In this title VII case, the appellant class, represented by Ophine Giles, Sara Tucker, and Rosalyn Stephens (Giles or class), appeals from a judgment of the United States District Court for the Northern District of Alabama in favor of the appellee Alabama State agencies and officials (Ireland).1 The district court found that the challenged employment practices did not have an adverse impact on blacks nor did those practices reflect disparate treatment of blacks. We affirm in part, vacate in part, and remand to the district court for resolution of two remaining facets of Giles’ discrimination claim that were not adequately resolved below.
Issues
Three principal questions are presented in this appeal:
(1) whether the district court abused its discretion in certifying the class;
(2) whether the district court erred in refusing to give evidentiary weight or collateral estoppel effect to the findings of discrimination in United States v. Frazer2 with respect to Giles’ discrimination claim; and (3) whether the trial court’s findings of fact are clearly erroneous, or its conclusions of law in error, with respect to the merits of Giles’ title VII claims.
Background
Partlow State School and Hospital for the Mentally Retarded is operated by the Department of Mental Health of the State of Alabama. Among its nonprofessional staff, Partlow maintains job classifications denominated mental health worker (MHW) I, II, and III.3 These are nonexempt, merit system positions. Service as a MHW-I is a prerequisite to advancement to MHW-II and, in turn, service as a MHW-II is a prerequisite to promotion to MHW-III. In August 1976, Partlow imposed a moratorium on promotions from MHW-I to MHW-II. The six original plaintiffs4 brought suit challenging a full range of allegedly discriminatory employment practices at Partlow.
Facts.
At the time of the moratorium on promotions, blacks were disproportionately concentrated in the lowest level mental health worker positions at Partlow. The MHW-I group was then two-thirds black. By 1980, less than 5 percent of MHW-III’s at Part-low were black. Circumstances at Partlow, however, cannot be fairly evaluated without an appreciation of the events leading up to the moratorium.
In 1970, then Chief District Judge Frank M. Johnson, Jr., held in Frazer that the State of Alabama was engaged in a systematic pattern and practice of racial discrimination with respect to state personnel *1370in the administration of federally financed grant-in-aid programs.5 As a result of the discriminatory employment practices of the State of Alabama, blacks were found to be concentrated at the lowest levels of state employment. The state was found to have maintained a variety of all black positions, all white positions, and segregated facilities. Further, Judge Johnson found that in many cases the recruiting examination, appointment, training, promotion, retention, and a variety of other personnel practices of the State of Alabama constituted unlawful racial discrimination in violation of the Constitution of the United States. Specific remedial orders were issued and requests for supplemental relief were granted by the district court in Frazer.
In 1976, in resolving one such request for supplemental relief, Judge Johnson found that the Department of Mental Health had hired a substantial number of blacks since the original decree in Frazer.6 That hiring, however, was found to be in response to the “right to treatment” cases, Wyatt v. Stickney and Wyatt v. Aderholt,7 which required the hiring of additional support staff by the Department of Mental Health in order to provide adequate patient care.8 The hiring of blacks by Partlow was not found to be remedial or pursuant to the original Frazer decree. In granting supplemental relief in Frazer, Judge Johnson specifically ordered that psychiatric aides I and II9 shall be eligible for promotions to higher rated and supervisory positions within their line of progression on the basis of their training and experience. The Department of Mental Health was also ordered to employ and recruit black professional personnel.10
In 1976, the Department of Mental Health announced a 5-year plan which projected a decrease in the number of patients at Partlow from approximately 1,200 in 1976 to approximately 800 by 1981. This forecast was based on a projected change in the circumstances of patient care from large institutions such as Partlow to locally based facilities. In fact, the patient population at Partlow fell to 712 by 1983. This projected decrease in patient load resulted in some reduction in the projected number of MHW-I’s and -II’s needed at Partlow.
Partlow sought to accommodate this decline by limiting the number of MHW-II’s. The trial court, however, found that a demonstrated need for additional MHW-II’s continued until at least January 1982.
