Davis v. Washington

512 F.2d 956, 168 U.S. App. D.C. 42, 10 Fair Empl. Prac. Cas. (BNA) 105, 1975 U.S. App. LEXIS 15900, 9 Empl. Prac. Dec. (CCH) 9980

Opinions

SPOTTSWOOD W. ROBINSON, III, Circuit Judge:

On behalf of all blacks who have unsuccessfully sought appointment to the District of Columbia Metropolitan Police Department since 1968, appellants intervened in this class action¹ to assert the claim that Test 21, a written entrance examination administered to all applicants, discriminates against blacks in violation of the Due Process Clause of the Fifth Amendment and federal civil rights laws.² The District Court con- eluded that Test 21 was nondiscriminatory in design and operation and was “reasonably and directly related to the requirements of the police recruit training program.”³ The court entered summary judgment for the defendants, appellees here,⁴ and this appeal followed.⁵ We hold that appellants have demonstrated on the record that Test 21 has a racially disproportionate impact,⁶ and that appellees have not met their heavy burden of showing that the test is related to job performance.⁷ Accordingly, we reverse the summary judgment and remand for further proceedings.

I. THE APPLICABLE LEGAL STANDARD

A. The Facts

Applicants are appointed to positions in the Metropolitan Police Department if they satisfy character and physical standards, have a high school diploma or the equivalent, and receive a raw score of 40 or above on Test 21.⁸ The test was developed by the Civil Service Commission for general use throughout the federal service as a measure of verbal ability, rather than specifically to measure the full range of skills required to perform the tasks of a police officer.⁹ The Department was not involved in the formulation of Test 21.

Appellants tendered to the District Court statistical evidence purportedly demonstrating that Test 21 has a disproportionate impact on black applicants. Among the applicants tested in the District of Columbia from 1968 through 1971, 57% of the blacks failed the test, as compared to a failure rate of 13% for whites.¹⁰ Although black applicants thus failed Test 21 at a rate more than four times greater than the rate for whites, appellees dispute the capability of these statistics to prove a racially disproportionate impact. They further contend that, regardless of the sufficiency of this evidence, Test 21 can be used in selecting police officers because it is sufficiently job related. In support of the latter proposition, they offer a validity study¹¹ conducted under the auspices of the Civil Service Commission demonstrating, they claim, that scores on Test 21 accurately predict performance in Recruit School, the Department’s police training program.

B. The Griggs Standard

In Griggs v. Duke Power Co.,¹² the Supreme Court considered the legality of standarized intelligence tests under Title VII of the Civil Rights Act of 1964.¹³ “If an employment practice which operates to exclude Negroes cannot be shown to be related to job performance,” the Court said, “the practice is prohibited.” ¹⁴ Once it is shown that a particular selection procedure has an exclusionary effect on minority applicants, there is a heavy burden on the employer to show that discriminatory procedure “bear[s] a demonstrable relationship to successful performance of the jobs for which they were used.” ¹⁵ In the instant case, if appellants’ evidence is sufficient to show an exclusionary impact on black applicants, and we hold that it is,¹⁶ then the Department must shoulder the heavy burden of justifying Test 21 by demonstrating that results on the test are satisfactorily related to successful performance of the job of a police officer.

The District Court ruled that appellants’ showings (a) that the number of black police officers in the Department is not proportionate to the population mix of the District, (b) that a higher percentage of blacks fail Test 21 than whites, and (c) that Test 21 has not been validated to show its reliability for measuring job performance, shifted the burden of proof to appellees.¹⁷ We reach the same conclusion on the basis of the racially disproportionate impact that Test 21 is shown to have.¹⁸ The District Court further ruled, however, that Test 21 is exonerated by a direct and reasonable relationship to the requirements of the Department’s police training program.¹⁹ We deem that insufficient, and finding Test 21 not otherwise demonstrated to be job related, we hold that appellees have not met their burden.²⁰

II. THE RACIALLY DISPROPORTIONATE IMPACT OF TEST 21

The evidence in the record establishes that the percentage of black failures on Test 21 is far greater than white failures.²¹ The cases hold, and we agree, that evidence establishing that significantly more blacks than whites fail a written entrance examination given to all applicants is sufficient, as a matter of law, to show the racially disproportionate impact of the examination.²² The disparity disclosed in this case — more than four to one — is larger than differences held sufficiently disproportionate in other cases.²³ Indeed, absent evidence revealing some other reason for the lopsided failure rates appearing here, it is difficult to imagine how disproportionate effect could ever be better demonstrated.

