclude that application of disparate treatment analysis, and its requirement that plaintiffs demonstrate the intent to discriminate, is the only proper method of scrutinizing subjective criteria. Further, we share the fear expressed in Griffin that the "[e]xclusion of such subjective practices from the reach of the disparate impact model . . . is likely to encourage employers to use subjective, rather than objective selection criteria." 755 F.2d at 1516. By this method, an employer could use a wide range of criteria during an interview or on a recommendation form without ever articulating these criteria clearly or, more importantly, without validating their necessity to the selection procedure. In our view, "[i]t could not have been the intent of Congress to provide employers with an incentive to use such devices rather than validated objective criteria." Id. Moreover, we cannot believe that it was the intent of Congress to allow unarticulated biases or unconscious prejudices to result in discriminatory hiring practices merely because they appear in the form of an unsupervised interview.

For reasons similar to those regarding the applicability of impact analysis to evaluate a multicomponent system, we reject USX's argument that requiring it to defend a challenge to its use of subjective criteria places an undue burden upon it.

In our view, proving business necessity is no more onerous in a case involving subjective practices than one involving objective practices, because in either case the employer is the person with knowledge of what his [or her] practices are and why he [or she] uses the methods and criteria he [or she] does, as well as the person with superior knowledge of precisely how his [or her] employment practices affect employees.

Atonio, 810 F.2d at 1486. We are persuaded that the district court was correct in its application of impact theory to the challenge that the class raised and, accordingly, we will affirm.

IV. DISPARATE TREATMENT

The district court denied the class's disparate treatment claim because it concluded that the class had failed to demonstrate a prima facie case under that theory. Green, 570 F. Supp. at 277 (Conclusions of Law, ¶ 8). The district court reached this conclusion because it found that the class had failed to "prove that the defendant intentionally and habitually discriminated against blacks on the basis of their race." Id. Although it found that USX was aware that it utilized an employment system that resulted in statistically significant adverse impact upon the hiring of black workers, the district court concluded that the evidence of intent required by Title VII and § 1981 was greater. The district court stated that the fact "`[t]hat an employer was aware of or totally indifferent to the racially discriminatory impact of its hiring policies is not of itself sufficient to establish a prima facie case and shift the burden of explanation to the defendant.'" Id. (Conclusions of Law ¶ 9) (quoting Gay v. Waiters' Dairy Lunchmen's Union, 694 F.2d 531, 552 (9th Cir. 1982)). It held that the class needed to show "an invidious purpose," id., or "a hostile, discriminatory motive" on USX's part in order to make the prima facie showing of disparate treatment. Id. at 278. In this analysis of the evidence necessary to establish a prima facie case under disparate treatment, however, the district court was in error.

The test that the district court employed placed a much more onerous burden upon the class than the Supreme Court has required for the establishment of a prima facie case of disparate treatment. In McDonnell Douglas, the Supreme Court set forth four requisites to the prima facie showing:

(i) that a [plaintiff] belongs to a racial minority; (ii) that he [or she] applied and was qualified for a job for which the employer was seeking applicants; (iii) that despite his [or her] qualifications, he [or she] was rejected; and (iv) that after his [or her] rejection, the position remained open and the employer continued to seek applicants from persons of the complainant's qualifications.

411 U.S. at 802, 93 S.Ct. at 1824. To be sure, this test was intended as a guidepost for courts in their implementation of disparate treatment analysis, and not as a rigid formula. We are unpersuaded, however, that the test employed by the district court in this case is in any way consistent with the McDonnell Douglas parameters. In our view, a showing of discriminatory intent does not require the explicit demonstration by a plaintiff of an employer's "invidious purpose" or "hostile motive." Rather, to make this showing, "[t]he plaintiff must prove by a preponderance of the evidence that she [or he] applied for an available position for which she [or he] was qualified, but was rejected under circumstances which give rise to an inference of unlawful discrimination." Texas Dep't of Community Affairs v. Burdine, 450 U.S. 248, 253-54, 101 S.Ct. 1089, 1094, 67 L.Ed.2d 207 (1981) (emphasis supplied).

