Hammon v. Barry

Opinion for the Court filed by Circuit Judge STARR. Dissenting opinion filed by Circuit Judge MIKVA. STARR, Circuit Judge: These consolidated cases come before us on appeal from the District Court’s judgment upholding the hiring provisions of an affirmative action plan developed by the District of Columbia Fire Department pursuant to a consent decree, 606 F.Supp. 1082. The United States challenges the race-conscious hiring provisions of the plan on the grounds that they violate Title VII of the Civil Rights Act of 1964 and the equal protection component of the Due Process Clause of the Fifth Amendment. We conclude that, because the District of Columbia has failed to establish the requisite predicate of discrimination and did not consider, let alone employ, alternatives to its race-preference plan, the hiring provisions cannot stand. I A The genesis of these cases can be traced to October 1980, when two black D.C. Firefighters, Messrs. Holmes and Sheffield, filed complaints with the District of Columbia Office of Human Rights (OHR) alleging that the D.C. Fire Department had, beginning in October 1979, acted in a racially discriminatory manner. The alleged unlawful acts included failing to give black residents notice and information about vacancies in the Fire Department and, more relevantly for our purposes, using unvalidated tests for hiring of entry-level firefighters. In June 1981, without the benefit of an evidentiary hearing, the Director of OHR issued a “Summary Determination” finding a prima facie case of discrimination in the Fire Department’s recruitment, hiring, and promotion practices and in its implementation of policies. In August 1981, the OHR Director ordered that 60 of the next 70 entry-level firefighter positions be filled with minorities, with entrance-on-duty [52] dates retroactive to April 27,1981, the date when 23 firefighters, only two of whom were black, were hired. Noting the District’s admission that it was “not in compliance with D.C. Law 1-63, as it related to the racial make up [of] the work force of the District of Columbia,” 1 the Director mandated “that minorities shall be given preference in all new hiring and promotions until the traditional patterns of racial segregation have been eliminated; minorities within the Department are no longer under-represented at all grade levels; and minorities employed by the Department approximate the percentage of minorities in the workforce for the District of Columbia.” Defendant’s Appendix (“D.A.”) at 34 (emphasis added). The District appealed from this ruling. In due course, the designated Appellate Officer issued a supplanting order remanding the Holmes-Sheffield complaints on grounds of incomplete consideration by OHR. The Appellate Officer directed that, in view of the seriousness of the allegations, a public evidentiary proceeding should be conducted. A lengthy set of adjudicatory hearings thereafter ensued before a Hearing Examiner. B Throughout the OHR administrative proceedings, the District adamantly denied any discrimination whatsoever. Indeed, the District’s position as pertinent to our purposes. was simple and straightforward: that its entry-level examination, Test 21, which had been employed throughout the 1970’s, was valid. The District emphasized that Test 21 had successfully run the validation gauntlet both in 1971 (OHR) and 1979 (0PM),2 had passed constitutional muster in Washington v. Davis, 426 U.S. 229, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976), and that the pass-fail level of the new examination that replaced Test 21, the Firefighters Service Test (FST), had been set at a modest level. Specifically, the passing mark for the new examination was no more than the “score one could expect to receive if one answered the questions randomly.” 3 Proposed Findings, supra note 2, II19, at 16. The District further maintained that, save for the hiring of the 23 entry-level firefighters early in 1981, the Department’s entry-level test and other hiring procedures had no adverse impact on blacks.4 Id. U 27, at 19. [53] After 50 days of trial, the OHR Hearing Examiner rejected the lion’s share of the Hammon-Sheffield allegations. He noted specifically that “the complainants failed to introduce any evidence in the record to show the job qualifications for the entry-level firefighters job either intentionally discriminated or had an adverse impact against Blacks.” Opinion of Hearing Examiner, Patrick Kelly (“HE Opinion”) ¶ 23, D.A. at 44. And he further found “no evidenciary [sic] basis ... in the record” for the allegation that the District had discriminated in its recruitment efforts.” Id. 1T1Í 27, 28, 47. D.A. at 45-46.5 The Hearing Examiner did, however, find what he deemed to be two blemishes on the Fire Department’s record. Of particular relevance to the issues now before us, the Department’s racial composition, by its own admission, failed to reflect the racial composition of the adult population of the District of Columbia.6 Id. ¶ 153. The significance of this perceived shortcoming, as we shall presently examine in greater detail, was that the Fire Department was not in conformity with the strictures of applicable D.C. law requiring racial balance in all employment categories. See supra note 1. In addition, the Hearing Examiner concluded that the Fire Department’s entry-level examination, coupled with the Department’s pre-hiring investigative processes, had an adverse impact on minorities in the initial hiring effected in early 1981. Id. 119, D.A. at 68. Because of the importance of the Department’s validation and use of its entry-level examination to the issues presented by this appeal, we pause , to observe that history, as recounted in the unchallenged and exhaustive opinion of the OHR Hearing Examiner. C Prior to 1980, entry-level testing for D.C. Fire Fighters was not the responsibility of the District of Columbia at all. Rather, that function was carried on by the federal government, initially by the old Civil Service Commission and subsequently (after the landmark 1978 legislation) by the Office of Personnel Management. Over the years, OPM (and its predecessor) had administered Test 21, the pencil-and-paper test that had successfully withstood constitutional challenge in Washington v. Davis, 426 U.S. 229, 96 S.Ct. 2040, 48 L.Ed.2d [54] 597.7 Because black applicants tended to cluster near the passing score of Test 21, rank ordering caused white applicants to be “inserted on the registers over Blacks.” H.E. Opinion 1129, D.A. at 46. But, rank ordering was not in fact employed in the Fire Department’s hiring process during the reign of Test 21. Rather, at the request of the Fire Department itself, OPM (and its predecessor) changed the process “in the early 1970 [sic] and subsequently every register was exhausted before a new examination would be administered,” id., D.A. at 47. According to the Fire Department, this change was sought specifically in order to increase minority representation in the D.C. Fire Fighter ranks. Id. At the dawn of the 1980’s, Test 21 was replaced with the Firefighters Service Test. According to the OHR findings, OPM and the Fire Department developed the FST “in order to have a test that had better predictability of job performance than Test 21.” Id. ¶ 30, D.A. at 47. The FST was developed over a five-year period, during which two OPM officials, Dr. Paul van Rijn and Ms. Sandra S. Payne, conducted a Criterion Related Validity Study.8 In addition to this considerable period of gestation, the new exam was developed in compliance with the EEOC’s Proposed Uniform Guidelines, which were concurrently in formation at that agency. In 1980, OPM turned over the FST to the Fire Department for the latter’s use without making “any recommendation to [the District of Columbia Office of Personnel] concerning ... adverse impact ... because OPM believed the job relatedness of the job analysis and criterion measures would not only validate use of the FST, but would also validate its use as a rank ordering advice [sic] in the selection procedure.” Id. 1133, D.A. at 48 (emphasis added). Despite OPM’s position, considerable evidence was adduced at the OHR hearings to the effect that the FST was not valid as a rank ordering device (as opposed to pass/fail). Id. H 36, D.A. at 49 (review performed in March 1981 by Personnel Decisions Institute); id. 1137-62, D.A. at 49-54 (study performed by Dr. Richard S. Barrett). Significant among the perceived flaws in the OPM validation effort was its failure to get at the actual job requirements of the firefighter mission, which, it was argued, called for the test formulator “to go out and take a look and see what actually happens” on the job. Id. 1143, D.A. at 51. More fundamentally, the District’s expert, Dr. Barrett, “questioned whether it was possible to develop a valid multiple choice test to measure a Firefighter Job.” Id. f 57, D.A. at 53. D Based on this testimony and other evidence of record, the Hearing Examiner, notwithstanding the professional effort devoted to developing and validating the FST and its undisputed pedigree in the halls of the federal personnel establishment, concluded broadly that rank-order use of the exam results should not be employed by the Department without satisfying the following standard: A rank order system should be used only if there is convincing evidence that the employer benefits by using a rank order system. Id. 1163, D.A. at 54. The “adverse impact” on blacks of employing a rank-order system had been realized, the Hearing Examiner concluded, in [55] November 1980 when the results of the FST’s inaugural administration were tallied. At that time, the D.C. Office of Personnel developed a “Certificate of Eligibles” containing the names of the individuals attaining the top 100 scores on the examination. Although 74.35 percent of the test takers were black, and only 21.5 percent were white, the top-ranking Certificate included the following racial composition: 72 — Whites 27 — Blacks 2 — Hispanics Id. H 80, D.A. at 58-59. With this Certificate in hand (which, it will be noted, consisted of 101, not 100, applicants), the Fire Department processed the prospective employees by conducting a background check. That check, which it appears ordinarily took longer to complete for black applicants, was carried out without written operating procedures by officials with no prior experience in background checks. Id. 1182, D.A. at 59. Although the Hearing Examiner found no overt act of discrimination in this respect, he considered the process itself, which resulted in a disproportionate number of white hires in the initial 1981 hiring,9 to be an unvalidated component of the selection procedure. Id. 1183, D.A. at 60. These two perceived flaws in the hiring process were the sole problem areas identified by OHR after the lengthy and thorough process of adjudicatory hearings. Accordingly, the Hearing Examiner limited his