The District appellees contend that the court could properly find that appellants had not made the necessary showing simply by considering the relationship between the percentage of blacks in the Metropolitan Police Department and the percentage of black population in the community.²⁴ They assert that other courts have been more concerned with that type of information than with pass-fail rates. Our research discloses, however, that population data have been considered judicially for only two purposes. There are a number of cases that have held population data alone sufficient to show racially disproportionate impact, but in the apparent absence of data on pass-fail rates.²⁵ Other courts have noted population data merely to corroborate a showing of racially disproportionate impact based on pass-fail rates.²⁶ There is no authority— and we decline to provide any — for the proposition that proof of a racially disproportionate impact must encompass both pass-fail rates and disparate population figures. We think the precedents establish that either demonstration is legally sufficient to shift the burden of establishing job relatedness to the employer.²⁷

Appellees also urge us to consider the affirmative efforts of the Metropolitan Police Department to recruit black officers. We think such efforts are irrelevant to the issue — the discriminatory effect of Test 21 itself. The employer’s lack of discriminatory intent was deemed irrelevant by the Supreme Court in Griggs; “Congress directed the thrust of the [Civil Rights] Act to the consequences of employment practices,” the Court admonished, “not simply the motivation.” ²⁸ Other courts have held that an employee challenging an employment practice as discriminatory need not prove a purpose on the employer’s part to discriminate; ²⁹ the only intent requirement is that the employer consciously perform the allegedly discriminatory act.³⁰ Thus it has been expressly held, and we agree, that efforts to recruit minority members have no bearing on a showing that an employment practice has a racially disproportionate impact.³¹ Although the Department, quite commendably, has succeeded in increasing the proportion of black officers through vigorous efforts, it is self-evident that use of selection procedures that do not have a disparate effect on blacks would have resulted in an even greater percentage of black police officers than exists today.³²

The evidence this case lays bare is further corroborated by the “substantial body of evidence that black persons and other disadvantaged groups perform on the average far below the norm for whites on generalized intelligence or aptitude tests.” ³³ Judicial decisions on the “ever-extending series of challenges to civil service examinations”³⁴ unequivocally establish that blacks are test-rejected more frequently than whites.³⁵ This phenomenon is the result of the long history of educational deprivation, primarily due to segregated schools, for blacks.³⁶ Until arrival of the day when the effects of that deprivation have been completely dissipated, comparable performance on such tests can hardly be expected.

III. JOB RELATEDNESS OF TEST 21

As a result of appellants’ showing that Test 21 has a disproportionate racial impact, there is a heavy burden on appellees to prove that the examination bears a demonstrable relationship to successful performance as a member of the Metropolitan Police Department.³⁷ Appellees contend that they have met this burden by establishing a direct relationship between applicants’ scores on Test 21 and recruits’ average scores on examinations given in the Department’s Recruit School.³⁸

We are compelled to view the evidence presented by appellees in that regard with some skepticism. The assertion of predictive value of Test 21 for achievement in Recruit School is based upon a correlation between Test 21 scores and scores on written examinations given during a 17-week training course. We think this evidence tends to prove nothing more than that a written aptitude test will accurately predict performance on a second round of written examinations, and nothing to counter this hypothesis has been presented to us.³⁹ But despite serious doubt as to the probative value of the evidence,⁴⁰ we are willing to assume for purposes of this appeal that appellees have shown that Test 21 is predictive of further progress in Recruit School. The ultimate issue in this controversy then becomes whether that kind of proof is an acceptable sub stitute for a demonstration of a direct relationship between performance on Test 21 and performance on the job.

Appellees assert that their validity study⁴¹ establishes that Test 21 is predictive of “trainability,” and that therefore the examination survives the Griggs standard. Appellants, on the other hand, have convincingly argued that the record evidence does not demonstrate a sufficient relationship between Test 21 scores and trainability. All entrants into Recruit School pass the final examinations with a grade of 70 or above; if a particular candidate has difficulty, he is given assistance until he succeeds in passing the examinations.⁴² The validity study revealed that persons with high Test 21 scores are more likely to achieve a final average exceeding 85 in Recruit School,⁴³ but there is no evidence to support the proposition that a candidate with an average below 85 is more difficult to train or will not be as good a police officer as a candidate with an average over 85.⁴⁴ Moreover, since applicants who scored below 40 on Test 21 have never been admitted to Recruit School, the validity study expressed no conclusion regarding the likely performance in Recruit School of Test 21 failures.⁴⁵ For these reasons, and because of the departmental policy that nobody fail Recruit School,⁴⁶ appellees have not shown that the admission of applicants who score below 40 on Test 21 into Recruit School would necessitate expanded training time ⁴⁷ or produce Recruit School failures.⁴⁸ We might add that the Recruit School averages apparently have not been used by the Department for any purpose other than the attempt to validate Test 21 in this case.⁴⁹ The' Griggs standard does not, in our opinion, permit validation by a criterion that the employer itself does not believe sufficiently job related.

