dissenting:
The lesson of the great decisions of the Supreme Court and the lesson of contemporary history have been the same for at least a generation: discrimination on the basis of race is illegal, immoral, unconstitutional, inherently wrong, and destructive of democratic society. Now this is to be unlearned and we are told that this is not a matter of fundamental principle *1352but only a matter of whose ox is gored. Those for whom racial equality was demanded are to be more equal than others. Having found support in the Constitution for equality, they now claim support for inequality under the same Constitution. Yet a racial quota derogates the human dignity and individuality of all to whom it is applied; it is invidious in principle as well as in practice. Moreover, it can as easily be turned against those it purports to help. The history of the racial quota is a history of subjugation, not beneficence. Its evil lies not in its name but in its effect; a quota is a divider of society, a creator of castes, and it is all the worse for its racial base, especially in a society desperately striving for an equality that will make race irrelevant.
Bickel, The Morality of Consent 133 (1975).
Today we direct the approval of an agreement by the City of Miami to institute racial and sexual quotas in hiring and promoting its employees. The agreement, like that recently upheld by the Supreme Court against Title VII attack in United Steelworkers and Kaiser Aluminum v. Weber,1 is “voluntary”: here the employer was inspired to adopt “voluntary” quotas by the Department of Justice, there by the EEOC and the Office of Federal Contract Compliance.2 Here, as in Weber, there is neither admission to nor finding of the maintenance by the employer of discriminatory employment practices in the past, unless the majority’s discussion of statistics at 614 F.2d p. 1338 is meant as such a finding.3 If it is, the majority has acted on incomplete data, data that ignore, among other things, age differentials among the ethnic groups of which we treat. Country-wide, the median age of Jewish-Americans is 46 years; that of Anglo-Saxons, 34; of Blacks, 22; and of Hispanics, 18.4 On these figures alone, it would scarcely be surprising to discover — in Miami or elsewhere — that Blacks and Hispanics had not yet penetrated upper-level management in numbers proportionate to their strength in the general population.
Statistics being notoriously slippery tools for amateurs such as we, a word of caution from a professional on our subject may not be amiss:
A major nonmoral, nonsocietal variable that is routinely ignored is age. As already noted, median age differences among American ethnic groups range up to decades. The median age of American Indians is only one-half that of Polish-Americans (twenty versus forty); the median age of blacks is a little less than half that of Jews (twenty-two versus forty-six). These differences affect everything from incomes and occupations to unemployment rates, fertility rates, crime rates, and death rates. For example, Cuban-Americans average a higher income than Mexican-Americans, who are a decade younger, but in the same age brackets it is the Mexican-Americans who earn *1353more. Any attempt to explain gross income differences between these two groups in terms of either discrimination by “society” or by their respective “ability” runs into the hard fact that the gross difference is the opposite of the age-specific difference. Similarly, blacks have lower death rates than whites, but this in no way indicates better living conditions or medical care for blacks, much less any ability of blacks to discriminate against whites in these respects. Blacks are simply younger than whites, and younger people have lower death rates than older people; on an age-specific basis, whites have lower death rates than blacks. Age differences also overshadow racial differences in unemployment rates: Blacks in the twenty-four to forty-four-year-old brackets have consistently had lower unemployment rates than whites under twenty — every year for decades, even though whites as a group have lower unemployment rates than blacks as a group. In short, the impact of age on statistical data is so great that to compare groups without taking age into account is like comparing apples and oranges. Yet “affirmative action” comparisons of group “representation” almost invariably ignore age differences.
T. Sowell, Knowledge & Decisions 257 (1980) (footnotes omitted and emphasis added).
Other neutral, statistical variables than age exist as well and are likewise ignored in our oversimplified comparisons.5 But the addition of this factor alone, group age differentials, belies the conclusion of past race prejudice implied in the majority’s confident assertions about “striking disparity in earnings” and statistics that speak “just as plainly for female and Spanish-named employees” as they do for blacks.6 Partial statistics always speak plainly but rarely, if ever, accurately.