recommendation at the conclusion of the 50-day trial to four: first, that the D.C. Office of Personnel “properly validate the [FST] in accordance with the [EEOC’s] Uniform Guidelines on Employee Selection;” second, that the D.C. Office of Personnel and the Fire Department establish written processing procedures to ensure that “minority applicants are processed through the background investigation and medical examination components of the Selection Process and presented to the Appointment Board at the same rate as White applicants”; third, that the “date of entry be the same for all applicants who are selected as Firefighters from the 1980 [examination] regardless of the actual date of entry” 10 and, critically, that “the selection list of applicants developed as a result of the 1980 examination shall remain in effect until exhausted”; and fourth, that the Fire Department submit to OHR; an “Affirmative Action Plan [“AAP”] which will eliminate racial disparity within the Department.” Id. 11111-4, D.A. at 69-70.11 The Examiner’s fourth recommendation was the one destined to figure large in the events ahead, including the present litigation. The Hearing Examiner’s recommendations were affirmed in all material respects by the designated Appellate Officer, and a Final Order was duly issued in November 1983. The District was thus exonerated of all charges of discrimination, save for (1) adverse impact if rank-ordering use were made of the FST results12 and (2) the use [56] of unstandardized processing procedures. Notwithstanding this virtually clean bill of health, the District was charged with fashioning an AAP going beyond erasing the two blemishes to the more sweeping goal of eliminating racial “imbalance” in the Department. The District was also directed to exhaust completely the 1980 list of eligibles and to give all firefighters appointed from the 1980 list the same date of hire. No review was sought by any party in the District of Columbia courts, thereby rendering the November 1983 Order the final resolution of the Holmes-Sheffield cases. E As one might suspect, the Fire Department’s needs for new firefighters continued during the foregoing, somewhat protracted proceedings. Accordingly, hiring of entry-level firefighters continued during and after the OHR proceedings that we have just chronicled. The results of that hiring process, which we believe considerably worthy of note, are as follows: in 1982, 67.5 percent of new hires were black; in 1983, 80.5 percent were black; and in 1984, 78.6 percent were black. In sum, the evidence of discrimination in the record was as follows: The Hearing Officer, after comprehensive adjudicatory proceedings, found that, as of early 1982, there was “insufficient factural [sic] evidence provided in the record to show that the job qualifications for the entry level firefighters intentionally discriminated or had an adverse impact against Blacks who met the minimum qualifications____” H.E. Opinion 114, D.A. at 65. The Hearing Officer found that “[bjased upon the evidence provided in this record, there is insufficient evidence to support Complainant’s [sic] claim that Respondents have discriminated against Blacks and District residents in their recruiting efforts____” Id. ¶ 5, D.A. at 65. Indeed, the Hearing Examiner found that the District had conducted an aggressive recruiting campaign directed at blacks. Id. 111124-28, D.A. at 44-46. Since the Hearing Examiner’s unchallenged findings were issued, the District has stepped up its hiring of blacks. F Notwithstanding this evidence of propriety in the District's hiring practices, the Hammon plaintiffs filed a discrimination suit in federal district court in March 1984. Hammon v. Barry, C.A. No. 84-0903 (D.D.C. filed Mar. 22, 1984). The catalyst for this suit was the District’s then-imminent plan to administer the FST once again, occasioned apparently by the near exhaustion of the 1980 list of eligibles. The complaint sought, among other things, “an Order requiring defendants to develop an affirmative action plan to eliminate racial disparity within the District of Columbia Fire Department,” Prayer for Relief 114, see also 11118-9, and an order prohibiting the. District from pursuing its plan to readminister the FST in late March 1984 or using that exam in the selection of new firefighters. Id. 1110; see also Complaint 111, 34, 35. On March 23, 1984, the District Court denied plaintiffs’ request for a temporary restraining order against readministering the FST.13 Two months later, on May 23, 1984, the court entered a consent decree. In that decree, the District agreed to validate an entry-level firefighter examination and other components of the selection process in accordance with the EEOC’s Uniform Guidelines, and to submit a proposed affirmative action plan to the court and OHR. In addition, the decree authorized the District to hire from the 1984 examination list only after exhausting the 1980 eligible list. Of especial note, paragraph 11 made clear that the decree was “neither an admission nor a finding that the [District] defendants have violated any law or regulation regarding prohibited discrimination.” In February 1985, the District submitted to the court its Affirmative Action Plan. The Plan addressed both hiring and pro [57] motions.14 Alarmed by the racial quotas in promotions imposed by the. AAP, eight white incumbent firefighters (the Byrne plaintiffs) and their union (Local 36) filed suit challenging the Plan’s promotion provisions as violative of Title VII and the equal protection component of the Fifth Amendment’s Due Process Clause.15 Byrne v. Coleman, C.A. No. 85-0872 (D.D.C. filed Mar. 8,1985). Three days later, the United States likewise filed suit challenging both the hiring and promotion provisions of the plan. United States v. District of Columbia, C.A. No. 85-0797 (D.D.C.). The three cases were consolidated, and the parties filed cross-motions for summary judgment on the basis of stipulated facts. On April 1,1985, after a one-day hearing, the District Court filed an opinion and order upholding the hiring provisions, but invalidating the promotion provisions as violative of Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e-2(a) (1982).16 Hammon v. Barry, 606 F.Supp. 1082 (1985). In sustaining the hiring provisions under Title VII, the court relied on the factors governing a plan’s permissibility identified in United States Steelworkers v. Weber, 443 U.S. 193, 99 S.Ct. 2721, 61 L.Ed.2d 480 (1979). Specifically, the court determined that the plan would not unnecessarily trammel the interests of whites because they had little if any expectation or entitlement to a job with the Fire Department; that the plan was sufficiently temporary in nature (eighteen months); and that although alternatives to the race-preference plan were apparently not considered, none suggested at the one-day hearing appeared to be feasible. Hammon, 606 F.Supp. at 1092-94. The court also emphasized that the AAP’s hiring provisions were “designed to break down an old pattern of racial segregation and hierarchy.” Id. at 1090. To reach this conclusion, the court looked back a generation ago to the early 1950’s when the Fire Department was officially segregated; in addition, the court recited facts as to (1) the existing percentages of blacks and whites in the work force, and (2) the current percentages of black and white applicants. It concluded from those two items of evidence that “vestiges” of discrimination remained. Id. at 1086-87, 1092. The court also upheld the hiring provisions against constitutional attack. Observing that “the ‘interest in ameliorating, or eliminating where feasible, the disabling effects of identified discrimination’ ” constitutes a compelling governmental interest, the trial court concluded that OHR’s “finding that the entry level test adversely affected black applicants” created “a sufficient predicate for the implementation of race-conscious affirmative relief.” Id. at 1095 (quoting Regents of the University of California v. Bakke, 438 U.S. 265, 298, 98 S.Ct. 2733, 2752, 57 L.Ed.2d 750 (1978)). The court then determined that the plan was “sufficiently narrowly tailored” because “the hiring provisions of the Fire Department’s AAP ask white applicants to shoulder only a minor burden.” Id. at 1096. Both the District and the United States appealed. On September 5,1985, this court dismissed at the District’s instance its appeal on the promotions issue and denied the Hammon appellees’ motion to dismiss the United States’ appeal on the hiring issue. The promotions provisions of the Affirmative Action Plan thus stand condemned; it is only the hiring provisions that are now before us. [58] II The United States’ challenge to the race-preferential hiring provisions raises both statutory and constitutional issues. The United States, supported by the Byrne plaintiffs (the eight white firefighters), charge that the hiring remedy sanctioned by the District Court violates both Title VII of the Civil Rights Act of 1964 and the equal protection component of the Due Process Clause. We observe preliminarily that the decisional law is infected with imprecision as to the differences between statutory and constitutional analysis (and, indeed, as to the proper mode of analysis itself) in the public employer setting. In the years since the three landmark affirmative action cases of the 1970’s, to which we shall presently turn, the Supreme Court has faced both Title VII and Equal Protection questions only in Local 28, Sheet Metal Workers v. EEOC, — U.S. -, 106 S.Ct. 3019, 92 L.Ed.2d 344 (1986).17 There, the Court stated that in a Title VII case, “a court should consider whether affirmative action is necessary to remedy past discrimination in a particular case before imposing such measures, and ... the court should also take care to tailor its orders to fit the nature of the violation it seeks to correct.” Id. at 3050. In constitutional analysis, it is clear that the Court requires that a race-conscious remedy (1) be designed to achieve a “compelling” or “significant” governmental interest, and (2) be properly tailored to meet that interest. The rigor with which a court is obliged to scrutinize the government’s professed interest and the remedy’s relationship to that interest, however, has not yet been agreed upon by a clear majority of the Court. See, e.g., id. at 3052; Wygant v. Jackson Board of Education, 476 U.S. 267, 106 S.Ct. 1842, 1846, 1857, 1861, 1867, 90 L.Ed.2d 260 (1986). It would appear that, as to a public employer, the general analytic framework is quite similar at the constitutional and statutory levels. In addition, certain other basic principles immediately suggest themselves. First, Congress obviously did not contemplate that Title VII would permit that which the Constitution forbids. Such a state of affairs would obviously be entirely repugnant to basic principles of constitutionalism and the supremacy of the higher law of the Constitution. Second, under well-established principles of restraint, courts should address questions of statutory propriety before turning to the gravely sensitive task of measuring the lawfulness of a public body’s action against the demands of the Constitution itself. See Ashwander v. Tennessee Valley Authority, 297 U.S. 288, 346-48, 56 S.Ct. 466, 482-84, 80 L.Ed. 688 (“ ‘It is not the habit of the Court to decide questions of a constitutional nature unless absolutely necessary to a decision of the case.’ ” (quoting Burton v. United States, 196 U.S. 283, 295, 25 S.Ct. 243, 245, 49 L.Ed. 482 (1905))). With these principles in mind, we set forth our analysis of the racially preferential hiring provisions under the rubric of Title VII. However, we do not feel it improper in this difficult area to seek guidance, as appropriate, from pertinent principles articulated by the Supreme Court in. the constitutional setting. And, as to one issue, we are constrained to look to the demands ordained by the Constitution itself. Ill A It is elementary in our law that a predicate of discrimination must exist be [59] fore race-conscious measures may lawfully be employed voluntarily or imposed by court decree. The decisional law uniformly holds that a record of discrimination constitutes the crucial condition precedent to the use of racial preferences. As chronicled below, it is established beyond the slightest doubt that this predicate is indispensable to the use, as here, of nonvictim-specific forms of racially based remedies. We begin with the three affirmative action cases handed down by the Supreme Court last Term. In all three cases, the members of the Court emphasized time and again the need to justify remedies that classify individuals on the basis of the immutable characteristic of race. Such remedies, we have been taught, are not to be invoked lightly; to the contrary, racial preferences are to be employed cautiously and reluctantly. See Local 28, Sheet Metal Workers v. EEOC, 106 S.Ct. at 3050 & n. 48. In Sheet Metal Workers, for example, Justice Brennan, speaking for himself and Justices Marshall, Blackmun, and Stevens, put the point quite succinctly in holding that a court may order, “in appropriate circumstances, affirmative race conscious relief as a remedy for past discrimination.” Id. at 3034 (emphasis added). The same theme is prominently featured in Justice Brennan’s opinion for the Court in the Vanguards case, Local No. 93 v. City of Cleveland, — U.S.-, 106 S.Ct. 3063, 92 L.Ed.2d 405 (1986). There, in holding that Title VII does not preclude the entry of a consent decree that provides relief to persons who are not victims of discriminatory practices, the Court observed that the District Court had found, based on evidence of recordman historical pattern of racial discrimination in promotions within the Cleveland Fire Department. Id. 106 S.Ct. at 3070. So too, the Sixth Circuit on appeal in that case had found “such relief justified by the statistical evidence presented to the District Court and the City’s express admission that it had engaged in discrimination.” Id. (emphasis added). Emphasizing the importance of voluntary compliance with the vital objectives of Title VII, a theme first sounded seven years previously in Weber, the Vanguards Court upheld the power of federal courts to sanction a race-conscious remedy where the parties to the consent decree are “seeking to eradicate race discrimination.” Id. at 3070. Six weeks prior to Sheet Metal Workers and Vanguards, the Court handed down the Wygant decision, where Justice Powell, writing for the plurality of Chief Justice Burger, then-Justice Rehnquist and Justice O’Connor, stated: This Court never has held that societal discrimination alone is sufficient to justify a racial classification. Rather, the Court has insisted upon some showing of prior discrimination by the governmental unit involved before allowing limited use of racial classifications in order to remedy such discrimination. 106 S.Ct. at 1847 (emphasis added). This bedrock requirement of our law was not challenged by the dissenting Justices in Wygant, save one (and that apparently only in the particular setting of education).18 Indeed, Justice Marshall, writing in dissent for himself and Justices Brennan and Blackmun in Wygant, decried the absence of a record and the efforts of the parties “to try their case before this Court.” Id. at 1858. The absence of record evidence led the three dissenting Justices to complain that “the plurality has too quickly assumed the absence of a legitimate factual predicate, even under the plurality’s own view, for affirmative action in the Jackson schools.” Id. We acknowl [60] edge the obvious ambiguity in that carefully crafted statement, but we observe that the Wygant dissenters went on to emphasize that “while racial distinctions are irrelevant to nearly all legitimate state objectives that are properly subject to the most rigorous judicial scrutiny in most instances, they are highly relevant to the one legitimate state objective of eliminating the pernicious vestiges of past discriminar tion.” Id. at 1891 (emphasis added). Indications to the same effect abound in the fourteen opinions filed in the three cases from last Term. Justice Powell in Sheet Metal Workers, for example, indicated that race-conscious relief could lawfully run to non-victims “in cases involving particularly egregious conduct,” id. 106 S.Ct. at 3054; see also id. at 3055 (“The finding by the District Court and the Court of Appeals that petitioners have engaged in egregious violations of Title VII establishes, without doubt, a compelling governmental interest sufficient to justify the imposition of a racially classified remedy.”). Justice White, who authored the Court’s opinion two Terms ago in Firefighters Local Union No. 1784 v. Stotts, 467 U.S. 561, 104 S.Ct. 2576, 81 L.Ed.2d 483 (1984), emphasized in his Vanguards dissent that a predicate of racial discrimination was necessary to warrant racial preferences: Under the present law, an employer may adopt or be ordered to adopt racially discriminatory hiring or promotion practices favoring actual or putative employees of a particular race only as a remedy for its own prior discriminatory practices disfavoring members of that race. 106 S.Ct. at 3081 (emphasis added). Justice White’s criticism that the Vanguards majority had paid “scant attention to this necessary predicate for race conscious practices,” id., was met with a vigorous disavowal by Justice O’Connor, who along with Justice Powell joined to provide the Vanguards majority. Justice O’Con-nor emphasized that “the Court’s opinion does not hold or otherwise suggest that there is no ‘necessary predicate for race conscious practices.’ ” Id. at 3080 (quoting dissenting opinion of White, J.). Quite apart from the specific expressions in this respect, the Court’s staunch refusal to sanction racial quotas or orders effecting a strict racial balance suggests that, to be valid, race-conscious remedies must in fact be remedial in nature. They are not to promote the Utopian goal, as some might see it, of a society in strict proportional racial balance in every employment setting, from brain surgery to basketball. Indeed, if the core element of remediation for past discriminatory practices is unnecessary, we are at a loss to explain the purpose informing the Sheet Metal Workers plurality’s elaborate exegesis of the principle that affirmative action is aimed at dismantling the structure of prior discrimination. 106 S.Ct. at 3031-57. There must, in essence, be a direct nexus between the race-preferential action and the remediation of discrimination. The Sheet Metal Workers plurality issued a stem warning in this respect: Although we conclude that § 706(g) [of Title VII, 42 U.S.C. § 2000e-5(g)] does not foreclose a district court from instituting some sorts of racial preferences where necessary to remedy past discrimination, we do not mean to suggest that such relief is always proper. Id. at 3050 (emphasis added). Quite to the contrary, Justice Brennan emphasized, a “court should exercise its discretion with an eye toward Congress’ concern that race-conscious affirmative measures not be invoked simply to create a racially balanced work force.” Id. (emphasis added). This broad admonition was followed immediately by the plurality’s specific teaching that “[i]n the majority of Title VII cases, the court will not have to impose affirmative action as a remedy for past discrimination, but need only order the employer or union to cease engaging in discriminatory practices and award make-whole relief to the individuals victimized by these practices.” Id. See id. at 3036. Racial preferences, the Sheet Metal Workers plurality emphasized, are the exception, not the rule. Indeed, Justice Bren [61] nan’s comprehensive opinion for the plurality identified only two permissible exceptions: where the employer or labor union “has engaged in persistent or egregious discrimination,” and where racial preferences “may be necessary to dissipate the lingering effects of pervasive discrimination.” Id. 106 S.Ct. at 3050 (emphasis added).19 The record before the Sheet Metal Workers Court abundantly satisfied the demands mandated by the discrimination predicate: the discriminatory practices of the New York City union were “pervasive and egregious.” Id. at 3057.20 The Vanguards case, which as we noted before held narrowly that § 706(g) of Title VII does not preclude entry of a consent decree that provides relief to nonvictims, likewise involved a clear, undisputed history of discrimination. After years of unsuccessfully defending against various claims of discrimination, Cleveland embraced a different posture: “[T]he City of Cleveland had eight years ... of having judges rule against [it]. You don’t have to beat us on the head. We finally learned what we had to do.” Vanguards, 106 S.Ct. at 3067. The Vanguards record thus included not only evidence of “ ‘a historical pattern of racial discrimination’ ” but an “express admission [by the City] that it had engaged in discrimination.” Id. at 3070. See supra text at 418. It cannot go unnoticed, therefore, that in both instances in which the Supreme Court sustained racial preferences last Term, the predicate of discrimination was firmly established, with plenty of room to spare.21 This should come as no surprise. Last Term’s outpouring by a sharply divided Court on the nettlesome subject of affirmative action was not, of course, written on a blank slate. Indeed, the opinions were written against the background of the three watershed, racial-preference cases of the 1970’s, Bakke,22 Fullilove23 and Weber, which, upon analysis, all share the salient characteristic of impaired access by certain (but by no means all) minorities to specific, identifiable educational, commercial, or job opportunities. Tellingly, the remediation theme was prominently