Appellees maintain that there is support for their position in applicable legal precedents.⁵⁰ Claims similar to that have been made before, and the reported decisions on the point have uniformly rejected a correlation between entrance examination performance and training performance as a means of validating the entrance examination.⁵¹ In a typical case, involving a written entrance examination given by the Philadelphia Police Department,⁵² it was contended that the examination accurately predicted performance in the Department’s Police Academy, as reflected in scores on Academy final written examinations.⁵³ The court refused to find the entrance test in that case job related for several reasons. There was no correlation between Academy performance and effective job performance; ⁵⁴ there was no showing that the Academy examinations were significantly different from the entrance exams, so the correlation might merely have indicated that the entrance examination predicated good scores on a second written test;⁵⁵ and there was no showing that applicants who failed the entrance examination would not do well in the Academy.⁵⁶ As further support for its conclusion, the court noted that no one had failed in the Academy in the past five years.⁵⁷ The similarity between that case and the case at bar is indeed striking, and the same result has been reached by at least three other courts.⁵⁸

Appellees further argue that convincing proof that an employment selection procedure, even one with a racially disproportionate impact, accurately predicts trainability might in some eases satisfy the Griggs standard.⁵⁹ Appellees have not, however, proffered such proof, nor have they persuaded us that trainability could be a proper criterion for validating Test 21. As long as no one with a score below 40 enters Recruit School, as long as all recruits pass Recruit School, as long as the Department’s actions concede that Recruit School average has little value in predicting job performance, and as long as there is no evidence Of any correlation between the Recruit School average and job performance, we enter tain grave doubts whether any of this type of evidence could be strengthened to the point of satisfying the heavy burden imposed by Griggs.

IV. CONCLUSION

The dissent is concerned, as are we, with preservation of the Department’s ongoing efforts to improve its professional standards.⁶⁰ The District Court declined to read the controlling legal principles so that “brakes [would] be placed upon efforts to upgrade recruiting and job standards in law enforcement work.” ⁶¹ The court apparently believed that the relief sought by appellants would necessarily lower recruitment standards and produce “a setback for blacks and whites alike.”⁶² We believe these fears are not well founded. Elimination of Test 21 could result in lower standards only if it were job related, and appellees have not come forward with satisfactory evidence that Test 21 is job related. The common sense theory of validity espoused by the dissent is clearly an inadequate response to appellants’ proof of racially disproportionate impact.⁶³ The rationale is equivalent to a finding of construct validity without satisfying the prerequisites delineated in Douglas v. Hampton.⁶⁴

Appellants’ complaint sought both declaratory and injunctive relief. They are, as a matter of law, entitled to a declaration that Test 21 is invalid, and the conclusions we reach have in other cases served as the basis for enjoining the administration of entrance examinations and ordering affirmative relief.⁶⁵ The focus here, however, has been on the substantive issues, and the record lacks the factual development essential to a sound determination as to whether an injunction against use of Test 21 is warranted. We leave that question open to consideration by the District Court, mindful that declaratory relief in accordance with this opinion might, in consequence of agreement by the parties or otherwise, obviate any need for the coercion of an injunction.⁶⁶

The judgment of the District Court is reversed. Appellants’ motion for partial summary judgment must be granted,⁶⁷ and appellees’ respective motions for summary judgment denied.⁶⁸ The case is remanded to the District Court for entry of appropriate orders, and determination and effectuation of a suitable remedy, in harmony with this opinion.

Footnotes

1

. The litigation originally was a class action challenging allegedly discriminatory practices in promotion of police officers within the Department. That action was decided adversely to the original plaintiffs, Davis v. Washington, 352 F.Supp. 187 (D.D.C.1972), and is not involved in this appeal.