But this feature of our decision is not especially remarkable, having as precedent not a few past efforts by us in the statistical line. Nor is our conclusion that Title VII, despite its language,7 does not outlaw employment discrimination on grounds of race or sex. We could scarcely conclude otherwise, given the Supreme Court’s determination in Weber that the spirit of Title VII accords with racial discrimination against whites, so long as it is reasonable in degree. Not these determinations, but another, afford today’s opinion its landmark quality.
For today’s decision sustains preferential treatment of classes based on race and sex, in the form of hiring and promotion quotas, against constitutional attack on 14th amendment grounds. As the opinion makes plain, the program of deliberate class preference that we here command is to be carried out, not in order to redress wrongs to particular individuals — to accord any given person a rightful place of which he or she has been deprived — or even to redress a discerned historic class discrimination by a particular employer, but to compensate for general societal discrimination in the past. I can ascribe no other meaning to the approving quotation of a portion of Mr. Justice Brennan’s minority opinion in Bakke at p. 1338, supra.
The majority cites a plethora of cases for the proposition that there is no “insurmountable barrier to the use of goals or quotas to eradicate the effects of past discrimination.” See p. 1335 supra (footnote omitted and emphasis added). In none of *1354these cases, however, has any circuit court approved the imposition of racial quotas based on a statistical disparity alone, as the majority does here. The majority correctly points out that this court, sitting en banc, approved racial quotas in Morrow v. Crisler, 491 F.2d 1053 (5th Cir.) (en banc), cert. denied, 419 U.S. 895, 95 S.Ct. 173, 42 L.Ed.2d 139 (1974). We did so, however, only upon the basis of a finding by the trial court8 that the Mississippi Highway Patrol had “historically engaged in unconstitutional discrimination in the employment of patrolmen,” id. at 1055, and that the patrol’s existing practices were discriminatory. Id. In fact, we specifically characterized the relief ordered there as intended to “offset the effects of past discrimination.” Id. at 1056. Assuming that Morrow retains vitality after Bakke,9 Morrow surely does not countenance the establishment of quotas on the mere basis of statistical disparities, without more; nor do either of the other cases from this circuit cited by the majority, see Franks v. Bowman Transportation Co., 495 F.2d 398 (5th Cir. 1974), modified, 424 U.S. 747, 96 S.Ct. 1251, 47 L.Ed.2d 444 (1976) (past racial discrimination had been demonstrated); Local 53 of the International Association of Heat & Frost Insulators & Asbestos Workers v. Vogler, 407 F.2d 1047 (5th Cir. 1969) (union admitted discrimination). In fact, not one of the cases enumerated by the majority at pp. 1335-1336, supra, approves a quota in the clear absence of any findings of past employment discrimination. I believe that the facts of Morrow represent the minimum threshold showing that must be made in order to support the imposition of racial quotas: namely, pervasive and established history of discriminatory employment practices by a specific employer against a particular class of individuals.
To support its quantum constitutional'^ leap, the majority interprets Mr. Justice Powell’s opinion in Bakke “to deny that specific findings of past illegal conduct are constitutionally required before a public employer can base decisions on race.” Maj. op., supra at p. 1337. This is not so: Mr. Justice Powell’s opinion makes clear beyond peradventure that the Supreme Court has “never approved preferential classifications in the absence of prove[n] constitutional or statutory violations.” 438 U.S. at 302, 98 S.Ct. at 2755 (footnote omitted and emphasis added). The fact that Justice Powell couched his discussion “in negative terms,” maj. op., supra at p. 1337 n.28, cannot obscure the fact that the majority today charts a course that the Supreme Court has never sanctioned by approving a preferential classification in the absence of a proven constitutional or statutory violation.
