dissenting:
On 31 July 1980 the majority of this panel reversed a district court ruling which had granted summary judgment to defendantsappellees in this Title VII suit on grounds of appellant’s failure to make out a prima facie case of racial discrimination. The panel opinion found two errors of law in the district court’s decision, first in its statement that it was “critical” for plaintiff-appellant to show racially discriminatory motive on the part of defendant, and second in its requirement that, to make a prima facie case, plaintiff must show he was “as qualified or more qualified than individuals who were promoted or detailed.” 1
In addition, the panel majority made its own assessment of the factual issue of appellant’s qualifications. The majority looked primarily to the Supreme Court’s standard in McDonnell Douglas Corp. v. Green, which held that a Title VII plaintiff can establish a prima facie case of racial discrimination by showing:
(i) that he belongs to a racial minority; (ii) that he applied and was qualified for a job for which the employer was seeking applicants; (iii) that, despite his qualifications, he was rejected; and (iv) that, after his rejection the position remained open and the employer continued to seek *521applicants from persons of complainant’s qualifications.2
Of these criteria only the second, in its requirement that plaintiff be “qualified,” is at issue in this case. The panel majority concluded that the weight of evidence on record so clearly indicated plaintiff’s qualifications for the jobs in question that this court could hold plaintiff to be “qualified” without remanding the issue for a district court finding based on the proper standard.
I agree with the majority that the district court applied an inappropriate legal standard in ruling that it was critical for plaintiff to prove discriminatory motive on the part of defendant. Although the Supreme Court has stated that for “disparate treatment” cases of this sort proof of discriminatory motive is “critical,”3 the Court has allowed such motive to be proven indirectly in certain circumstances, by inference from “the mere fact of differences in treatment.” 4 Because the district court’s opinion appears to require direct proof of discriminatory motive, I concur with the majority insofar as they vacate the district court’s grant of summary judgment on this erroneous legal basis.
I strongly disagree, however, with the majority’s treatment of the standard for permissible indirect proof of discriminatory motive and its rationale. The Supreme Court intended its McDonnell Douglas standard to identify those situations where, based on plaintiff’s initial evidence, the employer’s decision appeared “more likely than not” to be a discriminatory one.5 For the standard to accomplish this goal, the Court acknowledged that it must be applied flexibly, since in light of variations in circumstances of Title VII suits, the specified standard “is not necessarily applicable in every respect to differing factual situations.”6
In the present case the majority has applied the second McDonnell Douglas criterion inflexibly with an irrelevant and mechanical standard of “qualification” that bears little relationship to plaintiff’s ability to perform well in the managerial positions in question. As a result the majority’s concept of a prima facie case has little if anything to do with the likelihood that defendant denied a position to plaintiff for racially discriminatory reasons, which is what the McDonnell Douglas standard is supposed to determine. Although this approach finds support in some other lower court opinions, the present case illustrates a potential for the application of the McDonnell Douglas standard to lose touch with its purpose. Far from this outcome being required by the Supreme Court’s decisions in this area, it is not consistent with the reasoning of those decisions. I therefore feel compelled to detail my disagreement with the majority’s approach and set forth a more appropriate alternative consistent with the purpose of the McDonnell Douglas test.
The four positions to which appellant sought promotion by the Washington, D. C., Post Office are managerial in nature. Appellant was already among the highest ranking officers in the D. C. Post Office when he applied for promotion to the four positions of Mail Processing Officer, Acting Mail Processing Representative, Director of the Operations Division, and Customer *522Services Representative.7 To be “qualified” for any managerial position at this level requires above all that one have managerial ability-an ability to supervise, direct, and cooperate with others; a personal sense of responsibility and initiative; an ability to motivate other persons; and a familiarity with the substantive area of activity for which one is responsible. Such traits of managerial ability are intangible and defy measurement by “objective” means using figures or paper credentials.