We recognize and adhere to the prior holdings of the Supreme Court and of this Court with regard to the requisite intent that must be demonstrated in order for a plaintiff to prevail in a disparate treatment case, and we do not conclude that the district court erred by its scrutiny of the record for "hostile motive" or "invidious purpose." Our interpretation of the McDonnell Douglas and Burdine tests, and of the precedent of this Court, however, leads us to the conclusion that the evidence of that intent need not be so direct and uncontrovertible as "smoking gun" evidence to compel a rebuttal by the defendant, and to require that that rebuttal be substantive and not mere pretext. In our view, the presence of abundant circumstantial evidence from which the inference of discriminatory treatment can be reasonably drawn is sufficient to require that the employer respond, and that its response be closely scrutinized.

The burden of making the prima facie showing in these cases was not intended to be onerous. See Burdine, 450 U.S. at 253, 101 S.Ct. at 1093. The requirement of the prima facie showing was intended as a working tool to provide a means to determine which cases raise sufficiently compelling inferences of discrimination to require rebuttal. It was not intended as a hallmark of whether the complainant has proved his or her case. In some cases the inference of unlawful discrimination may be raised, and thus "[a] prima facie case of disparate treatment may be established by statistics alone if they are sufficiently compelling." Griffin, 755 F.2d at 1525 (citing Eastland v. TVA, 704 F.2d 613, 618 (11th Cir.), opinion modified and reh'g denied, 714 F.2d 1066 (1983), cert. denied sub nom. James v. TVA, 465 U.S. 1066, 104 S.Ct. 1415, 79 L.Ed.2d 741 (1984)). As the Supreme Court stated in Teamsters,

[s]tatistics showing racial or ethnic imbalance are probative . . . [in employment discrimination cases] only because such imbalance is often a telltale sign of purposeful discrimination; absent explanation, it is ordinarily to be expected that nondiscriminatory hiring practices will in time result in a work force more or less representative of the racial and ethnic composition of the population in the community from which employees are hired.

431 U.S. at 340 n. 20, 97 S.Ct. at 1857 n. 20; accord Croker v. Boeing Co., 662 F.2d 975, 991 (3d Cir. 1981) (holding that "[w]hen plaintiffs allege classwide racially discriminatory treatment in violation of Title VII, proof of discriminatory motive is essential, although the burden may be met in some situations by presentation of statistical evidence that permits an inference of racial discrimination") (citing Wetzel v. Liberty Mut. Ins. Co., 508 F.2d 239 (3d Cir.), cert. denied, 421 U.S. 1011, 95 S.Ct. 2415, 44 L.Ed.2d 679 (1975)); Massarsky v. General Motors Corp., 706 F.2d 111, 118 n. 14 (3d Cir.) ("disproportionate impact of an employment policy may itself be circumstantial evidence of an intent to discriminate"), cert. denied, 464 U.S. 937, 104 S.Ct. 937, 78 L.Ed.2d 314 (1983). When those statistics are combined with anecdotal evidence further implicating the presence of unlawful motive in hiring, "the cold numbers [are brought] convincingly to life," Teamsters, 431 U.S. at 339, 97 S.Ct. at 1856 and the plaintiff's prima facie case of discriminatory treatment is enhanced.

In the present case, the district court found that there was a significant racial disparity in USX's hiring, of which USX was aware. The district court also found that USX's hiring decisions were predicated upon unguided subjective criteria, that none of the hiring criteria that USX used were ever validated, and that inadequate records were kept by USX to evaluate properly the progress that it was making in its compliance with a consent decree that was the result of previous employment discrimination litigation. We are aware that USX had implemented a scheme by which, it asserts, it attempted to comply with the anti-discrimination laws. That fact, however, without more, does not satisfy the company's obligation to explain the discriminatory results that persisted in its hiring practices. Viewing the statistical disparity in light of the facts that USX's hiring procedure relied so greatly upon the subjective criteria such as the determination of its foremen in interviews, and that the expressed required qualifications for employment are so minimal, we are further convinced that the class succeeded in establishing a prima facie case of disparate treatment.