2

. Until 1972, governmental employers were not subject to Title VII of the Civil Rights Act of 1964, Pub.L.No.88 — 352, § VII, 78 Stat. 253, 42 U.S.C. § 2000e et seq. (1970), and governmental employees in the states sought relief from allegedly discriminatory practices under the Equal Protection Clause of the Fourteenth Amendment. E. g., Castro v. Beecher, 459 F.2d 725, 733 (1st Cir. 1972); Chance v. Board of Examiners, 458 F.2d 1167, 1169 (2d Cir. 1972); Carter v. Gallagher, 452 F.2d 315, 318 (8th Cir. 1971), cert. denied, 406 U.S. 950, 92 S.ct. 2045, 32 L.Ed.2d 338 (1972). Since the equal protection requirement is imposed on the Federal Government, and its actions in governing the District of Columbia, by the Fifth Amendment Due Process Clause, Bolling v. Sharpe, 347 U.S. 497, 74 S.Ct. 693, 98 L.Ed. 884 (1954); Bolton v. Harris, 130 U.S.App.D.C. 1, 4 n. 3, 395 F.2d 642, 645 n. 3 (1968), these decisions are plainly relevant to the constitutional contentions in the instant case. In addition to their constitutional base, appel: lants seek a statutory foundation for relief in the Civil Rights Act of 1870. Act of May 31, 1870, ch. 114, § 16, 16 Stat. 144, 42 U.S.C. § 1981 (1970).

Although appellants’ complaint did not allege a violation of Title VII of the Civil Rights Act of 1964, which then was inapplicable to the Federal Government, decisions applying Title VII furnish additional instruction as to the legal standard governing the issues raised in this case. See Part 1(B), infra. The many decisions disposing of employment discrimination claims on constitutional grounds have made no distinction between the constitutional standard and the statutory standard under Title VII. E. g., Davis v. Washington, supra note 1, 352 F.Supp. at 191; Bridgeport Guardians, Inc. v. Civil Serv. Comm’n, 482 F.2d 1333 (2d Cir. 1972); Fowler v. Schwarzwalder, 351 F.Supp. 721 (D.Minn.1972). Moreover, while the Title VII protections did not extend to appellants at the time of intervention, Congress has since amended Title VII to reach charges of racial discrimination in federal employment, Equal Employment Opportunity Act of 1972, Pub.L. No. 92-261, § 11, 86 Stat. 111, 42 U.S.C. § 2000e-16 (Supp. II 1972), and appellants unquestionably are entitled to the benefit of the amendment. See De Rodulfa v. United States, 149 U.S.App.D.C. 154, 164, 461 F.2d 1240, 1250, cert. denied, 409 U.S. 949, 93 S.Ct. 270, 34 L.Ed.2d 220 (1972), and cases cited; Womack v. Lynn, 164 U.S.App.D.C. 198, 504 F.2d 267 (1974).

3

. Davis v. Washington, 348 F.Supp. 15, 17 (D.D.C.1972).

4

. The defendants were Walter E. Washington, Mayor-Commissioner of the District of Columbia; Jerry V. Wilson, Chief of Police at the time of intervention (District appellees); and the commissioners of the United States Civil Service Commission (federal appellees).

5

. The District Court made the determination and direction authorized by Fed.R.Civ.P. 54(b). At the time summary judgment was granted, the claims of discrimination in promotion were still pending. See note 1, supra.

6

. See Part II, infra. We have taken Judge Friendly’s advice and used this term to avoid the pejorative connotations of synonymous phrases. Vulcan Soc’y v. Civil Serv. Comm’n, 490 F.2d 387, 391 n. 4 (2d Cir. 1973). See Boston Chapter, NAACP, Inc. v. Beecher, 504 F.2d 1017, 1019 (1st Cir. 1974); United States v. Chicago, 385 F.Supp. 543, 550 (N.D.Ill.1974).

7

. See Part III, infra.

8

. Davis v. Washington, supra note 3, 348 F.Supp. at 16.

9

. Id.

10

. Appellants’ Appendix at 57 (affidavit of C. Terrence Ireland, ¶ 2). Among the applicants tested outside the District in 1970 and 1971, the failure rate was 47% for blacks and 12% for whites. Id. These data were obtained from the District appellees in discovery proceedings, and include minorities other than blacks in the statistics given for whites. Thus, while “whites and others” would more accurately state the evidence presented to the court, for convenience we will refer to this category as “whites.”

11

. David L. Futransky, Relation of D.C. Police Entrance Test Scores to Recruit School Performance and Job Performance of White and Negro Policemen, mimeograph, November, 1967 (hereinafter cited as Futransky Study).

12

. 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971).

13

. See note 2, supra, and accompanying text.

14

. 401 U.S. at 431, 91 S.Ct. at 853, 28 L.Ed.2d at 164.