The majority also seeks support for its position in “Mr. Justice Powell’s citation with approval of two cases involving affirmative action remedies imposed under Executive Order 11,246, [citation omitted] without detailed findings of past discrimination.” Maj. op., supra, p. 1337. See 438 U.S. at 301, 98 S.Ct. at 2754, citing Contractors Association of Eastern Pennsylvania v. Secretary of Labor, 442 F.2d 159 (3d Cir.), cert. denied, 405 U.S. 854, 92 S.Ct. 98, 30 L.Ed.2d 95 (1971); Associated General Contractors of Massachusetts, Inc. v. Alshuler, 490 F.2d 9 (1st Cir. 1973), cert. denied, 416 U.S. 957, 94 S.Ct. 1971, 40 L.Ed.2d 275 (1974). In so doing, the majority ignores Mr. Justice Powell’s crucial assertion that “[e]very decision upholding the requirement of preferential hiring under the authority of Executive Order 11,246 has emphasized the existence of previous discrimination as a predicate for the imposition of a preferential remedy.” 438 U.S. at 301 n.40, 98 S.Ct. at 301 n.40 (citation omitted and emphasis *1355added). Here we have no finding of the existence of previous discrimination as a basis for imposing such a remedy. Mr. Justice Powell did say that “[s]uch [racial] preferences also have been upheld where a legislative or administrative body charged with the responsibility made determinations of past discrimination by the industries affected, and fashioned remedies deemed appropriate to rectify the discrimination.”10 438 U.S. at 301, 98 S.Ct. at 2754 (emphasis added). In the instant case we have been presented with no determination whatever by the Department of Justice that the City of Miami had a history of past discrimination.11 We have before us merely the fact that the Department of Justice gave its administrative approval to a particular remedy by signing the consent decree. I, for one, cannot equate the Department’s signing a consent decree with its having made a determination that a particular employer bas a history of past discrimination. If the two acts were in some way equivalent, the employer by signing the consent decree would in a sense be admitting a past history of discrimination, despite the express provision in the consent decree that its entry “shall not constitute an adjudication or an admission by any of the defendants of any violation of law.” For the majority to conclude that “[n]othing more is constitutionally required by the opinion of Mr. Justice Powell,” maj. op., supra at p. 1338, than the Department of Justice’s signature on the consent decree requires a tortured and, I fear, erroneous interpretation of Mr. Justice Powell’s opinion in Bakke that I cannot endorse.
Perhaps even more troubling to me is the majority’s suggestion that a court can impose quotas to rectify the effects of societal discrimination. Maj. op., supra at p. 1338. *1356This is precisely the course Mr. Justice Marshall would have us follow, see Bakke, 438 U.S. 265, 387, 98 S.Ct. 2733, 2798, 57 L.Ed.2d 750 et seq. (Marshall, J., dissenting), and it is implicit in the language of Mr. Justice Brennan cited by the majority at p. 1338. It is clear beyond peradventure, however, that the Supreme Court has not been willing to impose quotas to redress societal discrimination against a particular class. If societal discrimination were the litmus for the application of quotas, racial quotas could be imposed upon virtually any employer and against his “majority” employees, since discrimination, sadly, has been pervasive in our society’s past. See 438 U.S. at 387, 98 S.Ct. 2798 et seq. (Marshall, J., dissenting). Even to hint that past discrimination by society in general affords federal courts a license to impose quotas against particular employers, as I believe the majority does, is an action that I cannot condone and that the Supreme Court refused to take in Bakke, thus provoking Mr. Justice Marshall’s dissent.
With today’s decision, the wheel appears to turn full circle to the one-discredited reasoning of Plessy v. Ferguson, 163 U.S. 537, 16 S.Ct. 1138, 41 L.Ed.2d 256 (1896).12 Plessy found in the Constitution permission for deliberate legislative discrimination against blacks as a class; here we conclude that it permits planned and purposeful discrimination against whites and males because they are involuntary members of a class preferred unjustly in the past. Thus, the momentum developed in abandoning one injustice carries us into still another, its opposite. We lay it down that two constitutional wrongs do indeed make a right.