To assess plaintiff’s qualifications for these positions, however, the majority inexplicably did look first to paper credentials, citing appellant’s Master’s degree, course work towards a Ph.D., forty years experience in the Post Office, including supervisory positions, performance ratings not less than satisfactory, and participation in performance-enhancing courses and seminars. The majority also noted that appellant had been nominated as a candidate for D.C. Postmaster by the Regional Office (though the list recommended by the District Office did not include him).8 In addition, appellant was the second choice of the Promotion Advisory Board for two other positions, and he was promoted subsequent to the events at issue here.9
The problem with assessing qualifications based on such quantifiable, “objective” criteria is that these criteria bear very little relationship to an employer’s real-life decision about which candidate he would prefer for a managerial position. That decision must typically be based much more on a comparative assessment of the subjective managerial talents of candidates than on the basis of objective paper criteria. To say that a candidate meets certain minimal objective criteria for a managerial position says virtually nothing about whether a nondiscriminatory employer would choose him over other candidates. As a result, a judicial determination that a Title VII plaintiff is “qualified” in this sense for a managerial position says virtually nothing about whether the employer’s decision is “more likely than not” to be a racially discriminatory one which is the issue to be determined at the time tests for a prima facie case are applied.
Of course this problem is apt to occur inevitably when one assesses qualifications for a managerial position' according to an absolute test rather than a comparative test. If one looks only at the absolute qualifications of the individual isolated from his qualifications relative to other candidates, one is artificially limited to the objective, quantifiable type of criteria: even if they are not significant for real-life decisions, at least they are “something” for a court to go by.
But this dilemma occurs only if we rule out the possibility of a relative assessment of the less tangible qualifications of the various candidates for a managerial position. By a comparative technique one can assess how well one candidate has performed in the past compared to another candidate: compare the performance of organizational units for which the candidates were responsible, compare the subjective evaluations from their supervisors in the past, or perhaps compare their general reputations among those for whom they worked. This type of comparison would bear at least some relationship to employment decisions made in the real world. If a court found that a minority candidate was approximately as qualified as a nonminority candidate in this sense, the finding truly would bear some relation to the likelihood that the employer made his decision based on discriminatory racial grounds.
The district court attempted to do precisely this in requiring plaintiff to make some showing that he was “as qualified or more qualified than the individuals who *523were promoted or detailed.” 10 The majority rejects this approach for lack of authority, and it is true that appellee’s counsel was unable to cite authority for such an approach at oral argument.
But I cannot agree that the district court’s approach is “plainly at odds” with the McDonnell Douglas test.11 The McDonnell Douglas Court acknowledged the flexibility necessary if its prima facie case standard is to serve its intended purpose,12 and further disclaimed any intent to set a specific test for evaluating employment qualifications. On the contrary, the Court expressed a concern that qualifications tests “bear a demonstrable relationship to successful performance of the jobs.”13 More recently, the Court has stated that relative as well as absolute assessment of qualifications is relevant to the McDonnell Douglas standard. The Court explained that under McDonnell Douglas plaintiff proves discrimination indirectly by presenting evidence to exclude the two most common nondiscriminatory reasons for an employer's rejection of a candidate: “an absolute or relative lack of qualifications or the absence of a vacancy in the job sought.”14
In light of the purpose of the McDonnell Douglas test and the realities of employment decisions, this language should not be ignored, despite the lack of lower court cases that have followed the relative qualifications branch of this standard. The fact is that some employment decisions are made primarily by absolute, noncomparative criteria, while others depend predominantly on relative qualifications in a comparison among candidates. The former approach applies generally for positions that require objectively verifiable skills; if the applicant meets the objective standard for employment, he is “qualified.” The relative approach, on the other hand, applies to positions requiring traits that can only be evaluated subjectively, by comparison with other candidates; in this category managerial positions are the clearest example.