We find the rationale employed in Parson v. Kaiser Aluminum Chem. Corp., 575 F.2d 1374 (5th Cir. 1978), cert. denied sub nom. Local 13000 United Steel Workers of Am. v. Parson, 441 U.S. 968, 99 S.Ct. 2417, 60 L.Ed.2d 1073 (1979), and Rowe v. General Motors Corp., 457 F.2d 348 (5th Cir. 1972), useful to our resolution of the issue before us. In each of these decisions, the appellate court reviewed promotion procedures that resulted in a significant disparity between blacks and whites who sought elevation within a company. The appellate court in Parson noted similarity between the promotion procedure challenged in the case under its review, and the procedure that had earlier been invalidated in Rowe to the extent that both procedures relied heavily upon non-supervised evaluations by foremen, and utilized "standards [that were] vague and subjective . . . [without] safeguards in the procedure designed to avert discriminatory practices." 575 F.2d at 1385 (citing Rowe, 457 F.2d at 358-59). The court noted that "[t]hese factors result in `procedures which depend almost entirely upon the subjective evaluation and favorable recommendation of the immediate foreman.' They are therefore `a ready mechanism for discrimination against Blacks.'" Id. (quoting Rowe, 457 F.2d at 359). In light of the statistical disparity between promotions of whites and blacks, and the use of unguided subjective criteria used in the promotion procedure, the appellate court in Parson concluded that the district court had erred by failing to employ this standard to determine whether the plaintiff had proved his case of discrimination.

These are precisely the circumstances present in this case. USX's hiring of employees in its P M Department resulted in a statistically significant disparity between black and white applicants. It based its hiring decisions on unguided, subjective criteria that were never validated as essential to USX's employment practice, or determined to be nondiscriminatorily applied. We find that the presence of these factors is sufficient to raise an inference of unlawful discrimination that requires rebuttal by USX.

Having concluded that the class succeeded in making a prima facie showing of discrimination under the disparate treatment theory, we must now determine whether USX articulated a sufficient nondiscriminatory reason to rebut the inference that its hiring practice was discriminatory. USX asserts two reasons for its disproportional hiring: (1) that the class members' qualifications were less than those of the applicants ultimately hired; and (2) that a greater proportion of black applicants voluntarily "dropped out" of the hiring process prior to the making of hiring decisions. See Brief for Appellee/Cross Appellant at 37-39. Both of these arguments were explicitly rejected by the district court. Our review of the record reveals that the district court was correct in concluding that these assertions are wholly without merit, and we find them pretextual.

USX's "best qualified" hiring defense was predicated upon the notion that, although the basic employment requisites were minimal, the company nonetheless made substantive evaluations of the applicants and chose the "best." It asserts in this regard, that the disproportionate hiring merely reflected the fact that white applicants were better qualified. USX asserted specifically that "promotability" was one criterion that it employed to identify the better qualified applicants.

The record is completely devoid of evidence supporting this contention. It reflects only that USX employed a nonspecific amalgam of subjective criteria in reaching its hiring decisions. Significantly, the district court found that

the process by which [USX] divined who, among its applicants, were the best qualified was wholly subjective, consisting essentially of combining the gut reactions to the applicant of employees in the personnel office and one or more foremen in the plant. No evidence was presented or offered to show that applicants were tested in any way for ability to do specific tasks required of workers above the laborer level.