Like my Brothers, I respectfully confess that I cannot discern the Court’s precise position on this difficult subject from the opinions in University of California Regents v. Bakke, 438 U.S. 265, 98 S.Ct. 2733, 57 L.Ed.2d 750 (1978). I can, however, say with certainty that the Court was then — by the margin of one vote — unready to write into the Constitution again, as it once did in Plessy and as the majority does today, a license for governmentally engineered race discrimination when that seemed desirable as a matter of contemporary policy. With the Chief Justice,13 I confess that such a policy of uplifting the downtrodden holds deep, short-run appeal for me. If I were an elected official, holding a valid policy warrant, and if my views of the nature of the Constitution permitted me to do so, I might well endorse what we here do — though with grave misgivings about' the deliberate injustice occasioned to persons who did not choose to be born male or white. I am not a legislator, however, and broad policy is therefore not my domain; my concern is the Constitution.
Those like myself, who have not felt the lash of race or sex prejudice, can never know how deeply it cuts, and I am very far from weighing such wrongs and indignities lightly. But even bearing in mind the cynic’s gibe at the majestic evenhandedness with which the law forbids both rich and poor to sleep under bridges, I still cannot bring myself to believe that in the long haul the interests of our people are well served by constitutional decisions that recoil from one class injustice to another. I therefore would refuse, in this confused and confusing case, with the ink hardly dry on Bakke and Weber, to take such a constitutional giant step. Henceforth, if I read the majority correctly, employers — public and private — become liable to governmental pressure, and their “majority”14 employees to government-induced race and sex discrimi*1357nation, whenever a statistical “imbalance” is perceived, pressure unrestrained by any constitutional trammel except whatever a court may from time to time view as an unreasonable degree of racial or sexual discrimination. And so, in our circuit, “majority” and “minority” are to become words of art east free entirely from that numerical reference by which I had thought their use derived its main force and equity: a special concern for groups of citizens who are unfairly locked in disadvantaged positions by permanent and irrelevant common characteristics and who cannot, because of their numerical weakness, break out via the electoral process.15
I fully recognize that today’s holding does not want for either reason or redeeming features. As for reason, it well may be that when Title VII’s prohibitions against race discrimination — with their infinitely more explicit language and unambiguous history — were overthrown in Weber, it was ordained that those of the fourteenth amendment could not long survive. Nor is the majority rule freighted with such discredited trappings of racial prejudice or assumed class inferiority as informed the earlier “wisdom”: far from it; all that the present constitutional ruling countenances is straight class discrimination, not class prejudice. Finally, by the lights of our day, *1358the purpose of the holding is at least as well intended as was Plessy, which bowed in its own day to what the Court doubtless viewed as current realities.
But I, for one, do not find those lights clear enough for such work as rechiselling a license for class discrimination in constitutional stone. The Supreme Court pulled up at that brink in Bakke. Today’s majority goes on, but without me. I respectfully dissent.16
. 443 U.S. 193, 99 S.Ct. 2721, 61 L.Ed.2d 480 (1979).
. See Meltzer, The Weber Case, 3 Regulation 34, 42 (1979).
. And here, as in Weber, the appeal to reality in foregoing such findings is made: it is urged that to require such a finding or admission would be to subject the employer to “untold consequences.” Concurrence of Judge Thorn-berry, 614 F.2d 1351. Overlooked by that argument, however, are the “untold consequences” of permitting the imposition of racial quotas without such a finding for the employer’s “majority” employees, see n.14, infra, whose rights are bartered away by the employer in order to buy his private peace. In a head-to-head tort or contract case the rights of third parties are not implicated; the parties may bargain away their own rights for whatever reasons they may choose. In contrast, the consent decree at issue here profoundly affects the constitutional rights of third parties — the employer’s non-Black, non-Hispanic, or non-female employees — to be free from discrimination on account of their race or gender. See nn.14-16 and accompanying text, infra. If the city is to be allowed to evade untoward consequences to itself by visiting constitutional infringements on its white male employees, surely simple justice requires — at the barest minimum — that first it be shown that they, or at least their class, have been unfairly favored by the city in the past.
. United States Bureau of the Census figures, noted in Sowell, Myths About Minorities, 68 Commentary, No. 2, p. 33 (1979).