Under a comparative approach, plaintiff need not show that he was the most qualified candidate, only that he was approximately as qualified in light of comparative criteria. Although this is necessarily a somewhat vague standard that requires discretion on the part of the trial court, the absolute standard is also vague and discretionary when applied to cases that involve managerial positions with qualifications that cannot be stated in terms of objective criteria. The essential difference is that the relative standard is a discretionary one bearing some relation to actual hiring decisions for managerial positions, while the absolute standard is a discretionary one bearing little or no relation to actual managerial hiring decisions.
Significantly, the Supreme Court’s application of the absolute qualification approach has in fact been in the context of positions with objectively verifiable skills; hence the majority’s opinion here represents an unthinking extension of Supreme Court doctrine, an extension which is inconsistent with the doctrine’s overall purpose. McDonnell Douglas involved a plaintiff who had been fired from a job as a mechanic and laboratory technician.15 For such a job, the ability to perform the required skills can be objectively verified, far more easily at least than for a manager. Similarly, in *524Furnco Construction Corp. v. Waters,16 the Court applied the McDonnell Douglas standard in the context of bricklayers,17 for whom one could judge qualifications in objective, verifiable terms of ability and experience in performing a particular task. So long as we are dealing with nuts, bolts, and bricks, the majority approach would doubtless work well to identify discriminatory hiring practices.
To extend an absolute qualifications approach to managerial positions, however, not only defeats the purpose of the McDonnell Douglas standard, but also leads to distorted and unjust results. If courts establish minimal objective criteria for qualifications in managerial positions,18 then any supervisory employee satisfactorily performing his present job can claim to be “qualified” in this vague sense for promotion to the next higher position. The consequences are clear for any employer who contemplates promoting a superior nonminority employee over a marginally qualified minority employee: since the minority employee will automatically be able to make out a prima facie case of discrimination, it will make no sense for the employer to promote the superior candidate over the marginal candidate unless he has available sufficient resources of time and money to rebut the prima facie case in extended litigation.
The only alternative strategy for the employer would be to devise some type of “objective” qualifications test for managerial employees, which would preclude a prima facie case by showing that those passed over for managerial positions were not “qualified.” But the difficulty, not to say impossibility, of devising an objective test to identify those who would be best in a managerial position, simply highlights the absurdity of using an absolute, objective test rather than a relative one to determine managerial qualifications under McDonnell Douglas.
We recognize, of course, that comparative criteria may always come into evidence at the stage of a Title VII suit at which defendant attempts to rebut plaintiff’s prima facie case. As a matter of theory this may seem of equal value to a relative qualifications approach used at the threshold stage of assessing the sufficiency of plaintiff’s prima facie case. But as a practical matter of litigation expense, the former is likely to be significantly more costly than the latter. To impose this added expense of making an affirmative defense on a defendant, who has not even been shown more likely than not to have discriminated, is neither fair nor consistent with usual trial procedure.
It should be recognized that a relative test of qualifications would generally be more difficult to satisfy than an absolute, objective test. But this merely corresponds with the fact that racial discrimination is by nature more difficult to prove by indirect evidence for a highly subjective decision-such as the hiring of managers-than for decisions made according to objective and verifiable criteria. Courts must weigh the danger of unpunished discriminatory decisions in the hiring of managers, against the unfairness of subjecting innumerable innocent employers to the cost of overcoming a prima facie case built on thin air. By applying a relative qualifications test where appropriate, courts at least conform their decisions to the realities of hiring decisions and thus increase the probability of just results.
By contrast, the use of an absolute test of qualifications for managerial positions can easily force the burden of rebutting a prima facie case upon an employer who has never come close to making a discriminatory hiring decision. Even an employer who actively favors minority applications could come under the gun, if he just once passed over a marginal minority candidate for promotion to a managerial position in favor of a highly qualified nonminority candidate. The potential for a standard to lead to such possible injustice should give us pause.