Between 1976 and April 1982, Partlow denied promotions to the position of MHW-II to any person occupying the position of MHW-I. As a result of a reclassification study, nine black and five white MHW-I’s were promoted to MHW-II in April of 1982. Additionally, in 1976, Partlow converted from a medical model of patient care to a developmental model utilizing health care teams — denominated “nursing care teams.” Between October 1977 and January 6, 1982, various MHW-I’s were assigned to nursing care teams at Partlow. On those teams MHW-I’s were involved in dispensing medication and in the unsupervised treatment of patients. They were performing the same functions as the MHW-II’s oh those teams, yet, they received no increase in pay. While being paid as MHW-I’s, various individuals were also called upon to perform other MHW-II duties by serving as charge aides.
The nursing care teams were never intended to be permanent and were recently *1371disbanded and reformed. The reduction in the number of MHW-I’s and -II’s between 1976 and 1982 was accompanied by an increase in the number of licensed practical nurses (LPN’s) and registered nurses (RN’s) at Partlow. Reformation of the nursing care teams was necessary to ensure that medication is dispensed only by licensed personnel — LPN’s and RN’s — and not by MHW-I’s and -II’s.
Proceedings Below.
The six original plaintiffs brought this action, on their own behalf and on behalf of a class consisting of other black employees at Partlow, claiming discrimination under title VII11 and under 42 U.S.C. §§ 1982 and 1983. The original plaintiffs sought certification of this action as a class action racial discrimination suit challenging a full range of employment practices at Partlow including promotion, hiring, levels of pay, training programs, work assignment, and various other employment practices.
The trial court heard evidence on the practices at Partlow and concluded that, with respect to the discrimination claims, the requirements of Fed.R.Civ.P. 2312 are met in one regard only and not in other regards. On April 4, 1982, the court certified the class consisting of black employees at Partlow who on or after October 1,1977, have served in the capacity of MHW-I with respect to the claims that they have been underpaid or that they have been denied promotions to the positions of MHW-II and MHW-III because of their race.13
The class had filed a motion for summary judgment seeking a determination of liability based on offensive use of the Frazer decision as collateral estoppel. The district court, however, denied that motion. Ireland subsequently filed a summary judgment motion to treat the class action as one for supplemental relief under Frazer. That motion was also denied.
At the conclusion of the parties’ cases, Judge Pointer ruled from the bench in favor of Ireland. He found that, while both black and white MHW-I’s were underpaid, they were both treated substantially the same. The court examined the promotion registers and concluded that there was “no showing that blacks would have been or were affected more than whites by the decision not to promote anyone.” The court found that the decision not to hire or promote MHW-I’s to the MHW-II level between 1976 and 1981 was not the result of a decision, desire, or purpose to discriminate against blacks but, rather, was a consequence of the declining patient population at Partlow during that period.
The class filed a motion for a new trial and for amendment of the trial court’s findings. That motion was granted only to the extent of correcting a few of the court’s findings and was denied in all other respects. The trial court, in its September 21, 1982, supplemental opinion, found that *1372the differences between black and white employees based on the corrected figures wei;e too insubstantial to support an inference of disparate impact. The earlier judgment was sustained. Giles appeals.
Class Certification
Giles contends that the trial court failed to consider evidence of racial discrimination at all levels of the Partlow work force. First, Giles argues that the court below erroneously constricted membership in the class by excluding orderlies — a predominantly black position at Partlow. Further, Giles argues that by relegating nonprofessional plaintiffs who were not included in the class to supplemental relief under Frazer, the trial court departed from the law of the case and acted contrary to precedent under title VII.
In reviewing the district court’s decision regarding class certification under Fed.R.Civ.P. 23, we will not disturb the district court’s determination unless it can be shown that the district court abused its discretion.14
The court below did not expressly delineate why the various other discrimination claims, originally filed yet not certified for class treatment, were inappropriate for class consideration. The court stated only that these other claims “either are appropriate for resolution as supplemental to the Frazer matter or [are] due to be dealt with as individual claims.” The trial court concluded that the requirements of Fed.R. Civ.P. 23(a) and (b)(2) were not met with respect to these other claims. While a class may be certified even where the requirements of rule' 23(b)(2) are not met provided that some other paragraph of rule 23(b) is satisfied, paragraph (b)(2) is particularly applicable to a claim of unlawful discrimination against a class.
Giles argues that orderlies were improperly excluded from the class. Ireland, however, avers that, since the issue of inclusion of orderlies was not raised below, we cannot consider it on appeal.15 Giles originally sought certification of this action as an across-the-board class action discrimination suit in an attempt to represent all black employees at Partlow. The district court considered and rejected that claim. Thus, Giles’ claim, that orderlies should have been included in the class, was raised below and is properly before us on appeal.