featured in all three decisions. Nothing in the various opinions in those cases suggest judicial sanctioning of race-based social engineering schemes; rather, it was the task of remedying discrimination that warranted race-conscious remedies redounding to the benefit of non-identifiable “victims” of discrimination. In Bakke itself, a deeply divided Court justified the consideration of race in the admissions process only by virtue of the perpetuation of impaired access by certain minorities to Davis’ Medical School. Likewise, the various opinions in Fullilove em [62] phasized that the 10 percent government contract set-aside program at issue there was remedial in nature, specifically designed to reduce past racial and ethnic discrimination. In his plurality opinion, Chief Justice Burger dwelt on the fact that Congress created the set-aside program only in the wake of long experience with government contracts and after determining that a variety of impediments to equal opportunity existed in that arena. Indeed, previous Congressional efforts aimed at improving minorities’ record in bidding for government contracts had been unavailing. Fullilove, 448 U.S. at 463-67, 100 S.Ct. at 2767-69. The race-preference regime at issue was, as it were, a last resort, shielded in no small measure by the Court’s substantial deference to Congress, and particularly Congress' exercise of its broad powers under both the General Welfare Clause and section 5 of the Fourteenth Amendment. As the Fullilove plurality emphasized, the Court was faced with “the broad remedial powers of Congress. It is fundamental that in no organ of government, state or federal, does there repose a more comprehensive remedial power than in Congress, expressly charged by the Constitution with competence and authority to enforce equal protection guarantees.” Id. at 483,100 S.Ct. at 2777. But, even in view of this unparalleled reservoir of governmental power in a free society, the Court demanded that the legislative action be remedial in nature. This bedrock requirement was emphasized by Justice Powell in his concurring opinion: [The set-aside] is justified as a remedy that serves the compelling governmental interest in eradicating the continuing effects of past discrimination identified by Congress. Id. at 496, 100 S.Ct. at 2784 (emphasis added). Unlike Bakke and Fullilove, Weber did not involve a governmental classification based on race; rather, that decision addressed whether Congress had so regulated the private sector as to prevent an employer and union from voluntarily24 creating a set-aside program for blacks seeking to enter craftworker training programs. In concluding that Congress in Title VII had not chosen to prohibit outright such private arrangements, Weber’s five-member majority25 reasoned that the aims of the Kaiser-Steelworkers plan were commensurate with the purpose animating Congress in passing Title VII — to open up employment opportunities and thus eliminate the remaining vestiges of discrimination. As it was destined to do again seven years later in Vanguards and Local 28, however, the Court in Weber, pointedly emphasized the narrowness of its holding— “Title VII’s prohibition in §§ 703(a) and (d) against racial discrimination does not condemn all private, voluntary, race-conscious affirmative action plans.” Id. 4434, U.S. at 208, 99 S.Ct. at 2729 (emphasis added). The Kaiser-Steelworkers affirmative action plan at issue in Weber fell within permissible limits because, like Title VII itself, the plan “was designed to break down old patterns of racial segregation and hierarchy.” Id.; see also id. at 204, 99 S.Ct. at 2728. That invidious pattern was graphically demonstrated by two facts: first, “blacks had long been excluded from craft unions,” id. at 198, 99 S.Ct. at 2725, thus making Kaiser’s situation in Grammercy, Louisiana, similar to that obtaining in New York City in Sheet Metal Workers; and second, a gross statistical disparity existed in that “prior to 1974 only 1.83% (5 out of 273) of the skilled craftworkers at the [Kaiser] plant were black, even though the work force in the ... area was approximately 39% black.” Id. Clearly, the nongovernmental, race-preference plan in Web [63] er was “remedial” in the broad sense of opening up to minorities theretofore closed channels of employment opportunities. Thus, as we read the law, remediation of present discrimination (or extant results of its insidious prior operation) is the absolutely indispensable element of the legality of remedies which differentiate human beings on the basis of race. B In addition to the threshold requirement of discrimination, the law contains yet another hurdle before a race-preference program may properly be employed: the remedy crafted to address a violation must be tailored to fit that violation. This requirement is not new. The principle is rooted in the rich traditions of equity and firmly buttressed by common sense. If, by way of modest example, Smith constructs a building on Whiteacre that intrudes slightly onto Jones’ Blackacre, the remedy that fits the violation, one would naturally think, would be to require Smith to remove the portion of his building unlawfully situated on Blackacre. It would seem excessive and wrong to require Smith to demolish the entire building, rather than just the offending part of the structure, for such a “remedy” would partake more of the nature of a sanction or punishment upon Smith for trenching on Blackacre. This principle holds true in remedying violations of statutory and constitutional rights, as well as rights protected at common law. Thus, when, for example, Title VII liability is established, the remedy must be crafted carefully to cure the violation that activates the judiciary’s remedial arsenal. The Court made this broad rule clear in the school busing cases. Although approving in principle mandatory transportation of school children to effect desegregation in Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971), the Court emphatically did not sanction appellate abdication of reviewing with care the powers exercised by federal trial courts in respect of a public educational system. Rather, the Supreme Court directed the lower courts to view with grave caution sweeping “remedies” that disrupt basic institutions of our society in order to ensure that the mode of vindicating basic rights fits the violation. See, e.g., Milliken v. Bradley, 418 U.S. 717, 94 S.Ct. 3112, 41 L.Ed.2d 1069 (1974). The Supreme Court has employed this sound and venerable principle in the arena of state-imposed or -sanctioned racial preferences. Although we recognize full well that the Justices have not been of one mind as to the proper degree of scrutiny to apply to governmental line-drawing by race,26 all members of the Supreme Court appear to agree that courts must examine with great care the nexus between violation and remedy. For, “[rjacial classifications are simply too pernicious to permit any but the most exact connection between justification and classification.” Fullilove, 448 U.S. at 537, 100 S.Ct. at 2805 (Stevens, J., dissenting). [64] Wygant provides a strong basis for this conclusion. Justice Powell, joined by Chief Justice Burger, then-Justice Rehnquist, and Justice O’Connor, made the point this way in his plurality opinion: “[The means cho-' sen] must be ‘narrowly tailored to the achievement of the goal.’ ”27 Wygant, 106 S.Ct. at 1846 (quoting Fullilove, 448 U.S. at 480, 100 S.Ct. at 2776). In Wygant, this meant that even if the Jackson, Michigan School Board had engaged in discriminatory conduct, the reformed Board was not at liberty to remediate in any fashion it chose. The Board’s remedial impulses were constrained by countervailing considerations. Indeed, the layoff provision that deprived Wendy Wygant of her teaching job was not, the plurality concluded, “a legally appropriate means of achieving even a compelling purpose.” Id. 106 S.Ct. at 1852. In addition to the lead opinion in Wygant, Justice O’Connor in concurring expressly embraced the “narrowly tailored” standard. Id. at 1852. Three other Justices, to be sure, advocated the weaker standard of “substantially related.” But even under that less daunting yardstick a “rational relationship” between ends and means would appear not to be enough; rather, the remedy must be “tailored” substantially to the actual accomplishment of what Justices Brennan, Marshall, and Blackmun describe as “important government objectives.” Id. at 1847. Finally, Justice Stevens would require “scrupulous fairness” in adopting the practice for procedure and inquire whether the method employed “is specifically designed to achieve [the] objective.” Id. at 1869-70.28 IV We now apply these principles to the case at hand. A The record before us shows beyond dispute that the predicate of discrimination has not been established in this case. The issue, indeed, is not even close. The District Court entered the race-conscious remedy, incorporated in the consent decree, in March 1985. We are entirely at a loss, however, to discern any act of hiring discrimination, much less any pattern or practice of discrimination, that the court was trying to remedy. The facts are staggeringly to the contrary: the overwhelming majority of new hires in the Fire Department in recent years have been black. As we mentioned above, in 1983, no less than 80.5 percent of new hires were black. In 1982, 67.5 percent of new hires were black. These numbers clearly indicate that the Fire Department is not engaged in hiring practices that discriminate against blacks. Indeed, they powerfully suggest that the District Court’s approval of the racial pref [65] erence plan had nothing to do with “breaking] down old patterns of racial segregation and hierarchy.” Weber, 443 U.S. at 208, 99 S.Ct. at 2730. In addition, no serious contention has been advanced that this record contains the slightest indication of ongoing racial discrimination in hiring by the Fire Department. And no wonder, in light of the findings of the Office of Human Rights. Those conclusions, the reader will recall, are overwhelmingly to the effect that no discriminatory conduct infects the Fire Department’s practices and procedures (save arguably, for the April 1981 hiring). What is more, the District of Columbia does not stand before us like Cleveland in Vanguards as a self-confessed discriminator, or like Local 28 in Sheet Metal Workers as a lawless, obdurate bastion of blatantly discriminatory practices. Nor is it like Kaiser, with its craft union ranks overwhelming white, a powerful present-day demonstration of a prior regime of discrimination. Instead, as we have seen, the District has steadfastly and persuasively protested its innocence of any discriminatory activity. And, more importantly, the record before us shows that the District has,in fact conducted its hiring in a nondiscriminatory manner. To return briefly to the facts, between 1969 and 