. For example, the statistics failed to account for educational qualification for positions or for job experience.
. The majority notes with apparent satisfaction that its recited statistics have not been challenged. P. 1338. That is hardly surprising as there has been no trial at which to mount a challenge. Doubtless they are correct, as far as they go — which is not far enough to prove what the majority seems to believe they do.
. The central provision of that statute makes it unlawful for an employer “to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s race, color, religion, sex, or national origin . . 42 U.S.C. § 2000e-2.
. See Morrow v. Crisler, 479 F.2d 960, 962 (5th Cir. 1973) (panel opinion). “The evidence presented in this case amply supports the District Court’s conclusion that the policies and practices of the defendants constitute a pattern and practice of racial discrimination in hiring and employment in violation of the Fourteenth Amendment . . Indeed it did: at the time of the trial court decree, the Mississippi Highway Patrol had never in its history had a black patrolman.
. If Bakke overruled Morrow sub silentio, the majority’s approving citation of Morrow is without significance.
. In Contractors Association of Eastern Pennsylvania, supra, the Assistant Secretary of Labor had issued an order containing
findings that although overall minority group representation in the construction industry in the five-county Philadelphia area was thirty percent, in the six trades representation was approximately one percent. It found, moreover, that this obvious underrepresentation was due to the exclusionary practices of the unions representing the six trades. It is the practice of building contractors to rely on union hiring halls as the prime source for employees.
442 F.2d at 173 (emphasis added). Although findings of this ilk are at least one step removed from findings of past discriminatory conduct on the part of particular employers as in Morrow, there are nevertheless executive findings directly linking the hiring procedures of a group of employers to discriminatory union practices. And while I might prefer more before approving a quota remedy, see, e. g., Morrow v. Crisler, supra, I cannot abide less. Nor can Justice Powell, as I read his opinion in Bakke. No even remotely comparable findings to those in Contractors Association of Eastern Pennsylvania were made in the instant case; here there have been no “legislative or administrative determinations of past discrimination . . 438 U.S. at 301, 98 S.Ct. at 2754.
In Associated General Contractors, supra, the district court used statistics of racial disparity “as sufficient evidence to support an inference of racial discrimination.” 361 F.Supp. 1293, 1299 (D.Mass.) (emphasis added), aff'd, 490 F.2d 9 (1st Cir. 1973), cert. denied, 416 U.S. 957, 94 S.Ct. 1971, 40 L.Ed.2d 275 (1974). That inference, however, was not overcome “by any contrary evidence.” Id. The court concluded that racial imbalance in the Boston construction trades was “the result of past discriminatory practices on the part of many ‘entities’ in that industry.” 490 F.2d at 18 n.15. Thus, there was a trial-level, judicial determination of past discrimination predicated on an unrebutted inference. In the instant case, to the contrary, we have no judicial determination; we have at best only an inference of discriminatory practices made by ourselves, on demonstrably insufficient data. By signing a consent decree, the parties have chosen not to enter a forum in which the City of Miami could rebut that inference. I do not see how that decision operates to transform an inference into a factual finding. If a naked inference can give rise to the imposition of quotas, then Mr. Justice Powell’s careful analysis is drained of its substance and rendered meaningless.
. Such a unilateral determination by one party litigant would have an unsettling, boot-strap quality, even had one been made; and it may be doubted whether the Department, acting generally, constitutes such a “legislative or administrative body charged with the responsibility” of making determinations of discrimination in the employment context as Justice Powell had in mind in the passage quoted in text above. His reference is to the Department’s special responsibilities under section 5 of the Voting Rights Act. 438 U.S. at 302 n.41, 98 S.Ct. at 2755 n. 41.
. See Van Alstyne, Rites of Passage: Race, the Supreme Court, and the Constitution, 46 U.Chi.L.Rev. 775 (1979).
. Weber, 443 U.S. at 216, 99 S.Ct. at 2734, 61 L.Ed.2d at 497.