15

. Id. See McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802 & n. 14, 93 S.Ct. 1817, 1824 & n. 14, 1824, 36 L.Ed.2d 668, 678 & n. 14 (1973); Bridgeport Guardians, Inc. v. Civil Serv. Comm’n, supra note 2, 482 F.2d at 1337; Officers for Justice v. Civil Serv. Comm’n, 371 F.Supp. 1328, 1336 (N.D.Cal.1973); Smith v. East Cleveland, 363 F.Supp. 1131, 1147 (N.D.Ohio 1973).

16

. See Part II, infra.

17

. Davis v. Washington, supra note 3, 348 F.Supp. at 16.

18

. See Part II, infra.

19

. Davis v. Washington, supra note 3, 348 F.Supp. at 17.

20

. See Part III, infra.

21

. See text supra at note 10.

22

. Castro v. Beecher, supra note 2, 459 F.2d at 729 (passing rate 25% for blacks and 65% for whites); Bridgeport Guardians, Inc. v. Civil Serv. Comm’n, supra note 2, 482 F.2d at 1335 (passing rate for whites 3‘/2 times the black rate); Chance v. Board of Examiners, supra note 2, 458 F.2d at 1171 (passing rate for whites IV2 times greater than passing rate for blacks); United States v. Chicago, supra note 6, 385 F.Supp. at 550 (failure rate for blacks twice the rate for whites); Pennsylvania v. O’Neill, 348 F.Supp. 1084, 1089-90 (E.D.Pa.1972), aff’d in part and vacated in part, 473 F.2d 1029 (3d Cir. 1973) (1.82 times as many whites pass as blacks).

23

. Compare the cases cited supra note 22.

24

. See Davis v. Washington, supra note 3, 348 F.Supp. at 16. There is some disagreement concerning what figures might be relevant in this regard. Appellees contend that the Department recruits primarily within a 50-mile radius of the District of Columbia; they assert that the percentage of blacks in the Department compares favorably with the percentage of blacks in the 20-29 age group in the area of recruitment. Appellants argue, however, that the pertinent population figures are those for the District alone; and since the proportion of blacks in the District is much higher than in surrounding areas, they seek support in the comparison they make. We decline to resolve the dispute, because we deem neither comparison material to this appeal. See Boston Chapter, NAACP, Inc. v. Beecher, supra note 6, 504 F.2d at 1020 n. 4.

25

. Boston Chapter, NAACP, Inc. v. Beecher, supra note 6, 504 F.2d at 1020; Morrow v. Crisler, 479 F.2d 960, 961-62 (5th Cir.), aff’d on rehearing en banc, 491 F.2d 1053 (1974); Carter v. Gallagher, supra note 2, 452 F.2d at 323; Penn v. Stumpf, 308 F.Supp. 1238, 1243 (N.D.Cal.1970).

26

. Bridgeport Guardians, Inc. v. Civil Serv. Comm’n, supra note 2, 482 F.2d at 1335; Chance v. Board of Examiners, supra note 2, 458 F.2d at 1173; Officers for Justice v. Civil Serv. Comm’n, supra note 15, 371 F.Supp. at 1332-33 (N.D.Cal.1973); Smith v. East Cleveland, supra note 15, 363 F.Supp. at 1146; Castro v. Beecher, 334 F.Supp. 930, 935 (D.Mass.1971), aff’d in part and remanded in part, 459 F.2d 725 (1st Cir. 1972).

27

. See cases cited supra notes 25-26.

28

. Griggs v. Duke Power Co., supra note 12, 401 U.S. at 432, 91 S.Ct. at 854, 28 L.Ed.2d at 165 (emphasis in original).

29

. Boston Chapter, NAACP, Inc. v. Beecher, supra note 6, 504 F.2d at 1021; Bridgeport Guardians, Inc. v. Civil Serv. Comm’n, supra note 2, 482 F.2d at 1336; Vulcan Soc’y v. Civil Serv. Comm’n, 360 F.Supp. 1265, 1272 (S.D.N.Y.), aff’d in part and remanded in part, 490 F.2d 387 (2d Cir. 1973); Penn v. Stumpf, supra note 25, 308 F.Supp. at 1244.

30

. Hicks v. Crown Zellerbach Corp., 319 F.Supp. 314, 320 (E.D.La.1970).