*525In the present case we do not know whether defendant discriminated in its promotion decision. The possibility that we may be finding a prima facie case against a totally nondiseriminatory employer is suggested, however, by the district court’s findings that (1) during the period in question “blacks as well as whites were promoted or detailed to positions above plaintiff’s position,” (2) there has been a considerable increase in the number of high level black employees in the D.C. Post Office during and since the period covered by the complaint, and (3) at the present time “almost all high level positions are held by blacks.”19
On the present record it is unclear whether plaintiff was as qualified as other candidates for the managerial positions in question. I would remand to the district court for a finding on this issue based on relative as well as absolute criteria of qualification, but not requiring direct proof of discriminatory motive on the part of defendant. I therefore dissent from the majority’s entering a finding of qualification instead of remanding this issue to the district court.
ORDER
Upon consideration of the motions of both the appellant and the appellees to further extend the time within which to file petitions for rehearing and/or to suggest rehearing en bane, it is
ORDERED, by the Court, that the motions of both the appellant and the appellees are granted and the time within which parties may petition for rehearing and/or suggest rehearing en banc is granted to, and including, October 23, 1980.
Dissenting opinion from motions to grant extension of time to file rehearing filed by Circuit Judge EDWARDS.. Aikens v. United States Postal Serv., 642 F.2d 514 at 519 (D.C.Cir.1980).
. McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 93 S.Ct. 1817, 1824, 36 L.Ed.2d 668 (1973).
. International Bhd. of Teamsters v. United States, 431 U.S. 324, 335 n.15, 97 S.Ct. 1843, 1854-1855 n.15, 52 L.Ed.2d 396 (1977). The present case is an instance of alleged disparate treatment on grounds of race, rather than one of “disparate impact,” which involves “employment practices that are facially neutral in their treatment of different groups but that in fact fall more harshly on one group than another and cannot be justified by business necessity.” Id.
. Id.
. Furnco Constr. Co. v. Waters, 438 U.S. 567, 576, 98 S.Ct. 2943, 2949, 57 L.Ed.2d 957 (1978). See International Bhd. of Teamsters v. United States, 431 U.S. 324, 357-58 & n.44, 97 S.Ct. 1843, 1865-1866 & n.44, 52 L.Ed.2d 396 (1977); McDonnell Douglas Corp. v. Green, 411 U.S. 792, 801-02, 93 S.Ct. 1817, 1823-1824, 36 L.Ed.2d 668 (1973).
. McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802 n.13, 93 S.Ct. 1817, 1824 n.13, 36 L.Ed.2d 668 (1973).
. Aikens v. Bolger, No. 77-0303, Findings of Fact ¶¶ 5, 9 (D.D.C. 26 Feb. 1979).
. See id. ¶ 21.
. Aikens v. United States Postal Serv., 642 F.2d 514 at 518 (D.C.Cir.1980).
. Aikens v. Bolger, No. 77-0303, Conclusions of Law ¶ 8 (D.D.C. 26 Feb. 1979).
. Aikens v. United States Postal Serv., 642 F.2d 514 at 519 (D.C.Cir. 1980).
. See McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802 n.13, 93 S.Ct. 1817, 1824 n.13, 36 L.Ed.2d 668 (1973).
. Id. at 802 n.14, 93 S.Ct. at 1824 n.14 (quoting Griggs v. Duke Power Co., 401 U.S. 424, 431, 91 S.Ct. 849, 853, 28 L.Ed.2d 158 (1971)).
. International Bhd. of Teamsters v. United States, 431 U.S. 324, 358 n.44, 97 S.Ct. 1843, 1866 n.44, 52 L.Ed.2d 396 (1977) (emphasis added).
. McDonnell Douglas Corp. v. Green, 411 U.S. 792, 794, 93 S.Ct. 1817, 1820, 36 L.Ed.2d 668 (1973).
. 438 U.S. 567, 98 S.Ct. 2943, 57 L.Ed.2d 957 (1978).
. See id. at 569-71, 98 S.Ct. at 2945-2947.
. See, e. g., Rich v. Martin Marietta Corp., 522 F.2d 333, 347-48 (10th Cir. 1975).
. Aikens v. Bolger, No. 77-0303, Findings of Fact ¶¶ 15-16 (D.D.C. 26 Feb. 1979).