Green, 570 F. Supp. at 269 (Conclusion of Fact ¶ 110) (emphases supplied). Moreover, even the evidence that was presented by USX negated the contention that the company's selection process included any component designed to result in hiring the "best qualified." A study, performed by USX witness Dr. Finis Welch, was introduced to show that USX's hiring reflected differences in applicants' qualifications rather than racial bias. What that study demonstrated, however, was only that the hiring process was unseemly. Dr. Welch's "Winner/Loser" study revealed what the district court called a "bizarre" pattern of hiring by USX. Id. at 270 (Findings of Fact ¶ 119). Among the hiring practices shown by Dr. Welch's study, for example, were that applicants with seven years' experience as manufacturing operatives had no better chance of being hired by USX than did persons with no work experience; than an applicant with one year experience as an assembly line worker in an ice cream plant was more likely to be hired by USX than an applicant with 10 years experience as a steel mill laborer; and that applicants with college training or some clerical experience had a reduced chance of being hired.

We are aware that no one criterion is significant to every job and that, for some jobs, criteria such as college educations, which are in themselves valuable, are not valuable to the employer in light of his or her employment needs. In the present case, however, USX articulated no logically comprehensible standard that it used to identify applicants who were "better qualified" to do the jobs for which USX was hiring. Moreover, it presented no evidence demonstrating that it evaluated applicants non-discriminately to determine who those persons were. "Ironically," as the district court noted, the one criterion that Dr. Welch's study did identify as significant was that "being white gave an applicant an `edge,' that is, an increased possibility of being hired." Green, 570 F. Supp. at 270 (Findings of Fact ¶ 119).

USX's second articulated rationale for the disparity in hiring is similarly unpersuasive. Its assertion that a disproportionate share of black applicants voluntarily withdrew from the hiring process is not supported by the evidence of record. USX relies upon the testimony of Robert Talley, an employee in the company's personnel office who, without any statistical expertise, reviewed hiring data between 1973 and 1974 and concluded that a statistically proportionate share of "job opportunities" was extended to blacks. The district court rejected Talley's conclusions as "unreliable" because of "numerous problems with his data and his methods." Green, 570 F. Supp. at 271 (Findings of Fact ¶ 126). The district court heard testimony from Larry Edwards, Talley's supervisor at USX, who testified that "there were no reliable [USX] records from which rates of applicant drop-outs or failures of physical exams could be determined." Id. (Findings of Fact ¶ 127). Moreover, Edwards testified that "the phenomenon of `dropping out' during the interviewing process, or after a job offer had been extended, occurred `across the board,' and . . . did not occur disproportionately often among blacks." Id. In light of this evidence, and of the class's rebuttal of Talley's conclusions because of errant data, the district court concluded that "[USX had] not explained the statistically significant shortfall in black hiring for entry-level P M jobs, vis-a-vis black representation in the applicant pool, during the class period . . ." Id. at 272 (Findings of Fact ¶ 132). We also conclude that this asserted justification for the shortfall in USX hiring is not supported by the evidence of record.

Accordingly, we find each of the reasons relied upon by USX to be pretextual rationalizations for discriminatory hiring. We conclude, therefore, that USX is liable for employment discrimination under the disparate treatment theory of liability, and we will reverse the district court's decision to the contrary.

V. REMAINING ISSUES ON THE CLASS'S APPEAL

The remaining issues raised by the class concern the measure of damages that the district court awarded. Although, as stated above, we will not vacate the entire award based upon our holding that USX was liable to the class under the disparate treatment theory, we find it necessary to vacate part of the district court's decision regarding damages and remand for rehearing. We conclude that the district court erred in the rationale that it employed to deny the class's claims for prejudgment interest and front pay and thereby abused its discretion.

A. Prejudgment Interest

The question of whether to award prejudgment interest in this case is committed to the sound discretion of the district court. The particular rationale that the district court relied upon to deny a full measure of prejudgment damages, however, appears inconsistent with established precedent regarding the award of such relief. In its damage award, the district court expressed the view that the "defendant's lack of intentional discrimination is . . . [a] mitigating factor" in its behalf weighing against the award of prejudgment interest that dated back to the acts of discrimination. Jt. App. at 183 (Damages Opinion ¶ 291) (emphasis in original). Therefore, the district court awarded interest from the date of the liability judgment to the damages decision only. Although we defer ordinarily to the discretionary authority of the district courts in such cases, and we recognize that in reaching a determination on such awards, the district court may reasonably consider the extent and determinability of the employer's liability, nothing in the cases suggest that the award of prejudgment interest should be predicated upon a finding of intentional discrimination or bad faith, or that such an award should be offset by the absence of such bad faith. In fact, the Supreme Court has instructed that the opposite view is more consistent with the purpose in granting remedies for Title VII discrimination.