31

. United States v. Jacksonville Terminal Co., 451 F.2d 418, 443 (5th Cir. 1971), cert. denied, 406 U.S. 906, 92 S.Ct. 1607, 31 L.Ed.2d 815 (1972); Smith v. East Cleveland, supra note 15, 363 F.Supp. at 1146; Western Addition Community Organization v. Alioto, 330 F.Supp. 536, 540 (N.D.Cal.1971). See also Johnson v. Goodyear Tire & Rubber Co., 491 F.2d 1364, 1372-73 (5th Cir. 1974).

32

. The District Court noted that factors other than Test 21 may also have limited black recuitment. The court stated that many blacks pass Test 21 but fail to report for duty, “thus showing how small a part the Test per se results in failure to recruit more blacks.” 348 F.Supp. at 16-17. The record does show that only 48% of blacks who passed Test 21 between January 1, 1970 and September 4, 1970 reported for duty. Appellants’ Appendix at 72-73 (affidavit of James M. Murray). The same evidence also reveals, however, that only 49% of the whites who passed Test 21 in the same period reported for duty. To attach any significance to the black failure is to ignore the fact that the same phenomenon occurs among whites. Furthermore, during the same time period, blacks comprised 72% of the applicants taking Test 21 in the District of Columbia, but only 56% of those who passed and 55% of all new officers reporting for duty. If the examination had no disparate effect, one would expect approximately 72%, rather than 55 or 56%, of the successful applicants to be black.

33

. Arrington v. Massachusetts Bay Transp. Auth., 306 F.Supp. 1355, 1358 (D.Mass.1969).

34

. Kirkland v. State Dep’t of Correctional Services, 374 F.Supp. 1361, 1364 (S.D.N.Y.1974).

35

. In every one of the more than twenty reported decisions involving this issue cited in this opinion, the final judicial result has encompassed a conclusion of racially disproportionate impact.

36

. Griggs v. Duke Power Co., supra note 12, 401 U.S. at 430, 91 S.Ct. at 853, 28 L.Ed.2d at 163-164. See also Gaston County v. United States, 395 U.S. 285, 89 S.Ct. 1720, 23 L.Ed.2d 309 (1969); Hobson v. Hansen, 269 F.Supp. 401, 419-21 (D.D.C.1967), aff’d sub nom., Smuck v. Hobson, 132 U.S.App.D.C. 372, 408 F.2d 175 (en banc 1969).

37

. See text supra at note 15. The federal appellees intimated to the District Court that the issue of job relatedness must be reached in this case even in the absence of any showing of racially disproportionate impact. In their memorandum of points and authorities in support of their motion for summary judgment, they acknowledged that 5 U.S.C. § 3304(a) (1970) requires that examinations given pursuant to Civil Service Commission authority be job related. See D.C.Code § 4-103; Note, Legal Implications of the Use of Standardized Ability Tests in Employment and Education, 68 Colum.L.Rev. 691, 729 (1968). Appellants have not claimed that they have been deprived of any rights guaranteed by § 3304(a). Because appointments to the Metropolitan Police Department are generally subject to civil service rules, there is a serious question as to whether the validity of Test 21 should be determined by the Equal Employment Opportunity Commission regulations or by the Civil Service Commission regulations, which may be inconsistent with the EEOC guidelines. Compare 29 C.F.R. §§ 1607.1 et seq. (1973) (EEOC), with 5 C.F.R. §§ 300.101 et seq. (1974) (CSC). The EEOC guidelines have been followed in the decisions dealing with validation of employment tests, and have been recognized as controlling at least one circuit. United States v. Georgia Power Co., 474 F.2d 906, 913 (5th Cir. 1973). See also Griggs v. Duke Power Co., supra note 12, 401 U.S. at 433-34, 91 S.Ct. at 854-855, 28 L.Ed.2d at 165-166. Castro v. Beecher, supra note 2, 459 F.2d at 725; Vulcan Soc’y v. Civil Serv. Comm’n, supra note 6, 490 F.2d at 394 n. 8; Carter v. Gallagher, supra note 2, 452 F.2d at 320; Officers for Justice v. Civil Serv. Comm’n, supra note 15, 371 F.Supp. at 1337; Davis v. Washington, supra note 1, 352 F.Supp. at 191. The EEOC guidelines have also been substantially adopted by the Office of Federal Contract Compliance. 41 C.F.R. §§ 60-3.1 et seq. (1974). In any event, the issue is largely academic in view of our holding in Douglas v. Hampton, 512 F.2d 976, decided today.