1980, an average of 41.8 percent of those hired by the Fire Department each year were black. In 1981, when the District employed what it believed was a test validated for use as a rank-ordering device, the Department hired only 23 new firefighters, two of whom were black.29 Since 1981, however, an average of no less than 75.5 percent of those hired each year have been black.30 We pause to observe that, as the District Judge noted, conflicts exist in the case law with respect to the proper definition of the relevant labor market.31 In [66] upholding the quota plan here, the District Court opted in favor of the applicant pool, which was 74.53 percent black in 1980 and 64.6 percent black in 1984, as opposed to the population or labor pool of the Washington Metropolitan Area. (Roughly half of those hired by the Department at entry levels are from the suburbs). According to the 1980 Census, only 29.3 percent of the metropolitan area population within the eligible age range (20 to 28 years of age) for entry-level positions are black. But even relying on the applicant pool, we are at a loss to understand how the District Court arrived at its implicit conclusion that the District’s hiring practices had been discriminatory. As we have seen, the average percentage of black hires each year since 1981 has been greater than even the 74.35 black percentage in the 1980 applicant pool. And, the percentage of black hires is grossly, favorably disproportionate to the percentage of blacks in the eligible population (a fact likely attributable to the District’s recent campaign aggressively to recruit black firefighters).32 See supra note 28. On this record, there were, in sum, no present day impediments to black hiring. There were no relics from yesteryear to overcome. There were no vestigial remnants demanding to be extirpated root and branch. The Plan’s hiring provisions were not designed to dismantle the structure of discrimination. That dismantlement had long since occurred. The crucial condition precedent — discrimination in need of remediation — was therefore wanting. In its absence, the Plan cannot lawfully stand. B Upon analysis, the only discrimination-related concern articulated by the District in this court is its professed desire to avoid the adverse impact of a rank-order use of its hiring test, the FST. This modest goal, upon which the District confers the fancy label of “compliance” relief,33 is [67] painfully incompatible with the fundamental requirement of “narrow tailoring,” to which we now turn. Oral argument in this case graphically demonstrated that the District has been unimaginative in the extreme in seeking to avoid a regime of racial preferences. Indeed, the District’s posture is downright anomalous. It professes a keen interest in having a pencil-and-paper test for entry-level firefighters, yet it used taxpayer money to hire an expert (Dr. Barrett) who lambasted the very idea of employing such an exam for firefighters.34 As if that oddity were not enough, the District set the pass-fail level of the exam at a remarkably humble level — at a score where mere random answering would gamer a passing mark. Dart-throwing methods of test answering sufficed under the District's screening methodology. Not surprisingly, under the District’s benign egalitarianism, almost everybody who showed up to take the test in 1980 and 1984 passed it. The District nonetheless complains that its hands are tied. It earnestly wants to employ a pencil-and-paper test, it says, but is effectively precluded from selecting among those who passed the test on a random (or lottery) basis because the District would, the District bemoans, lose Congressional funding if it were to do so. See Continuing Appropriations, 1985, Pub.L. No. 98-473, 98 Stat. 1837 (Oct. 12, 1984). The District claims that it therefore must hire on a rank-order basis, but has been unable thus far to validate the FST for that use. The District argues that various dangers lurk everywhere were it to employ some other test, so it is bound to stay the course with the FST. The District’s protestations ring hollow. Indeed, the District has refused even to consider seriously alternative tests which the Hammon plaintiffs themselves strenuously maintain are employed satisfactorily and nondiscriminatorily in other jurisdictions.35 These tests, the Hammon [68] plaintiffs urge, are readily at hand if the District would but avail itself of them. The United States echoes the Hammon plaintiffs’ protestations in this respect. Other alternative courses of action, not based on race, also suggest themselves upon a moment's reflection. For instance, the District could have exhausted the 1984 eligibility list. Or, the District could have moved in the intervening years to validate the FST for rank-order use (as the experts at OPM thought had already been done as of 1979). Or, of course, the District could have ceased to use the test, especially since, as we have seen, the passing score was set at such a meek and lowly level that the exam results, at least in the lower strata, would seem virtually meaningless. Thus, even assuming arguendo that the District’s hiring procedures constituted a “violation” necessitating remedial action, its race-based hiring methods were not narrowly tailored to accomplish its remedial purposes. The District, had it paused to consider the matter, could surely have found an avenue of relief more congruent with the demands of Title VII and the Due Process Clause. Cf Wygant, 106 S.Ct. at 1852. Quite apart from the District of Columbia’s failure, we are constrained to conclude that the District Court in hastily approving the affirmative action plan likewise fell into error. The court’s action does not comport with Justice Brennan’s recent reiteration of the well-established principle that “the court should also take care to tailor its orders to fit the nature of the violation it seeks to correct.” Sheet Metal Workers, 106 S.Ct. at 3050. In its opinion, the District - Court took pains to note: “[T]he Court is concerned that reasonable alternatives were not seriously discussed.” Hammon, 606 F.Supp. at 1094. Nevertheless, the court went on to uphold the hiring provisions, finding an “absence of clear options.” This will not do. The District Court has a duty to determine whether a workable, race-neutral alternative exists, and if one does exist, to rule in favor of that course of action. The District Court may not blindly embrace a racially preferential scheme at the first invitation. It must act, in Justice Brennan’s words, “cautiously.” Circumspection, not docket management, must be the guiding principle in these sensitive cases involving as they do basic human rights and human dignity. In sum, the conclusion irresistably suggests itself that the race-preferential hiring scheme approved by the District Court is not, as the law commands, narrowly tailored to cure the “violation.” Indeed, it is not tailored at all. It cannot stand. C One might be led to wonder why the District of Columbia would embark upon this course of insisting upon administering an unvalidated test, with the pass-fail mark set at a meaningless level, and then employing a quota remedy to avoid the effects of any rank-order usage of that “test,” especially when other tests, acceptable to the plaintiffs, were at hand. The answer to this enigma lies not just in the District’s desire to avoid a lottery but in its effort to achieve a strict racial balance in its workforce in compliance with District of Columbia law. We turn now to examine that effort, and whether fidelity to the strictures of D.C. law can justify a racial quota. In 1976, the District of Columbia City Council passed D.C.Law 1-63. That measure provides that every agency of the District of Columbia government “shall ... develop an affirmative action plan.” See supra note 1. Such plans, the law mandates, shall have as their goal “full representation,. in jobs at all salary and wage levels and scales, in accordance with the representation of all groups in the available work force of the District of Columbia, including, but not limited to, Blacks, Whites, Spanish-Speaking Americans, Native Americans, Asian Americans, females and males.” The “available work force of the District of Columbia” is sweepingly [69] defined in terms of raw population. Specifically, as set forth in D.C.Code § 1-507, “ ‘available work force’ means the total population of the District of Columbia between the ages of 18 and 65.” This statutory objective could not be clearer. Each affirmative action plan is to achieve a specific racial balance.36 The Affirmative Action Plan approved by the lower court in this case is equally clear. “The long-range goal of the Fire Department in the hiring of entry-level personnel is expressed in D.C.Code 1-507.” The parties to this litigation expressly recognized as much. The parties’ Statement of Stipulated Material Facts at ¶ 53 trumpets with admirable brevity: “The AAP was intended to achieve compliance with D.C.Law 1-63.” That proportional representation was being sought is scarcely surprising. Back in 1981 in the OHR proceedings, the Fire Department candidly admitted that it was “not in compliance with D.C.Law 1-63, as it related to the racial make-up [of] the work force of the District of Columbia.” OHR Director’s Summary Determination at 7. It is abundantly clear that at least for the past five years, the District has assiduously sought to achieve racial representation in the Fire Department workforce proportional to that in the District’s population at large. It has, in brief, faithfully sought to bring itself into conformity with the explicitly race-conscious demands of D.C. law. This is an impermissible goal under our law. To draw lines based on race, our Constitution demands that government have the most weighty reason for doing so, one that is “compelling” or at least “important.” A government-inspired or -mandated effort to attain racial balance not only fails to satisfy that daunting standard, but is itself suspect. As Justice Powell stated in Bakke, “[a governmental] purpose ... to assure [the representation of] ... some specified percentage of a particular group merely because of its race ... must be rejected not as insubstantial but as facially invalid.” Bakke, 438 U.S. at 307, 98 S.Ct. at 2757. And that position, as we read the cases, represents the law of the land. Since Bakke, the principle has been forcefully reiterated that “[preferring members of any one group for no reason other than race or ethnic origin is discrimir nation for its own sake.” Id. Such preferences are forbidden by the Constitution. In Fullilove, Justice Stewart, in dissent, stated that “since the guarantee of equal protection immunizes from capricious governmental treatment ‘persons’ — not ‘races’ —it can never countenance laws that seek racial balance as a goal in and of itself.” Fullilove, 448 U.S. at 529, 100 S.Ct. at 2801. But we