In Albemarle Paper Co., the Court held that backpay awards could not be conditioned upon the showing of bad faith. The Court noted that Congress had intended that that relief go to the "consequences" of discriminatory employment practices, not simply to the motivation for the discrimination. Albemarle Paper Co., 422 U.S. at 422, 95 S.Ct. at 2373. The Court stated that if such awards were available

only upon a showing of bad faith, the remedy would become a punishment for moral turpitude, rather than a compensation for workers' injuries. This would read the "make whole" purpose right out of Title VII, for a worker's injury is no less real simply because his employer did not inflict it in "bad faith."

Id. Although Albemarle Paper Co. specifically addressed back pay awards, the Court's primary focus concerned the "make whole" purpose of Title VII damage awards. This purpose is equally applicable to the award of prejudgment interest, and therefore the rationale of Albemarle Paper Co. is applicable to this case.

B. Front Pay

The district court's denial of front pay to the cless was predicated upon its view that the amount of such an award would require too much speculation. The speculation that was required resulted from USX's inability to hire the class members at the time that judgment against it was entered, and concerned when USX would be able economically to make any new hires for the P M jobs. The district court found that, "[i]n all likelihood, [USX] will not resume hiring until at least 1990 or 1991." Jt.App. at 180 (Damages Opinion ¶ 274). The class sought an award of the mitigated wages for 1984 and 1985 (the two years immediately succeeding the liability judgment) as an appropriate estimate of front pay. The district court concluded that the "proposed estimate was `sheer guesswork,' and would require [it] to speculate wildly" as to the proper amount of damages. Id. at 181 (Damages Opinion ¶ 277). In this conclusion, however, we find t hat the district court was in error.

The decision whether to award front pay as relief in Title VII cases rests certainly within the discretion of the trial court. Dillon v. Coles, 746 F.2d 998, 1006 (3d Cir. 1984). The exercise of that discretion, however, must be sound, and it must be consistent with the rationale and purpose underlying front pay, which, in furtherance of the overall remedy scheme of Title VII, is "to make a victim of discrimination `whole' and to restore him [or her] to the economic position he [or she] would have occupied but for the unlawful conduct of his [or her] employer." Maxfield v. Sinclair Int'l, 766 F.2d 788, 795 (3d Cir. 1985), cert. denied, 474 U.S. 1057, 106 S.Ct. 796, 88 L.Ed.2d 773 (1986); cf. Albemarle Paper Co., 422 U.S. at 416, 95 S.Ct. at 2371 (exercise of discretion should not be dictated by the court's "`inclination, but . . . its judgment; and its judgment is to be guided by sound legal principles'") (quoting United States v. Burr, 25 F.Cas. 30, 35 (C.C.Va. 1807) (No. 14, 692d) (Marshall, C.J.)); Franks v. Bowman Transp. Co., Inc., 424 U.S. 747, 770-71, 96 S.Ct. 1251, 1267, 47 L.Ed.2d 444 (1976) (reversing the district court's denial of grant of seniority as remedy for discrimination and stating that "[d]iscretion is vested not for purposes of `limit[ing] appellate review of trial courts, or . . . invit[ing] inconsistency and caprice,' but rather to allow the most complete achievement of the objectives of Title VII that is attainable under the facts and circumstances of the specific case") (quoting Albemarle Paper Co., 422 U.S. at 421, 95 S.Ct. at 2373 (brackets in original) (citations omitted)). As we concluded in Dillon, the award of this type of relief "necessarily implicates a prediction about the future." 746 F.2d at 1006. The fact that that prediction is difficult should not work to defeat the purpose of the relief.