The district court heard testimony that the position of orderly is not subject to state merit system procedures. The MHW-I, -II, and -III positions, on the other hand, are merit system positions. Thus, different recruiting, employment, and promotion policies are applicable to the orderly position than to the mental health worker positions. The underpayment claim does not apply to orderlies because they were not required to perform MHW-II duties. Further, orderlies were not affected by the moratorium on promotions to MHW-II. The claims of discrimination with respect to orderlies are basically the same generalized claims — challenging concentration in low level positions and obstacles to advancement — that were raised and resolved in Frazer.
While these generalized claims of disproportionate representation of blacks are common to both orderlies and mental health workers, we cannot conclude the trial court abused its discretion in determining that the overlap is not sufficient to satisfy the “commonality” requirement of Fed.R.Civ.P. 23(a)(2). Orderlies have no interest in the two principal claims of the MHW-I plaintiffs: underpayment and denial of promotion opportunities.
*1373Giles argues that the certification of the class by the district court departs from the law of the case established by Judge McFadden’s November 21, 1980, order denying Ireland’s summary judgment motion. Judge McFadden held only that the original plaintiffs were not precluded by the Frazer decision from seeking additional and individual relief. That ruling, however, does not compel the conclusion that those plaintiffs, who were excluded when the class was later certified, are entitled to pursue relief in this proceeding.
The vitality of private enforcement mechanisms under title VII is well established in light of the general intent of Congress to accord parallel or overlapping remedies against discrimination.16 The availability of private enforcement avenues, however, does not entitle a plaintiff to participate in any particular action. Giles’ reliance on the law of the case doctrine is misguided. The policy of efficient use of judicial resources, urged by Giles in this regard, underlies Fed.R.Civ.P. 23 as well. That policy will not vitiate the specific requirements of the rule.
On the basis of the record before us, Giles has not established that the trial judge abused his discretion in certifying the class.
Collateral Estoppel
After ruling from the bench on the class certification issue, Judge McFadden proceeded to deny Giles’ motion for summary judgment. Giles sought to establish Ireland’s liability by asserting the collateral estoppel effect of the findings of discrimir xtion in Frazer. The court below refused to accord Frazer res judicata or collateral estoppel effect noting that the issues in this case must be resolved on the basis of facts that have transpired since judgment was entered in Frazer.
Giles argues on appeal that discrimination in the MHW-I, -II, and -III positions was clearly established in Frazer and that those discriminatory practices continued unabated at Partlow. Seeking to use the Frazer findings of discrimination offensively, Giles contends that there has been no change in the facts or law and that Ireland should not have been allowed to “relitigate” the issue of liability. Ireland responds that (1) the court in Frazer entered no specific findings on the situation as it existed at Partlow and (2) since the findings in Frazer predate the challenged policies, those findings are not applicable to this case.
To prevail, Giles must establish that the trial court abused its discretion in refusing to accord Frazer collateral estoppel effect.17 The Eleventh Circuit, in Williams v. Bennett,18 articulated the circumstances in which the doctrine may be applied, thus:
Preclusive effect will be given to the adjudication of an issue litigated in a prior proceeding if the issue in the subsequent proceeding is identical to the one involved in the prior action, the issue was actually litigated, and the determination of the issue was necessary in the prior action. * * * In Parklane Hosiery Co. v. Shore, 439 U.S. 322 [331-32], 99 S.Ct. 645 [651-52], 58 L.Ed.2d 552 (1979), the Supreme Court granted federal trial courts broad discretion to permit the offensive use of collateral estoppel [footnote omitted], indicating, however, that the exercise of this discretion is circumscribed by considerations of fairness to the defendant in foreclosing relitigation of an issue. * * * A trial judge may justifiably refuse to allow offensive use *1374of collateral estoppel if the defendant did not have an incentive to litigate the issue vigorously in the prior proceeding or if the defendant did not have a full and fair opportunity to litigate it. [Citations omitted.]