need not repair to dissents, no matter how forceful, to discern the governing principle that seeking racial balance will not do. As we saw more recently in the Title VII setting of Sheet Metal Workers, Justice Brennan admonished that “the court should exercise its discretion with an eye towards Congress’ concern that race-conscious affirmative measures not be invoked simply to create a racially balanced work force.” Sheet Metal Workers, 106 S.Ct. at 3050; see also id. at 3051. See supra text at 422. The Constitution and Title VII are thus congruent on this vital issue — discrimination on grounds of race is absolutely forbidden if the purpose is to achieve racial balance. Here, D.C.Law 1-63, and the District’s Affirmative Action Plan adopted in an effort to comply with that statute, clearly seek racial balance as a goal in and of itself. We conclude that the statutory goal of a racially balanced workforce is not only an inadequate ground upon which to support the AAP, but is a ground that stands condemned by Title VII and the higher law of the Constitution. V To summarize our holding, we conclude that the District has not demonstrated the requisite predicate of discrimination so as to justify use of a race-conscious hiring [70] remedy;37 that both the District of Columbia and the District Court failed to employ, and indeed even seriously to consider, non-race-based alternatives; and finally, that the District’s admitted goal of achieving racial parity is constitutionally invalid. Accordingly, the challenged portion of the judgment of the District Court is Reversed. ADDENDUM Our dissenting colleague strenuously argues that “vestiges” of past discrimination remain in the Fire Department, and that the race-conscious hiring plan at issue represents a proper remedial vehicle for extirpating these remnants of yesteryear. To demonstrate these “vestiges,” the dissent features what it considers the “striking” underrepresentation of black firefighters. Dissenting Op. at 435, 441. This tack does not withstand analysis, however, as the perceived disproportion is premised on an entirely artificial comparison of the number of black firefighters and the number of blacks in the District’s workforce. Id. It is undisputed that the Fire Department has consistently hired from the entire Washington metropolitan area; indeed, approximately half of the District’s entry-level hires are from the suburbs. According to the 1980 Census, only 29.3 percent of the metropolitan area population (within the eligible age range) is black, which is less than the percentage of blacks in the Fire Department. The dissent’s much vaunted dis-proportionality thus disappears when subjected to the most rudimentary analysis.1 The dissent also contends that the law favors voluntary remedial action. We agree. The dissent then goes on to suggest, however, that because the AAP was adopted by a political majority, its validity should be subject to less exacting review. Dissent Op. at 440, 442. The dissent’s deferral to the will of the majority is odd in the extreme where, as here, the members of the benefited “minority” comprise the political majority. Even those members of the Court advocating “intermediate scrutiny” for race-conscious policies are of the view that “any statute must be stricken that stigmatizes any group or that singles out those least well represented in the political process to bear the brunt of a benign program.” Bakke, 438 U.S. at 361, 98 S.Ct. at 2784. See also id. at 357, 98 S.Ct. at 2782; Fullilove, 448 U.S. at 519, 100 S.Ct. at 2795. Only recently, Justice Marshall, joined in dissent by Justice Brennan, reiterated this basic principle in Munro v. Socialist Workers Party, — U.S. -, 107 S.Ct. 533, 93 L.Ed.2d 499 (1986). The dissent’s position, which entirely ignores the composition of the political majority, finds no support in law or reason under the circumstances presented here. In our view, the dissent has, with all respect, failed to come to grips with the fundamental issue in this case: whether a public employer that not only is not characterized by closed channels of opportunity for minorities but to the contrary has been hiring a disproportionately large number of blacks at entry levels for a number of years, can employ a race-based hiring quota when non-race-conscious remedies are [71] readily at hand. In our view, the principled resolution of this issue is clear — a race quota cannot stand under these circumstances. Our law requires no less than invalidation of a gerry-built scheme that easily could have been avoided had the parties been animated by the bedrock principle in our constitutional and legal edifice that human beings should, save for compelling reasons, be free of classifications fashioned on the immutable characteristic of race.2 Footnotes [1] D.C. Law 1-63 contains two provisions of note: Every District government agency shall develop and submit to the Mayor and Council an affirmative action plan. D.C.Code Ann. § 1-508 (1981). The goal of affirmative action in employment throughout the District government is and must continue to be, full representation, in jobs at all salary and wage levels and scales, in accordance with the representation of all groups in the available work force of the District of Columbia, including, but not limited to, Blacks, Whites, Spanish-Speaking Americans, Native Americans, Asian Americans, females and males. As used in §§ 1-507 to 1-514, "available work force” means that total population of the District of Columbia between the ages of 18 and 65. Id. § 1-507. [2] As to the use of Test 21 throughout the 1970’s, the District maintained that no evidence had been offered as to what, if any, adverse impact Test 21 had produced. To the contrary, the District asserted that "[ejven as early as 1971 the OHR conducted a validation analysis on Test 21 and found that there was a high correlation between scores on Test 21 and job performance.” Respondents’ Proposed Findings of Fact and Conclusions of Law ("Proposed Findings”) If 7 at 11 (citing Intervenor Local 36, Exhibit 67). According to the District, a subsequent validation analysis, employing job performance measures, conducted by OPM in April 1979 (six months prior to the filing of the Holmes-Sheffield complaints), led to the conclusion "that Test [21] was [a] valid prediction of future-job performances." Id. [3] The District agreed that the test should be employed solely for pass/fail screening purposes, not as a rank ordering device: Given the high level of adverse impact from rank order use of the FST, the evidence submitted to support its validity for that purpose is inadequate. Id 1f 24 at 18; see also id. ¶ 26. [4] Likewise, the District denied any discrimination with respect to promotions, including any adverse impact occasioned by any of its procedures. Id. ¶¶ 42, 44, 51, 52-58 (no adverse impact); ¶¶ 57-59 (job-relatedness of promotion test); ¶¶ 64-66 (non-discriminatory access to exam manuals); ¶¶ 70-73 (effects of assignments to Fire Prevention Division); ¶¶ 84, 88-89, 90-93 (specific promotion decisions, and operation of the Promotion Board); ¶¶ 101-103 [53] (job assignments and training); ¶¶107-111 (statistical evidence of complainants’ expert “show[ed] very small differences of achievement of Blacks and Whites in the DCFD. Even in the salary area Blacks and Whites were close.”). [5] Similarly, the Hearing Examiner found the Fire Department’s promotion process to be free of discrimination. As to the promotions examination, which had been vigorously attacked by the complainants, the allegations of adverse impact on minorities were found to be "unsupported on the basis of this record." Id. ¶ 102. In addition, the Hearing Examiner found the promotions examination to be job-related, id. ¶ 103, and concluded that "in developing promotional examinations the Fire Department has made efforts to produce examinations that are objective, job-related and fair to all members of the Department.” Id. Other charges of discrimination were found to be entirely without foundation. See, e.g., id. IT 111 ("no evidence of any discrimination by the Promotion Board in performing its functions”); IT 113 ("no evidence of discrimination with respect to the functions of the Promotion Board and the Fire Chief in making promotions to the ranks of Sergeant, Lieutenant and Captain.”); ¶ 118 ("no basis to conclude that any discrimination of any type exists in promotions to the level of Battalion Fire Chief and none is found.’’); ¶ 122 (no basis for finding discrimination against Deputy Chief Joseph Kitt); ¶ 123 (no basis for concluding that "any discrimination exists in the selection for positions above the rank of Battalion Fire Chief’); 11124 (no evidence of adverse impact in the selection of technicians); ¶ 130 (“[Cjomplainants' claim with respect to study time in the Fire Prevention Division discriminating against Blacks is totally unsubstantiated by evidence in this record.”); ¶ 152 (complaint that black firefighters "have been denied access to [examination] manuals on the basis of their race is unsubstantiated by this record and totally devoid of merit”); ¶¶ 155-63 (allegations of two incidents of denials of promotions found to be "without merit”); see also fflf 169-70, 174, 177, 180 & 186. Cf. Bishopp v. District of Columbia, 788 F.2d 781 (D.C.Cir.1986) (D.C. Fire Department discriminated against white males by denying them promotions). [6] The Department "admitted that they [sic] were not in compliance with D.C.Law 1-63 as it relates to the racial make-up of the work force of the District of Columbia.” Id.; see supra note 1. [7] Test 21 was administered to candidates for the Metropolitan Police Department as well. In Washington v. Davis, the Supreme Court rejected the challenge that a standardized test, specifically Test 21, was constitutionally infirm by virtue of its disproportionately adverse impact on blacks. [8] The EEOC Uniform Guidelines on Employee Selection Procedures, 29 C.F.R. § 1607 (1986), provide suggested standards for employment test validity studies. The guidelines endorse three types of validity studies: content validity studies, construct validity studies, and criterion-related validity studies. A criterion-related validity study, such as was conducted by Dr. van Rijn and Ms. Payne, consists of “empirical data demonstrating that the selection procedure is predictive of or significantly correlated with important elements of job performance.” Id. § 1607.5B. The technical standards for such studies are described at § 1607.14B. [9] Speed apparently played a pivotal role in the selection process. According to the Hearing Examiner: It is clear from the record that the determining factor as to whether an applicant was hired in this first group was not his/her place on the list, but how quickly he was processed by the Police and Fire Clinic. Id. U 84, D.A. at 60. [10] This provision was not, however, framed as a remedial order to