In the present case, it was undisputed that the time at which USX would be economically able to hire new employees was uncertain, although almost assuredly not before seven or eight years following the determination of its liability to the class. In this light, an award to the class of front pay in an amount reflecting mitigated wages for two years following the judgment of liability, appears more as a reasonable compromise between the liability for harm caused by USX's discrimination and the class members' ability to mitigate, than "wild speculation." Moreover, to the extent speculation would be inherent in the award, USX and not the class should have borne the risk that the ultimate conclusion would prove in fact erroneous. "The risk of lack of certainty with respect to projections of lost income must be borne by the wrongdoer, not the victim." Goss v. Exxon Office Sys. Co., 747 F.2d 885, 889 (3d Cir. 1984) (citing Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555, 51 S.Ct. 248, 75 L.Ed. 544 (1931)).

A review of our decision in Goss may prove instructive, and elucidate why this case is not one in which the speculation necessary should have precluded the award. In Goss, we affirmed the district court's award of front pay. We rejected as requiring insupportable speculation Goss's assertion that she was entitled to a greater measure of front pay. Goss, who had been employed as a salesperson for Exxon Office Systems and whose income there had been based upon commissions from sales, argued that she was entitled to a front pay award that included an amount intended to provide additional compensation to her salary with her new employer. She asserted that she was entitled to this additional measure of compensation because it was unlikely that she would earn as much as a salesperson for the different company as she would have earned at Exxon. This assertion was apparently predicated upon Goss's belief that, because she had more knowledge of the Exxon product and more experience selling it, her success selling it would be greater than her success selling the new product, and thus her commission salary would have been larger. In its denial of this request, the district court concluded that whether or not Goss would be less successful in her new job than she would have been at Exxon depended in significant part upon future market conditions. The district court thus noted that a difference of sales would not result only from Goss' own skills or anything over which Exxon could exercise control. The district court recognized that not only could it not predetermine whether in fact Goss would be less successful as a salesperson in her new job, but that it also could not predetermine whether her income would not have similarly diminished during the same time period if she had been still employed by Exxon. The court concluded that this type of speculation was unreasonable. It therefore awarded front pay only for that time period following Exxon's liability for back pay and prior to Goss' start in a commensurate position with her new employer.

We affirmed this judgment. The type of speculation that the district court avoided in that case, however, is distinguishable from that required in the present case. Here, the district court was asked to fix some amount that would help to offset future harm that would certainly be caused by USX's past discrimination. The class contended that two years' mitigated wages were sufficient and necessary to compensate its members adequately for the injury caused by USX. This amount may in fact be greater (or lesser) than the actual damages to which the class was entitled. It is not our charge to determine whether that request reflected in fact an appropriate measure. The district court, however, should not have completely refused to award front pay merely because some speculation was necessary. USX should have been presented with the opportunity and the burden to demonstrate that the class members failed in their obligation to mitigate their losses or that for some other reason the award of two years' mitigated wages was unreasonable. USX, however, and not the class, must ultimately bear the risks of speculation. As the Supreme Court has noted, "whatever of uncertainty there may be in this mode of estimating damages, it is an uncertainty caused by the defendant's own wrongful act; and justice and sound public policy alike require that he should bear the risk of the uncertainty thus produced." Story Parchment Co., 282 U.S. at 564, 51 S.Ct. at 251 (citations omitted).

VI. REMAINING ISSUES ON USX'S CROSS-APPEAL

Finally, we address the remaining issues raised by USX in its cross-appeal. We find that each of the contentions raised is without merit, and therefore we will affirm the district court's judgment as to those claims.

A. Class Certification

USX argues that the district court erred by certifying the class because the claims raised were not sufficiently typical as required by Fed.R.Civ.P. 23(a). USX asserts that, because its personnel who were charged with hiring responsibility based their decisions upon varied subjective criteria, each instance of alleged discrimination was necessarily distinct and, therefore, that the certification of the class was improper.