The precise issue before the district court here was not fully litigated in Frazer, nor was there any opportunity to litigate it. Frazer dealt with public enforcement of title VII against a pattern and practice of racial discrimination in state employment. The Frazer court never addressed the moratorium on promotions challenged by Giles because that practice did not begin until after the most recent order granting supplemental relief in Frazer was entered. Thus, Frazer provides no basis for collateral estoppel in this proceeding. The trial court correctly proceeded to evaluate Partlow’s employment practices giving the Frazer decision “significance” and “circumstantial value” with respect to the issue of discrimination, yet, declining to give it collateral estoppel effect. The trial court has not been shown to have committed error.
Discrimination
Giles asserts on appeal that a violation of title VII may be found on the evidence of record in that the challenged employment practices perpetuate the effects of past discrimination. The challenged employment practices take two forms: (1) underpayment and (2) denial of promotion. Although Giles asserts that they are intertwined, they are treated separately in Giles’ brief. We address them seriatim.
This case was tried to the court below on both disparate treatment and adverse impact theories of proof. As plaintiff, Giles bore the initial burden of making out a prima facie case of discrimination.19 Under a disparate treatment theory, (1) Giles must establish a prima facie case; (2) Ireland would then be required to offer a legitimate, nondiscriminatory reason for the policy; and (3) Giles would then have the opportunity to establish that the proffered reason was pretextual.20 Proof of discriminatory intent is an essential element of Giles’ prima facie case under a disparate treatment theory, although intent can in some situations be inferred from the mere fact of differences in treatment.21
Under an adverse impact theory, on the other hand, proof of discriminatory motive is not required.22 Under that theory, Giles was required to demonstrate that the challenged practice had a substantial adverse impact on a protected group. If so, the court would then have been required to determine whether the policy constitutes a business necessity and, if so, whether adequate alternatives with a lesser adverse impact would serve the employer’s needs.23
To prevail, Giles must establish that the trial court’s findings of fact, whether of subsidiary or of ultimate fact,24 are clearly erroneous25 or that the court erred as a matter of law.
Underpayment.
Individuals employed as MHW-I’s at Partlow were regularly required to per*1375form MHW-II duties without an increase in pay. The trial court rejected Ireland’s contentions that these additional duties were within the MHW-I job scope and that they were imposed only on an emergency basis. Rather, the court found that these additional duties were imposed on a regular basis and could have been anticipated. MHW-I’s were required’ to serve on nursing care teams, doing the same work as MHW-II’s, and to act as charge aides, a MHW-II duty, while being paid as MHW-I’s. On the basis of these additional duties, nine black and five white MHW-I’s were reclassified as MHW-II’s in April 1982.
While the trial court questioned the propriety of Partlow’s employment practice of requiring MHW-I’s to perform MHW-II duties without a corresponding pay increase, it determined that discrimination had not been established. The court below found that both blacks and whites were underpaid pursuant to this employment practice. The statistical approach to the evidence presented by Giles was rejected in favor of Ireland’s statistical treatment of the evidence. The underpayment policy was found to affect blacks and whites equally and, thus, not found to be discriminatory.
Giles, on appeal, attacks the evidentiary basis for the district court’s ruling on the underpayment claim. Giles contends the evidence shows that proper wages were paid to 10 times as many whites as blacks. The issue, in Giles’ view, is not whether the policy of underpayment affected both whites and blacks but, rather, whether Partlow favored whites in paying a proper wage.
We see no substantive significance in the semantics urged on us by the class. Ireland’s payment of proper wages is in issue only inferentially. It is the disparate treatment of, or adverse impact on, black MHW-I’s that forms the basis of Giles’ underpayment claim under title VII. Whether the underpayment claim is evaluated under a disparate treatment or an adverse impact theory, resolution of that claim on this record depends fundamentally on the quality of Giles’ statistical proof. Giles’ characterization on appeal of the underpayment claim appears to be no more than a diaphanous attempt to illuminate the statistical evidence in the most favorable light.
The court below declined to rely on Giles’ benchmark — 53 percent black; 47 percent white — which was based on all workers at Partlow. The court heard testimony that the 53-47 benchmark proffered by Giles’ expert witness, Dr. Rogers, failed to account for individual differences in education, experience, and seniority. Further, that benchmark was not adjusted for the recent influx of black MHW-I’s at Partlow resulting from the Wyatt decisions.26 All of these factors affect the validity of Giles’ benchmark and, consequently, the quality of Giles’ statistical proof.