. To state the matter so makes plain that what in fact we do here is to require class discrimination in favor of two minorities (Blacks and Hispanics) and a majority (women) against another minority (white males). Thus we do what Mr. Justice Powell clearly said we could not do in Bakke :
[T]he difficulties entailed in varying the level of judicial review according to a perceived “preferred” status of a particular racial or ethnic minority are intractable. The concepts of “majority” and “minority” necessarily reflect temporary arrangements and polit*1357ical judgments. As observed above, the white “majority” itself is composed of various minority groups, most of which can lay claim to a history of prior discrimination at the hands of the state and private individuals. Not all of these groups can receive preferential treatment and corresponding judicial tolerance of distinctions drawn in terms of race and nationality, for then the only “majority” left would be a new minority of white Anglo-Saxon Protestants.
438 U.S. at 295-96, 98 S.Ct. at 2751.
. But cf. Frontiero v. Richardson, 411 U.S. 677, 686, 93 S.Ct. 1764, 1770, 36 L.Ed.2d 583 (1973), (plurality opinion), lending credence to the view that women, though a majority, are to be viewed as a “minority” because “vastly underrepresented” in governmental positions of power. The calculus suggested is one infinitely complicated by the circumstance that, in our society, power and influence are exerted in myriad ways, many of great subtlety. Such a case-by-case balancing of results, dependent upon ad hoc determinations of what groupings in our society command the most power from time to time, seems to me an exercise for which judges are ill fitted.
And if any group, regardless of its numbers, can be a “minority,” competition for the favors attending that status seems well-nigh inevitable, with those who have officially received it fighting off aspirants to it and struggling to preserve their status as such. Indeed, it appears to be in progress at the present time. See Van Alstyne, op. cit. n.12, supra, at 805-807, quoting from The Washington Post of May 22, 1979:
A highly unusual and controversial proposal under consideration by a city agency would eliminate a large number of local ethnic groups from eligibility for lucrative minority contracts with the District of Columbia government.
The proposal, by the staff of the Minority Business Opportunity Commission, would eliminate from minority status all Hispanics bom in Europe and South and Central America, leaving only those Hispanics bom in Mexico or Puerto Rico eligible for the contracts.
The proposal also would eliminate from minority status persons bom in Vietnam, India, China, Korea and Africa.
******
“Oh, my God,” one influential Hispanic leader with ties to the Barry administration said when informed of the staff proposal. “If Marion agrees to do something like that, then it’s really going to be a political war between Hispanics and blacks, and that would be terrible.”
Carlos Rosario, acting director of the Office of Latino Affairs for the District, said he was not familiar with the proposal. But, he said, “I don’t see why if they have Spanish heritage why they should lose their [eligibility for minority contracts]. If they are bom minorities, they should have the same breaks as others.”
See also The Wall Street Journal, March 21, 1980, at 16, col. 2:
In the department of racial and ethnic privileges, “minority group” members get special aid from the Small Business Administration. The legislation was written to help blacks and Hispanics, and now also covers American Indians and “Asian Pacific Americans.” We hear the Pakistani-Americans are about to have a go at it, but the program is in a tizzy because a group of Hasidic Jews has just petitioned the SBA to be counted as a minority.
While the Hasidim can point to all the usual kinds of disadvantages, their petition is opposed by the author of the legislation, Rep. Parren J. Mitchell. “I might be sympathetic toward their economic conditions,” Rep. Mitchell allows, but to put the philosophical point, “there’s a limited pie.”
. At the conclusion of my writing, I have been given my Brother Thomberry’s concurring remarks. It may well be possible, as he observes, to place a different sense than mine upon some of what we say here, but there can be no doubt whatever about what we do. Nor am I so much taken as he by the circumstance that we deal with a consent decree, where the “consent” is a compelled one, and the agreement is of that special kind by which A and B agree together about the constitutional rights of C and D. I can see no reason to give our imprimatur to this particular constitutional violation when we would doubtless strike down a proposed agreement to disregard any other provision of the Bill of Rights.