This contention rests in part upon the same rationale that USX asserts with regard to the applicability of disparate impact analysis to this case — that where subjective criteria are employed, it is necessary to evaluate each instance of discrimination independently. As we noted above, we find nothing in Title VII that requires that result. The argument is similarly unpersuasive in this context. It completely misperceives the typicality requirement of Rule 23. That rule requires that class representatives present sufficiently common issues of law and fact upon which the class action is based, to assure that the interests of the absent class members will be adequately represented. As we noted in Eisenberg v. Gagnon, 766 F.2d 770, 786 (3d Cir.), cert. denied sub nom. Weinstein v. Eisenberg, 474 U.S. 946, 106 S.Ct. 342, 88 L.Ed.2d 290 (1985), however, "`[t]ypical' is not identical." Our review under this first requirement of Rule 23 is to assure that the claims presented by the class representatives are consistent with those presented by the entire class. "[T]ypicality entails an inquiry whether `the named plaintiff's individual circumstances are markedly different or . . . the legal theory upon which the claims are based differs from that upon which the claims of other class members will perforce be based.'" Id. (quoting Weiss v. York Hospital, 745 F.2d 786, 809 n. 36 (3d Cir. 1984), cert. denied, 470 U.S. 1060, 105 S.Ct. 1777, 84 L.Ed.2d 836 (1985)). We are satisfied that the district court correctly applied this test in its certification of the class, and therefore we will not disturb its judgment.

Similarly, we find that the district court was correct in its decision not to decertify the class. In essence, USX argues that the specific claim presented by the named plaintiff Danley became moot when the class as a whole dropped from its claim allegations of discrimination regarding the particular year in which Danley had made his application to USX. For that reason, USX contends that the district court should have decertified the plaintiff class. This assertion is also incorrect.

The Supreme Court has instructed, in a similar case in which the named representative could not maintain a viable individual claim, that the class can maintain its claims so long as "`a live controversy [remains] at the time'" the case is reviewed. Franks, 424 U.S. at 755, 96 S.Ct. at 1260 (quoting Sosna v. Iowa, 419 U.S. 393, 402, 95 S.Ct. 553, 559, 42 L.Ed.2d 532 (1975)) (brackets in original). The Court held in Franks that, "[a]lthough the named representative no longer had a personal stake in the outcome, . . . `[w]hen the District Court certified the propriety of the class action, the class of unnamed persons described in the certification acquired a legal status separate from the interest asserted by [the named representative],' . . . and, accordingly the `cases or controversies' requirement of Art. III of the Constitution was satisfied." Id. at 753, 96 S.Ct. at 1259 (quoting Sosna, 419 U.S. at 402, 95 S.Ct. at 559). These same circumstances are present in this case. Despite the fact that Danley's individual claim was no longer viable, the class retained claims that clearly presented the district court and this Court "`with a case or controversy in every sense contemplated by Art. III of the Constitution.'" Id. at 755, 96 S.Ct. at 1260 (quoting Sosna, 419 U.S. at 398, 95 S.Ct. at 557).

In Goodman, we noted that a broad view of Rule 23 in Title VII claims was proper. Thus we found that, although racial bias in the discharges during employee probationary periods was a class claim, and that none of the named representatives were dismissed during their probationary periods, the claims of discharge in employment overall were nonetheless sufficiently common to justify class certification. See 777 F.2d at 123. We noted, however, that an expansive view of Rule 23 does not obviate the necessity that each requirement of that rule be met. Accordingly, we found with regard to another aspect of the class's claims, racial discrimination in the initial assignment of new employees, that the named plaintiffs were not sufficiently representative of the class. That finding, however, is distinct from the issue presented in the present case. In Goodman, we had determined that the named representatives' claims regarding discrimination in initial assignments fell outside the applicable limitations period. Thus, the question of the sufficiency of the named parties of the class was jurisdictional because their claims were no longer actionable. In this case, the named representative, Danley, did not prevail on his claim because the evidence supporting discrimination during the year in which he was hired was insufficient. That conclusion, however, did not affect the commonality of Danley's claims with the remainder of the class. For this reason, we conclude that it was not an abuse of discretion for the district court to deny the motion for decertification of the class.