Rather than analyzing the underpayment claim in relation to the benchmark proffered by Giles, the trial court analyzed the underpayment claim in relation to the numbers of MHW-I’s who were (1) required to serve on nursing care teams, (2) required to serve as charge aides, and (3) reclassified in April 1982 based on their actual duties. The trial court found that these are the class members who were underpaid at Partlow pursuant to the challenged employment practices.
Roughly, 66 percent of MHW-I’s at Part-low were black. Of that group, blacks underpaid on nursing care teams represented only 59 percent of all underpaid MHW-I’s serving on nursing care teams. Similarly, the nine reclassified blacks represent only 54.5 percent of MHW-I’s required to work as charge aides and 64 percent of all reclassified MHW-I’s.
On the basis of these figures, the court below found that no higher percentage of blacks than whites were classified and paid as MHW-I’s and required to perform the duties of MHW-II’s on the nursing care teams. The trial court also found that no greater percentage of blacks than whites, *1376as a percentage of persons of those races in the MHW-I category, were required to serve as charge aides. Finally, the court found that the reclassification of nine black and five white MHW-I’s in 1982 also failed to establish discrimination.
Relative to the benchmark adopted by the court below, we cannot conclude that the trial court’s finding — that blacks were not influenced to a greater degree than whites — is clearly erroneous. Nor can we conclude on the basis of the record before us that the trial court’s rejection of Giles’ proffered benchmark and its use of the black composition of the MHW-I category as a benchmark were clearly erroneous. Thus, we hold that the trial court did not err in holding that Giles had not established a claim of discrimination under title VII with respect to the underpayment of black MHW-I’s at Partlow.
Denial of Promotions.
No promotions of anyone holding the position of MHW-I to the position of MHW-II were made at Partlow from 1976 until April 1982. The trial court, however, found that a demonstrated need for additional MHW-II’s continued at Partlow during that period. The trial court also found that, as a consequence of the predominantly black composition of the MHW-I job classification, the decision not to promote any MHW-I to MHW-II would appear to fall more heavily on blacks than it would on whites.
The court below, however, proceeded to examine the register of eligibles27 from which appointments would have been made during the period the policy was in effect. After examining the top half of the 1977 and 1979 registers,28 the trial court found that no adverse impact on blacks was shown.29
The district court went on to determine whether an intent to discriminate had been established. The number of patients at Partlow, as predicted, had fallen dramatically by 1982. Additionally, Partlow had converted from a developmental model to a medical model of health care. Both of these factors ameliorated the need for MHW-II’s. Further, Partlow maintained the approximate ratio of MHW-II’s to MHW-I’s.30 On the basis of these three factors in the absence of direct evidence of discriminatory intent, the trial court found that the moratorium on promotions was not the result of a decision or desire or purpose to discriminate against blacks.
Giles, on appeal, challenges the trial court’s reliance on the registers, urging that the entire population of MHW-I’s is the appropriate pool of applicants who have suffered injury to their advancement opportunities. It would have been futile to apply for promotion avers Giles, and, accordingly, each MHW-I has suffered injury.
At trial, Giles introduced the testimony of Dr. Rogers on the futility of applying for promotion to MHW-II. The law will not require futile acts and Giles quite properly amplifies the vitality of that principle under title VII. It is well recognized that “unlawful employment practices may be so successful as to totally deter victims of gross and pervasive discrimination from *1377applying.”31 The futility doctrine is applied where the nonapplicant meets the not always easy burden of proving that he or she would have applied for the job had it not been for the employer’s discriminatory practices.32
Not all MHW-I’s were qualified and eligible for promotion to MHW-II. A large number of MHW-I’s at Partlow apparently did not view applying for promotion a “vain and useless exercise.” The 1977 and 1979 registers each contain the names of approximately 100 eligibles.33 Apart from Dr. Rogers’ testimony, which failed to persuade the trial court, Giles offered no direct evidence to establish the futility of applying for promotion.
We cannot agree with Giles that the decision of the Supreme Court in International Brotherhood of Teamsters v. United States,34 requires that the futility rule be invoked in this case. In Teamsters, the Court stated that35
[individual nonapplicants must be given an opportunity to undertake their difficult task of proving that they should be treated as applicants and therefore are presumptively entitled to relief accordingly.