38

.For whites, 92% of the recruits with raw Test 21 scores over 60 averaged over 85 in Recruit School, 80% of white recruits who scored from 52 to 60 had that average, and 63% of white recruits who scored from 40 to 51 attained an 85 average. For blacks, 78% of the scorers over 60 achieved an 85 average, 70% of black recruits who scored from 52 to 60 had such an average, and 54% of black recruits who scored from 40 to 51 average over 85 in Recruit School. These figures yield a correlation of .46 for whites and .39 for blacks. Futransky Study, supra note 11, at 2. For the purposes of this appeal, we accept these figures as statistically significant. See Boston Chapter, NAACP, Inc. v. Beecher, supra note 6, 504 F.2d at 1024 n. 13.

39

. Compare Harper v. Mayor & City Council, 359 F.Supp. 1187, 1202-03 (D.Md.), modified and aff’d, 486 F.2d 1134 (4th Cir. 1973), with Pennsylvania v. O’Neill, supra note 22, 348 F.Supp. at 1091. “[I]n many cases the apparent value of tests in predicting training success is spurious in that training success is measured by scores on other paper and pencil tests. In order for tests to be meaningfully evaluated, training success must be measured by some sort of performance evaluation which demonstrates whether high scorers on the original test are actually able to learn more quickly or more effectively to perform the job or jobs in question, rather then simply to score well on a subsequent test.” Cooper & Sobol, Seniority and Testing Under Fair Employment Laws; A General Approach to Objective Criteria of Hiring and Promotion, 82 Harv.L.Rev. 1598, 1649 (1969).

40

. The Recruit School examinations are not in the record, and there is no other basis in the record for confirming or disputing the hypothesis stated in text. Although the textual discussion questions whether appellees have met their heavy burden in this regard, we think it unwise to dispose of the case on that basis. It is uncertain whether any evidence could be generated to test the hypothesis, and to remand for consideration of the issue would not serve the interests of judicial economy in light of the overall result we reach in this case.

41

. Futransky Study, supra note 11.

42

. Id. at 1.

43

. Id. at 2. See note 38, supra.

44

. Indeed, the Futransky Study indicates that the contrary may be true. The study found no significant correlation between final Recruit School averages and job performance as measured by positive and negative incidents —commendation, appointment to a position of responsibility, below-average rating by supervisor, resignation with prejudice and disciplinary action. Id. at 5.

45

. Id. at 8. The difficulties in drawing any inference about job performance of failures solely from the performance of passers of an examination have been judicially noted. Boston Chapter, NAACP, Inc. v. Beecher, supra note 6, 504 F.2d at 1025; United States v. Georgia Power Co., supra note 37, 474 F.2d at 916. If appellees’ study had demonstrated a positive correlation between low scores on Test 21 and failures in Recruit School, the data might have supported an inference that those who score below 40 on Test 21 are more likely to fail in Recruit School. No conclusion could be drawn in this regard, however, because there were no failures in Recruit School.

46

. Davis v. Washington, supra note 3, 348 F.Supp. at 16; Futransky Study, supra note 11, at 8.

47

. Cf. Harper v. Mayor & City Council, supra note 39, 359 F.Supp. at 123 n. 38.

48

. Cf. Pennsylvania v. O’Neill, supra note 22, 348 F.Supp. at 1091.

49

. See id. at 1090 n. 6.

50

. They first rely on a statement in Spurlock v. United Airlines, Inc., 330 F.Supp. 228 (D.Col.1971), aff'd 475 F.2d 216 (10th Cir. 1972), in which the court held the requirement of a college degree job related to aircraft piloting. The court found “a direct and substantial correlation between successful completion of the training program and a college degree.” 330 F.Supp. at 235. Unlike the controlling force placed on Test 21 by the Metropolitan Police Department, however, the degree requirement was waived if an applicant had sufficient flight time, id. at 235; cf. Griggs v. Duke Power Co., supra note 12, 401 U.S. at 436, 91 S.Ct. 849; and the minimum flight time requirement, was found to predict • actual job performance. Id. at 235.

Other cases relied upon by appellees are also distinguishable. In Castro v. Beecher, supra note 26, there was no effort to relate test scores to training performance, 334 F.Supp. at 942; the court merely noted that the police department there involved had not even tried to make such a showing, id. at 942, and we see nothing intimating that the court would have held that a convincing showing of such a relationship validated the examination. Finally, in Buckner v. Goodyear Tire & Rubber Co., 339 F.Supp. 1108 (N.D.Ala.1972), aff'd 476 F.2d 1287 (5th Cir. 1973), the court sustained a test that predicted performance in a training program, but held that the training program was the job: “there is a demonstrable relationship between the tests used and the job (i. e., participation in the four year apprenticeship program) for which the tests are used.” Id. at 1114. Apprentices in Buckner had their education subsidized by the employer, were on the payroll while being trained, and went through an 8000-hour four-year program at a cost to the employer of $36,000 per apprentice. The facts of the case at bar would not justify a similar finding; the cost of training an individual police recruit is hardly close to $36,000 and the training program is relatively short.