eliminate discrimination. Indeed, the Hearing Examiner expressly provided that the date of entry was to be used only for future promotional eligibility "and shall not be construed as entitling any applicant selected pursuant to these recommendations to retroactive governmental benefits, such as back pay, insurance, etc.” Id. ¡3 n. *, D.A. at 69 n. *. [11] The Hearing Examiner also issued a standard anti-retaiiation admonition and made recommendations about attorney’s fees and monitoring the AAP and selection procedures. [12] The Hearing Examiner found that the District "was aware of the possible adverse impact that would be caused if the FST was used operationally, but decided to use the test operationally as a rank ordering selection procedure.” Id. 1f 34, D.A. at 49. It must be remembered, however, that the District administered the FST in November 1980 and employed it as a rank-order device in January 1981, before OPM’s validity study was impeached. At that time, the District thought that the test was valid, and therefore properly administered and used for rank-order purposes despite its possible adverse impact. Upon learning that the FST was not validated [56] for rank-order usage, the District promptly adjusted its procedures. [13] The test was thereafter administered in March, April, and July 1984. [14] Under the hiring portion of the AAP, 12 certificates (lists) of approximately 120 candidates each are developed from separate lists of white males, white females, black males, black females, Hispanic males, and other males based on the candidates’ scores on the FST. The AAP requires that the composition of each certificate approximate that of the group of persons who pass the FST: blacks (60%), whites (35.1%), Hispanics (2.4%), females (7%), and others (2.6%). Finally, the AAP mandates that each academy class of firefighters be composed of at least 60% blacks and 5% females. [15] The latter provision is, of course, triggered by virtue of the D.C. Fire Department’s status as a public entity. [16] The court’s order, which precluded implementation of the plan, was stayed insofar as the hiring provisions were concerned on April 26, 1985. [17] In Sheet Metal Workers, the Court held that the affirmative action order requiring the union to admit a certain percentage of nonwhites to union membership satisfied the requirements of Title VII. Sheet Metal Workers, 106 S.Ct. at 3052. The Court was then faced with the petitioners' constitutional claim, which it resolved without addressing whether the Constitution placed limits on that which Title VII would otherwise permit. Id. at 3052-53. As we presently note in more detail, it seems to us that for obvious reasons Congress would not have intended that Title VII’s scope extend to unconstitutional employment practices. In addition, we observe that in at least one other instance, the Court has determined that a provision of the Civil Rights Act of 1964 is commensurate with the demands of the Constitution. Bakke, 438 U.S. at 287, 325, 98 S.Ct. at 2746, 2766. [18] Justice Stevens concluded that "it is not necessary to find that the Board of Education has been guilty of racial discrimination in the past to support the conclusion that it has a legitimate interest in employing more black teachers in the future.” Id. 106 S.Ct. at 1867 (Stevens, J., dissenting). The justification for racial line-drawing in Justice Stevens’ view was supplied by the "public interest in educating children for the future," rather than by "some sort of special entitlement [of minority teachers] to jobs as a remedy for sins committed in the past.” Id. The present case, of course, does not involve the unique and sensitive area of education, leaving us without the government's fundamental interest in education to support the use of racial considerations. See also id. at 1852; Regents of University of California v. Bakke, 438 U.S. 265, 311-15, 98 S.Ct. 2733, 2759-61, 57 L.Ed.2d 750 (1978) (opinion of Powell, J.). [19] The plurality indicated that decision was not needed as to *‘[w]hether there might be other circumstances that justify the use of court-ordered affirmative action." Id. But, returning in the same breath to the ubiquitous theme of "discrimination as predicate," the plurality instructed the inferior courts to "consider whether affirmative action is necessary to remedy past discrimination in a particular case," id., and articulated a principle to which we will presently turn in the text, namely that “the court should also take care to tailor its orders to fit the nature of the violation it seeks to correct." Id. 106 S.Ct. at 3050-51 (citation omitted). [20] Discrimination of the grossest sort by Local 28 cannot be gainsaid. Over two decades previously, the New York State Commission for Human Rights had ordered the union, which had never had a single black member or apprentice, to cease its blatantly discriminatory practices. Justice Powell was moved, in concurrence, to characterize the union’s history of discrimination as "contemptuous.” Id. at 3054. "It would be difficult," he wrote, "to find defendants more determined to discriminate against minorities.” Id. at 3055. This is the stuff of which racial-preference remedies are, under present law, lawfully made. [21] We hasten to add that the Court in Vanguards did not pass final judgment on Cleveland’s plan. Instead, the Court held more narrowly that as a matter of law § 706(g) of Title VII did not ban federal district court approval of the consent decree. The ability of the union or others to challenge the decree under § 703 or the Fourteenth Amendment was expressly left unresolved. Id. 106 S.Ct. at 3080; see also id. ' (O’Connor, J., concurring). [22] 438 U.S. 265, 98 S.Ct. 2733. [23] Fullilove v. Klutznick, 448 U.S. 448, 100 S.Ct. 2758, 65 L.Ed.2d 902 (1980). [24] The Weber Court expressly chose not to address the contention that the affirmative action plan at issue there represented an attempt by Kaiser and the Union to comply with Exec. Order No. 11246. Id. 443 U.S. at 209, 99 S.Ct. at 2730. That contention, needless to say, touched rather directly on the asserted voluntariness of the plan. Cf. Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1964). [25] We observe that the majority included the late Justice Stewart, a dissenter in Fullilove, and Justice White, the author of Stotts and a dissenter in Sheet Metal Workers. [26] The Court has invoked the doctrine of strict scrutiny (Chief Justice Burger, Justice Powell, Justice White, then-Justice Rehnquist and Justice O'Connor); an intermediate but nonetheless demanding standard of review (Justices Brennan, Marshall, and Blackmun); and a strict requirement of procedural fairness (Justice Stevens in Fullilove and Wygant). However, where, as here, the racial classification is employed by the political majority to disadvantage the political minority, it would appear that Justices Brennan, Marshall, and Blackmun would also apply the full panoply of "strict scrutiny” analysis, including the "narrowly tailored” prong. Indeed, in their collective Bakke dissent, they opined that "any statute must be stricken that stigmatizes any group or that singles out those least well represented in the political process to bear the brunt of a benign program.” Bakke, 438 U.S. at 361, 98 S.Ct. at 2784. See also id. at 357, 98 S.Ct. at 2782; Fullilove, 448 U.S. at 519,100 S.Ct. at 2795 (Marshall, J., concurring). Cf. Munro v. Socialist Workers Party, — U.S.-, 107 S.Ct. 533, 93 L.Ed.2d 499 (1986). We see nothing in the opinions to suggest the truly remarkable proposition that, in a free society, a "discrete and insular minority” is destined forever to remain precisely that, even when by virtue of demographic shifts yesterday’s minority becomes todays majority. And, we would think, when control of the political process is obtained, one would be hard pressed to discern any institutional failures in the political process disabling the new majority. Cf. J. Ely, Democracy and Distrust 73-104 (1980). [27] Three of those four Justices flatly rejected the standard of "reasonableness" as the appropriate guide for measuring the legality of the specific remedial action. A closer fit than "reasonableness” was required. [28] This court has recognized the importance of tailoring the remedy to fit the violation as well. In Thompson v. Sawyer, 678 F.2d 257, 294 (D.C. Cir.1982), the court stated in the context of Title VII: In carefully honing a quota, the court should be guided by at least four important considerations. First, it should draw the quota in as limited and short-term a fashion as possible, consistent with effective eradication of discriminatory practices. Second, the court should target the quota specifically at the discrimination to be brought to an end. Third, to the extent possible, it should avoid infringing on the competing, legitimate concerns of others. Finally, the court should consider whether alternative, equally effective methods could supplement or supplant resort to a quota. See United States v. City of Chicago, 663 F.2d 1354 (7th Cir.1981). The Thompson court approved in principle a quota-type remedy in the context of a government employer that had demonstrated a marked "inability or unwillingness to recognize discrimination” which had been blatantly visited upon female employees, and in an industry characterized by "longstanding and pervasive discrimination____” Id. at 295. The court was quite clear that such remedies "must be fashioned and used gingerly, with care to see that they wound as little as possible." Id. at 294. The court then went on to direct the trial court as follows: "the decree should be tailored to the discrimination to be corrected____” Id. at 295. Accord Segar v. Smith, 738 F.2d 1249, 1293-95 (D.C.Cir.1984) (despite "pervasive” discrimination by the government employer, Id. at 1264, 1289, the device of promotion goals and timetables “must be used cautiously”). [29] Under a view most favorable to the Hammon plaintiffs, the District engaged in discriminatory hiring practices in 1981 when it hired the 23 firefighters from the initial Certificate of 101 candidates who had achieved the highest scores on the FST during the prior year. In addition, the District carried out the actual processing of these candidates in a manner that impacted adversely on the applications of blacks. These, however, are the sole elements of "discrimination” in the record. And it is not disputed that the District immediately took steps to remediate these perceived shortcomings. First, the District fully exhausted the list of those who had successfully taken the FST, which the reader will recall was virtually everyone, rather than relying upon a rank order cut-off score. Second, it appears (and is unchallenged in this litigation) that the District remedied its processing procedures to eliminate that element of adverse impact on black candidates. These steps were either completed or well under way