The Court’s discussion of nonapplicants in Teamsters, however, was in the context of determining entitlement to relief. Disparate treatment was found in that case on the basis of the experiences of actual applicants. We do not perceive the anomaly argued by Giles where the group entitled to relief is larger than the class of claimants whose experience formed the basis for the determination of adverse impact. Limitation of the pool of potential applicants, for the purpose of analyzing a potential adverse impact, does not prevent other members of the larger pool of potential applicants, who appeared initially not to have been affected by the discriminatory practice, from attempting to establish their entitlement to relief once an adverse impact has been shown.
Giles has not established that the trial- court’s failure to find that potential applicants were excluded or deterred from applying, or that they viewed application as a futile act, is clearly erroneous. The trial court did not err in limiting its analysis of adverse impact to actual applicants for promotion. Further, nothing in this record indicates that the district court erred in relying on the registers of eligibles as reflecting those individuals who were potentially adversely impacted by the denial of promotions.
We cannot agree with Giles that the court below considered only the 1977 register. In its August 3, 1982, bench opinion and in its supplemental opinion entered September 21,1982, the trial court expressly discussed and analyzed both the 1977 and the 1979 registers.
Since no promotions were made at Partlow under the challenged policy, the district court was unable to determine precisely who had been denied promotion. The trial court also recognized that it was presented with insufficient evidence to determine exactly how many MHW-I’s would have been promoted in the absence of the challenged policy. The trial court was compelled, however, to construct some reasonable, albeit arbitrary, measure to assess the implications of the failure to promote. It chose to look at the top half of both the 1977 and 1979 registers.
In the absence of more precise evidence of who would have been promoted, the top half of each register reflects those individu*1378ais who would have been more likely to secure promotion in the absence of the challenged employment practice. Further, use of the top half of each register does not unnecessarily restrict the size of the pool of persons potentially affected by the practice. We conclude that the trial court did not err in proceeding to analyze adverse impact on the basis of the top half of the 1977 and 1979 registers.
Of the top half of the 91 eligibles on the 1977 register, 23 are black and 22 are white. Of the top half of the 113 MHW-I’s on the 1979 register, 38 are black and 19 are white.
In view of the balanced composition of the first register, we conclude that the district court’s finding, that the evidence failed to establish an adverse impact on blacks during the 30 months the first register was in force, was not clearly erroneous. The second register, however, reveals that blacks were burdened more heavily than whites by the no promotion policy during the period the second register was in force. The trial court, however, found that the second register failed to establish an adverse impact in view of (1) the preceding 2V2-year period in which no impact was shown and (2) the further concentration of blacks on the second register resulting from certification of more blacks than whites on a register from which no one had been appointed.36
While we cannot conclude on the basis of the record before us that the trial court’s failure to find an adverse impact on the basis of the second register alone is clearly erroneous, we hold that the trial court’s analysis was inadequate to allow it to find that the challenged policies did not adversely impact on blacks at Partlow.
Giles argues on appeal that the no promotion policy is not supported by any business necessity and that adequate alternatives having a lesser adverse impact were available. The district court, however, entered no findings or conclusions on these issues, presumably because it concluded either that Giles had failed to establish a prima facie case or that Ireland had rebutted Giles’ prima facie case, rendering it unnecessary to reach these issues.
The burden on blacks reflected in the second register indicates that, although the policy of not promoting to MHW-II was fair in form, it deprived a greater number of blacks than whites of the opportunity for advancement. Even were the result of the policy an appropriate racial balance, that would not preclude Giles from establishing a prima facie case on the basis of deprivation of the opportunity for advancement. The district court was required to look more closely at the challenged employment practice and its effects before concluding that the practice had no adverse impact.
The court below failed to consider two facets of Giles’ claim of discrimination with respect to the no promotion policy. First, the district court did not adequately consider the effect of the no promotion policy to raise a barrier to advancement opportunities that was not based on job performance and that has not been found to be supported by business necessity. Second, the district court failed to consider whether the failure to promote, although it may not itself have had an overt adverse impact, perpetuated the effects of past discrimination.