51

. Officers for Justice v. Civil Serv. Comm’n, supra note 15, 371 F.Supp. at 1337 (police officers); Smith v. East Cleveland, supra note 15, 363 F.Supp. at 1148-49 (police officers); Harper v. Mayor & City Council, supra note 39, 359 F.Supp. at 1202-03 (firemen); Pennsylvania v. O’Neill, supra note 22, 348 F.Supp. at 1090-91 (police officers).

52

. Pennsylvania v. O’Neill, supra note 22.

53

. Id. at 1090.

54

. Id. See note 44, supra.

55

. Id. at 1091. See note 40, supra.

56

. Id. See text supra at note 45.

57

. Id. See text supra at note 46.

58

. See cases cited note 51, supra.

59

.Section 703(h) of the Civil Rights Act of 1964, 42 U.S.C. § 2000e-2(h) (1970), is the authority asserted by appellees for this proposition. They refer to a statement by Senator Tower, the author of § 703(h), that the provision would allow employers to “give general ability and intelligence tests to determine the trainability of prospective employees.” 110 Cong.Rec. 13,492 (1964). A decision as to whether § 703(h) permits validation by correlation to trainability, rather than job performance, is not a simple or insignificant endeav- or. One professional commentator notes that “from validation studies of tests with training criteria, one must generalize with great caution about the validity of those tests for [job] proficiency. . . . [T]he abilities and traits which are important for success in training are not very similar to those which determine success in the performance of the actual job.” E. Ghiselli, The Validity of Occupational Aptitude Tests 119 (1966). We decline to express any opinion on the merits of this issue on these facts, for appellees have not proved a case that presents the question.

We do note, however, that the legislative history of the 1972 amendments to Title VII of the Civil Rights Act of 1964, see note 2, supra, reveals congressional concern over use of general ability tests in hiring and promotion by the Civil Service Commission. H.R. Rep.No.238, 92d Cong., 1st Sess. 24 (1971), U.S.Code Cong. & Admin.News, 1972, p. 2137.

60

. See Davis v. Washington, supra note 3, 348 F.Supp. at 17. But see Pennsylvania v. O’Neill, supra note 22, 348 F.Supp. at 1103.

61

. Davis v. Washington, supra note 3, 348 F.Supp. at 17.

62

. Id. at 18.

63

. See Douglas v. Hampton, supra note 37, 168 U.S.App.D.C. at 70, 512 F.2d at 984. The dissent correctly notes that Griggs involved a distinguishable job situation, but many subsequent decisions have followed the same analysis when dealing with police officers and fire personnel. See cases cited notes 25, 26 & 29, supra, and note 65, infra.

64

. There has been no attempt to establish that proof of the empirical validity of Test 21. is not feasible. See Douglas v. Hampton, supra note 37, 168 U.S.App.D.C. at 72, 512 F.2d at 986. Perhaps more importantly, there is no evidence that directly relates the constructs of Test 21 to job performance. See id. at 72, 512 F.2d at 986. In the absence of such evidence, general concepts of common sense are inadequate aids to meaningful analysis of the crucial issues in this case.

65

. E. g., United States v. Chicago, supra note 6, 385 F.Supp. at 561; Officers for Justice v. Civil Serv. Comm’n, supra note 15, 371 F.Supp. at 1342; Smith v. East Cleveland, supra note 15, 363 F.Supp. at 1152; Harper v. Mayor & City Council, supra note 39, 359 F.Supp. at 1218; Pennsylvania v. O’Neill, supra note 22, 348 F.Supp. at 1104.

66

. Cf. Steffel v. Thompson, 415 U.S. 452, 461, 94 S.Ct. 1209, 39 L.Ed.2d 505 (1974); Samuels v. Mackell, 401 U.S. 66, 73, 91 S.Ct. 764, 27 L.Ed.2d 688 (1971).

67

. The motion was for a partial summary judgment because it did not involve the promotion issues. See note 1, supra.

68

. We note that the District Court granted the federal appellees’ motion and also dismissed the action as to them, while it did not dismiss as to District appellees. It is unclear why this difference exists, and we intimate no view as to the propriety of the dismissal on grounds not discussed in this opinion.