B. Summer Hires

Finally, we review USX's contention that the district court erred by its award of damages to applicants for summer employment. USX argues that "[n]owhere in the District Court's liability decision . . . did the Court find as a matter of fact or law that USX unlawfully discriminated against black applicants for summer employment in 1972 and/or 1973." Brief For Appellee/Cross-Appellant at 42. This argument, however, misstates the findings of the district court. We conclude that the district court's opinion clearly included applicants for summer employment in the class to which USX was liable.

In its order dated August 26, 1980 certifying the class, the district court stated that the plaintiff class was comprised of "[a]ll black persons who unsuccessfully sought employment at the Fairless Hills, Pennsylvania Plant of [USX] between July (11) (sic), 1972, and the [date of this order]." Green, 570 F. Supp. at 256. That order was never amended to exclude summer applicants from the plaintiff class, and there is no indication that the district court contemplated that this sub-class was to be severed from the overall class. To the contrary, in several instances in its opinion, the district court made particular reference to the summer applicants. See id. at 257 (Findings of Fact ¶ 16) ("[a]pproximately 15% of new P M hires during the class period were hired for summer jobs"); id. at 271 (Findings of Fact ¶ 128) (questioning the statistical evidence presented by USX's expert, and noting that part of the reason for the doubt was that the expert's data did "not include applications submitted by persons seeking summer jobs, although they are members of the class") (emphases supplied). We are convinced by these references, and by our review of the record, that the district court made findings of fact regarding the summer hires that justified inclusion of this sub-class in the damage award. In this light, USX's challenge regarding the summer applicants is more appropriately characterized as an assertion that the record does not support the liability determination that the district court made. On this question, our standard of review is whether the court's findings were clearly erroneous, see Pullman, 456 U.S. 273, 102 S.Ct. 1781, and our conclusion is that they were not.

In our review of the propriety of the factual findings of the district court, we need only determine whether the decision that it reached was reasonable — not whether a different conclusion would also have been reasonable. Before we may properly intrude upon any factual finding that the district court made, we must be "left with the definite and firm conviction that a mistake has been committed." United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 542, 92 L.Ed. 746 (1948). In this case, we are not in this way convinced of any error by the district court.

Although much of the data regarding summer hires for the years in question was unavailable, we note that the reason for that unavailability was USX's destruction of all of the applications for employment prior to 1975. See Green, 570 F. Supp. at 261 (Findings of Fact ¶ 49). Notwithstanding this fact, the class's statistical expert, Dr. Litwin, was able to present findings to the district court that "[USX's] hiring had a statistically significant negative and disparate impact upon blacks during the overall class period [,] . . . [and this conclusion was] the same whether summer hires were included or excluded." Id. at 265 (Findings of Fact ¶ 83) (emphasis supplied).

In our view, the inferences that the district court drew from these facts were reasonable. We are not left with the certain impression that the district court's findings that the summer hires should be included in the overall class for the purpose of determining liability were clearly erroneous. Therefore, we will affirm this aspect of the district court's judgment.

VII. CONCLUSION

For the foregoing reasons, we will affirm the judgment of the district court as to its holding of liability of USX to the class under the disparate impact theory of discrimination. We will also affirm that part of the judgment in which the district court ruled that class certification was proper, and that summer applicants were properly included in the class to whom liability is owed. We will reverse the judgment of the district court to the extent that it did not also find that USX was liable to the class under the disparate treatment theory of discrimination.

In accordance with this holding, and in light of our additional conclusions that the district court erred in the rationale that it employed in its denial of prejudgment interest and front pay to the class, we will affirm the damage award in part, vacate that award in part and remand for rehearing, consistent with this decision, of these issues of damages.