Barring Promotional Opportunity to Mental Health Worker III
Giles argued separately that an adverse impact could be found from denial of *1379promotional opportunities to MHW-III. No MHW-I’s could advance to MHW-III so long as Partlow pursued a policy of not promoting to the MHW-II level. Only those individuals who were already MHW-II’s could advance to MHW-III — a group that at Partlow was disproportion',tely white during the period the policy was in force: only 18 percent of MHW-II’s were black and only 14 percent of MHW-III’s were black.
Incumbents of Mental Health Worker Jobs at Partlow: 10-1-77 to 6-1-82
Mental Health Worker Category Number of Black Employees Number of White Employees
I 504 (66%) 254 (34%)
II 23 (18%) 108 (82%)
III 5 (14%) 31 (86%)
Thus, the evidence of record would appear to establish that blacks were barred in disproportionate numbers relative to whites from advancing to the position of MHW-III.
The district court did not in its bench opinion address this facet of Giles’ claim. Rather, the court below attempted to resolve the issue in a footnote to its supplemental opinion:
By suspending the opportunity for promotion to MHW-II, the decision also had the effect of postponing any opportunity for the next promotion to Mental Health Worker-Ill. At oral argument on their post-judgment motion, the plaintiffs asserted that the court’s decision had not addressed their claims regarding this position. On reflection, the court concludes that plaintiffs’ claims regarding the MHW-III position are dependent upon their claims regarding the MWH-II [sic] position.
We cannot agree with the district court’s assessment.
While the statistical evidence on impact with respect to the MHW-II position is equivocal, the underrepresentation of blacks in the pool of potential candidates for promotion to MHW-III would appear to be compelling. Further, the evidence of record fails to relate the justifications for the no promotion policy to the MHW-III position. MHW-III is a supervisory position and the court below made no findings on the need, or the reduced need, for MHW-III’s in response to declining patient population at Partlow. There is no evidence that MHW-III’s were utilized on nursing care teams; therefore, their displacement in favor of certified personnel is not established. No findings were entered on the relative ratios of MHW-III’s to other classes of mental health workers. In short, the district court’s reasoning appears to be inapplicable to the claims regarding the MHW-III position.
Even discounting the implications of the second register with respect to the greater burden of the policy on blacks than on whites, an equal impact on blacks and whites did not preclude the district court from finding a violation of title VII. The Supreme Court has held barriers to advancement violative of title VII, even where the result of such barriers is an appropriate racial balance.
In Connecticut v. Teal37 the Court expressly rejected a “bottom line” approach to a claim of racial discrimination in promotions under title VII:
We hold that the “bottom line” does not preclude respondent employees from establishing a prima facie case, nor does it provide petitioner employer with a defense to such a case.
sje $ jjs $ sjs *
“[Title VII] proscribes not only overt discrimination but also practices that are fair in form, but discriminatory in operation. The touchstone is business necessity. If an employment practice which operates to exclude Negroes cannot be shown to be related to job performance, the practice is prohibit*1380ed.” [Griggs v. Duke Power Co., 401 U.S. 424, 431, 91 S.Ct. 849, 853, 28 L.Ed.2d 158 (1971).]
******
Griggs recognized that in enacting Title VII, Congress required “the removal of artificial, arbitrary, and unnecessary barriers to employment” and professional development that had historically been encountered by women and blacks as well as other minorities. 401 U.S., at 431 [91 S.Ct. at 853]. * * *
******
* * * A disparate-impact claim reflects the language of § 703(a)(2) and Congress' basic objectives in enacting that statute: “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.” 401 U.S., at 429-430 [91 S.Ct. at 853] (emphasis added). * * * In other words, § 703(a)(2) prohibits discriminatory “artificial, arbitrary, and unnecessary barriers to employment,” 401 U.S., at 431 [91 S.Ct. at 853], that “limit ... or classify ... applicants for employment ... in any way which would deprive or tend to deprive any individual of employment opportunities.” (Emphasis added.)
* * * Congress’ primary purpose was the prophylactic one of achieving equality of employment “opportunities” and removing “barriers” to such equality. Id., at 429-430 [91 S.Ct. at 852-853]. * * *
******
* * * In considering claims of disparate impact under § 703(a)(2) this Court has consistently focused on employment and promotion requirements that create a discriminatory bar to opportunities. This Court has never read § 703(a)(2) as requiring the focus to be placed instead on the overall number of minority or female applicants actually hired or promoted. * * * [Emphasis in original.]