before the present litigation was even filed, not to mention the date of the court-authorized quota hiring provisions framed in the spring of 1985. [30] Thus, the District of Columbia Fire Department has been hiring blacks at an average rate of close to 50 percent per year since 1969. This bottom-line result scarcely bespeaks what Weber described as "a continuing pattern of racial segregation and hierarchy.” This is especially true since until 1980 the District recruited broadly from the entire metropolitan area, the population of which is almost 70 percent white. Beginning in 1980, the Fire Department stepped up its recruiting efforts targeted at residents of the District, a development that coincided with the enactment of a D.C. law requiring new hires to become residents of the District of Columbia within six months of hiring. Theretofore, no residency requirement obtained for firefighters wishing to serve in the Nation's Capital. [31] Under the case law, the critical comparative group is that of the qualified applicant pool, which may or may not be the entire work force in the area. Thus, when the job qualifications involved are ones that relatively few possess, statistical presentations that faU to focus on those qualifications do not have significant probative value. Hazelwood School District v. United States, 433 U.S. 299, 308 n. 13, 97 S.Ct. 2736, 2742 n. 13, 53 L.Ed.2d 768 (1977). See also Cook v. Boorstin, 763 F.2d 1462, 1469 (D.C.Cir. 1985) (case law establishes that "the percentage of minority employees selected for a given position should be compared to the percentage of minority candidates in the pool of persons qualified for that position")', Metrocare v. Washington Metro. Area Transit Auth., 679 F.2d 922, 930 (D.C.Cir. 1982) (“[Sjtatistics must compare the percentage of blacks hired for given jobs with the percentage of blacks qualified for those positions."); Mayor of Philadelphia v. Educational Equality League, 415 U.S. 605, 620-21, 94 S.Ct. 1323, 1333-34, 39 L.Ed.2d 630 (1974); New York City Transit Auth. v. Beazer, 440 U.S. 568, 586 n. 29, 99 S.Ct. 1355, 1366 n. 29, 59 L.Ed.2d 587 (1979). Because the entry-level qualifications for a D.C. firefighter are apparently minimal, however, the District Court’s use of the applicant pool was not necessarily incorrect. Internation [66] al Brotherhood of Teamsters v. United States, 431 U.S. 324, 339-40 n. 20, 97 S.Ct. 1843, 1856-57 n. 20, 52 L.Ed.2d 396 (1977); Hazelwood, 433 U.S. at 308 n. 13, 97 S.Ct. at 2742 n. 13. We note, however, that the District Court failed to perform any meaningful analysis with respect to the applicant pool, once it had identified that pool as the relevant labor market. The court merely recited the percentage of blacks who applied in 1980 and 1984 and then found that "the plan's hiring quota of 60% blacks is reasonable." Hammon, 606 F.Supp. at 1092. The court did not even compare the applicant percentages to the percentages hired, let alone determine whether the 1980 and 1984 applicant flows had any relevance to past hiring percentages, in light of the District’s aggressive 1980 campaign to recruit blacks. [32] Notably, the complaint in the present lawsuit and .the underlying administrative proceeding before the Office of Human Rights did not allege violations prior to the fall of 1979. The District Court nonetheless relied on the fact that the Fire Department had been segregated in the early 1950’s. The significance of that fact to a present-day hiring remedy is nowhere elucidated, nor is the slightest logical nexus between the generation-old violation and the present-day remedy apparent. To the contrary, the historical fact of segregation in that unhappy era preceded Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873 (1954), which as every schoolchild knows reaffirmed the bedrock constitutional right of all Americans not to be subjected to line-drawing on grounds of race. Other than semi-ancient history, the District Court looked to the raw numbers of minorities in various supervisory positions, principally at more senior levels save for the most senior. That data, of course, goes to the issue of discrimination in promotions, not hiring. And, it will be recalled, the District Court’s Solomonic judgment overturned the promotions portion of the affirmative action plan, a question that neither the District nor the Hammon plaintiffs have brought before us. [33] "Compliance" relief is a concept articulated by the Second Circuit. In Guardians Ass’n v. Civil Serv., 630 F.2d 79 (2d Cir.1980), that court drew a distinction between what it characterized as "compliance relief’ and "affirmative relief.” The court asserted: Compliance [relief] involves restricting the use of an invalid exam, specifying procedures and standards for a new valid selection procedure, and authorizing interim hiring that does not have a disparate racial impact.25 Affirmative relief involves interim hiring at any ratio greater than what is necessary just to avoid a disparate racial impact and any required long-term hiring targets or ratios. Id. at 108. In footnote 25, the court suggested: An example [of compliance relief] would be an interim minority hiring ratio of 30%, where the minority percentage of the applicant pool or the relevant work force is also 30%.... Id. According to the court, certain circumstances warrant compliance relief, while others (more egregious) would give rise to affirmative relief: [67] When it has been established that a selection procedure has been unlawfully used, an appropriate compliance remedy should forbid the use of that procedure, or its disparate racial impact, and may properly assure the establishment of a lawful new procedure. When it also appears that the employer has discriminated prior to the use of the challenged selection procedure, then it may also be appropriate to fashion some form of affirmative relief, on an interim and long-term basis, to remedy past violations (citations omitted). Id. It seems to us that our colleagues in New York are drawing a distinction between when the court can merely remedy the violation at hand (compliance relief) and when it can implement a more global remedy (affirmative relief). This distinction, it would seem, is simply a different way of stating the requirement that the remedy be tailored to fit the violation. In either instance, it is clear that the court requires a present-day Title VII violation as a threshold matter. In Guardians, the court had found that the employer’s use of an exam with significant disparate racial impact violated Title VII, and was sufficient to invoke the court’s equitable compliance relief powers. The court reiterated the requirement of a Title VII violation in Kirkland v. New York State Dept. of Correctional Services, 711 F.2d 1117 (2d Cir.1983). As the court put it, ”[s]imply stated, the remedies provided by a Title VII settlement, especially those containing race-conscious relief, must be substantially related to the objective of eliminating the alleged instance of discrimination____’’ Id. at 1132. See also Berkman v. City of New York, 705 F.2d 584, 595 (2d Cir.1983). The Second Circuit’s "compliance relief’ is therefore not available, as the District asserts, "simply to avoid a disparate racial impact.” D.C.Brief at 17. Instead, it is available under Second Circuit law only to avoid the disparate racial impact resulting from a present Title VII violation. As we are not faced with any Title VII violation, so that a remedy is necessary to bring the District back into compliance with Title VII, we reject outright the District’s attempt to characterize the AAP as “compliance relief.” And, we observe that even if the April 1981 hiring from the Certificate of Eligibles were considered a Title VII violation, the remedy employed — exhaustion of the list of those who passed the exam — surely brought the District back into compliance with Title VII. [34] The District commissioned Dr. Richard S. Barrett to review the adequacy of the Criterion-Related Validity Research Study conducted by OPM. H.E. Opinion |37, D.A. at 49-50. See text supra at 416-17. During the course of Dr. Barrett’s explication of the study’s deficiencies at the 50-day trial, he expressed doubt as to the possibility of developing a valid multiple choice test for firefighter hiring. Id. ¶ 57, D.A. at 53. [35] Courts have found tests employed by fire and police departments valid on a number of occasions. See, e.g., Clady v. County of Los Angeles, 770 F.2d 1421, 1434 (9th Cir.1985); Detroit Po [68] lice Officers Ass’n v. Young, 446 F.Supp. 979, 1007-08 (E.D.Mich.1979); Friend v. Leidinger, 446 F.Supp. 361, 367-75 (E.D.Va.1977), aff'd, 588 F.2d 61 (4th Cir.1978); Bridgeport Guardians v. Bridgeport Police Dep't, 431 F.Supp. 931, 936-39 (D.Conn.1977). [36] This feature did not slip by unnoticed. The bill that ultimately was enacted as D.C.Law 1-63 was vetoed by then-Mayor Walter Washington on grounds of unconstitutionality. [37] The Hammon plaintiffs strenuously assert that, given adequate opportunity, they can provide ample evidence of discriminatory conduct by the District. Our decision today in no wise prevents them from seeking such relief in the trial court. That question is simply not before us, and we thus have no occasion to address it. [1] The dissent charges that we failed to accord the District Court’s "factual findings” the deference mandated by the “clearly erroneous” standard of review. Dissent Op. at 441. An analysis of the two "findings," however, reveals that the charge is without foundation. First, as we make clear in the text of our opinion, we in no wise dispute that the Fire Department was segregated in the early 1950’s. The pertinence of this undisputed fact to the need vel non for a present-day race-conscious remedy is nowhere articulated, save for the incantation of "remaining vestiges” of this bygone era. But in that respect, the District Court's conclusory assertion that "there still exist some vestiges of the aforementioned past discrimination,” Hammon, 606 F.Supp. at 1087, is based on a recitation of the current racial composition of the Fire Department, without more. See supra n. 32. The dissent's reliance on this latter "finding" — made without the benefit of an evidentiary hearing or any other fact-finding device and unsupported by any analysis — highlights the weakness of the dissent's position. [2] Nothing in the Supreme Court’s recent decision in United States v. Paradise, — U.S.-, 107 S.Ct. 1053, 94 L.Ed.2d 203 (1987), alters our opinion in this case. In Paradise, the Court found a compelling state interest in the employer’s "pervasive, systematic, and obstinate discriminatory conduct." Id. 107 S.Ct. at 1065. The Court then determined that the District Court's one-for-one promotion plan was narrowly tailored, concluding that "strong measures were required in light of the Department’s long and shameful record of delay and resistance." Id. at 1074.