In short, the District Court’s dismissal of respondents’ claim cannot be supported on the basis that respondents failed to establish a prima facie case of employment discrimination under the terms of § 703(a)(2). The suggestion that disparate impact should be measured only at the bottom line ignores the fact that Title VII guarantees these individual respondents the opportunity to compete equally with white workers on the basis of job-related criteria. Title VII strives to achieve equality of opportunity by rooting out “artificial, arbitrary, and unnecessary” employer-created barriers to professional development that have a discriminatory impact upon individuals. * * * [Emphasis in original.]
The district court’s findings and conclusion with respect to advancement to the MHW-II position do not adequately address the effect of the no promotion policy on the advancement of blacks to the MHW-III position. We remand to the district court for resolution of this issue and we turn now to the remaining unresolved facet of Giles’ claim.
Perpetuation of Past Discrimination
Throughout the course of this litigation, Giles has averred that the effects of past discrimination, in combination with the recent hiring of blacks by Partlow, have concentrated blacks at the MHW-I level causing the no promotion policy to have a disparate impact on blacks. The effects of past discrimination at Partlow are well documented in Frazer and in the trial court’s findings below. Blacks were concentrated in the lowest level positions at Partlow as a result of past discrimination. The hiring required by the Wyatt cases served to further concentrate blacks at the MHW-I level. The no promotion policy impaired promotional opportunities for all MHW-I’s. Consequently, although the policy was found to lack a discriminatory motive, intent, or purpose, the record establishes that the policy more heavily burdened blacks than whites.
*1381In Griggs v. Duke Power Co.,38 the Supreme Court condemned under title VII practices which, although themselves neutral, have the effect to perpetuate past discrimination.
In short, the Act does not command that any person be hired simply because he was formerly the subject of discrimination, or because he is a member of a minority group. Discriminatory preference for any group, minority or majority, is precisely and only what Congress has proscribed. What is required by Congress is the removal of artificial, arbitrary, and unnecessary barriers to employment when the barriers operate invidiously to discriminate on the basis of racial or other impermissible classification.
* * * The Act proscribes not only overt discrimination but also practices that are fair in form, but discriminatory in operation. The touchstone is business necessity. If an employment practice which operates to exclude Negroes cannot be shown to be related to job performance, the practice is prohibited.
s(s sfc s}s ;}: H:
* * * We do not suggest that either the District Court or the Court of Appeals erred in examining the employer’s intent; but good intent or absence of discriminatory intent does not redeem employment procedures or testing mechanisms that operate as “built-in headwinds” for minority groups and are unrelated to measuring job capability.
* * * But Congress directed the thrust of the Act to the consequences of employment practices, not simply the motivation. More than that, Congress has placed on the employer the burden of showing that any given requirement must have a manifest relationship to the employment in question. [Emphasis in original.]
The failure to promote would appear to operate to “freeze” blacks in the lowest mental health worker categories at Part-low, impairing their opportunity for advancement. The no promotion policy is not related to job performance and was not found by the district court to be justified on the basis of business necessity. Thus, we conclude that under Griggs the trial court’s analysis is insufficient to exonerate the challenged policy.
Conclusion
Giles has not established that the trial court abused its discretion in certifying the class under Fed.R.Civ.P. 23. Nor did the trial court err in denying Giles’ motion for summary judgment. The district court’s findings of fact with respect to the challenged practices and their effect on the class members at Partlow have not been shown to be clearly erroneous. The district court did not err in holding that Giles did not establish a claim of discrimination under title VII with respect to the underpayment of black MHW-I’s. While we cannot conclude, on the basis of the record before us, that the district court erred in failing to find the challenged no promotion policy had an adverse impact on blacks, the trial court’s analysis is not sufficient to support its finding that the policy does not have an adverse impact on blacks. The court below failed to address adequately either the impact of the policy on the opportunity of blacks for advancement at Partlow, or the potential for the policy to “freeze” the effects of past discrimination at Partlow.
Accordingly, the judgment below is affirmed in part and vacated in part, and the case is remanded to the district court for further proceedings, consistent with this opinion, to determine whether the challenged no promotion policy deprived or tended to deprive blacks of employment opportunities or whether it operated to freeze the effects of past discrimination at Partlow.
AFFIRMED IN PART, VACATED IN PART, AND REMANDED.