Opinion
No. 72-1009.
September 7, 1972. Rehearing and Rehearing En Banc Denied November 17, 1972.
A. J. Cooper, Jr., Mobile, Ala., Jack Greenberg, Jeffry Mintz, William L. Robinson, New York City, for plaintiffs-appellants.
Mylan R. Engel, Fred G. Collins, Mobile, Ala., for defendants-appellees.
Appeal from the United States District Court for the Southern District of Alabama.
Before BELL, GOLDBERG and RONEY, Circuit Judges.
Plaintiffs, black officers of the Mobile Police Department, sued the defendants claiming that various practices of the Police Department discriminated against Negro officers on account of their race.
We agree with the plaintiffs' statement contained in their brief that the district court, in granting substantially all relief sought on the subject of racial assignment of officers, in ordering changes to reduce or eliminate the discriminatory impact of seniority and service ratings, and in requiring that instruction in intergroup relations be given to all officers and that the defendants undertake affirmative efforts to recruit black officers, has made possible substantial progress toward the achievement of the elimination of unlawful racial discrimination and the elimination of the vestiges of past discrimination.
Plaintiffs' sole issue on this appeal, however, is that the district court, in fashioning a remedy, did not enjoin the use of a written test, which they contend is discriminatory as to blacks, given to promote officers to the rank of sergeant. The district court found that the test is job-related. We affirm the judgment of Chief Judge Pittman on the basis of his order and decree reported at 331 F. Supp. 1134 (S.D.Ala. 1971).
Affirmed.
ON PETITION FOR REHEARING AND PETITION FOR REHEARING EN BANC
The Petition for Rehearing is denied and the Court having been polled at the request of one of the members of the Court and a majority of the Circuit Judges who are in regular active service not having voted in favor of it, (Rule 35 Federal Rules of Appellate Procedure; Local Fifth Circuit Rule 12) the Petition for Rehearing En Banc is also denied.
GOLDBERG, Circuit Judge, dissents from the denial of the Petition for Rehearing and the Petition for Rehearing En Banc.
It is with great reluctance that I dissent in this case, for I am conscious of the synoptic analysis of the problems surrounding testing procedures and of the enlightened decree entered by the distinguished trial judge. Allen v. City of Mobile, S.D.Ala. 1971, 331 F. Supp. 1134. Despite the innovations and courage implicit in the trial judge's reformation of the hiring practices of the police force of Mobile, however, I am convinced that he stopped just short of this case's Rubicon by failing to wade into the deeper waters and to seine for a more optimal test for police promotions. In addition, I am compelled to conclude from the findings of fact of the able district judge that the traditional requirements of equity mandate more immediate relief than that afforded in the decree affirmed by the majority.
The area of occupational and promotional testing is both new and confusing to the courts. So-called "objective" tests were once hailed as the definitive answer to "subjective," often discriminatory, hiring or promotion procedures. But it has become increasingly clear as analysis becomes more sophisticated that there can be other, much more subtle, forms of discrimination lurking in "objective" testing. It is now recognized that a test can be impeccably "objective" in the manner in which the questions are asked, the test administered, and the answers graded, and still be grossly "subjective" in the educational or social milieu in which the test is set. See generally U.S. Comm. on Civil Rights, For ALL the People . . . By ALL the People (1969); Comment, "Legal Implications of the Use of Standardized Ability Tests in Employment and Education," 68 Colum.L.Rev. 691 (1968).
I am persuaded that neither the able district judge nor the majority of this panel has applied an appropriate standard of review when a court is confronted with the admittedly difficult problem of reviewing tests. I do not know that I can provide here a more appropriate standard, but I can suggest some guidelines in the context of this case that seem to me to confront the deeper issues regarding testing.
A test alone is not talismanic; it should, in my opinion, be placed in its own context of valid predicative force for the appropriate position of skill and, in some circumstances, of its discriminatory effect. This Court has already concluded that promotional tests, as well as hiring tests, are subject to judicial scrutiny. See United States v. Jacksonville Terminal Co., 5 Cir. 1971, 451 F.2d 418. It is beyond question at this point in the nation's history that discriminatory state employment practices are constitutionally invalid, save for those rare cases in which the state can show a substantial interest in maintaining a practice shown to be discriminatory. Appellants have, to paraphrase Mr. Justice Holmes' now diluted dictum, no constitutional right to be policemen. But they do have a constitutional right to "be free from unreasonable discriminatory practices with respect to such employment." Whitner v. Davis, 9 Cir. 1969, 410 F.2d 24, 30. See also Griggs v. Duke Power Co., 1971, 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158; Castro v. Beecher, 1 Cir. 1972, 459 F.2d 725; Chance v. Board of Examiners, 2 Cir. 1972, 458 F.2d 1167; Carter v. Gallagher, 8 Cir. 1971, 452 F.2d 315; Rolfe v. County Board of Education of Lincoln County, Tennessee, 6 Cir. 1968, 391 F.2d 77; Wall v. Stanly County Board of Education, 4 Cir. 1967, 378 F.2d 275; cf. Wieman v. Updegraff, 1952, 344 U.S. 183, 73 S.Ct. 215, 97 L.Ed. 216; Norwalk CORE v. Norwalk Redevelopment Agency, 2 Cir. 1968, 395 F.2d 920. And of course even though police work is unquestionably sensitive, that sensitivity cannot ipso facto justify an unconstitutional procedure. See Washington v. Lee, M.D.Ala. 1966, 263 F. Supp. 327 (three-judge court); aff'd per curiam, 1968, 390 U.S. 333, 88 S.Ct. 994, 19 L.Ed.2d 1212; Morrow v. Crisler, S.D. Miss. 1971, No. 4716 [Feb. 12, 1971, Oct. 4, 1971], appeal pending, No. 72-1139 (written test for highway patrol enjoined as unvalidated); Baker v. City of St. Petersburg, 5 Cir. 1968, 400 F.2d 294; Castro v. Beecher, supra.
The patrolman in the instant case demonstrated beyond question that the Mobile police department was rife with discriminatory procedures, discrimination that the trial judge specifically found and that the majority accepts. Since the Mobile police department was wrenched from its "whites only" status in 1954, only one black patrolman has been promoted to sergeant (in 1962) in a force in which 12.4% of the officers are black and in a city approximately one-third black. It appears from the record that this one sergeant was placed in positions within the department that precluded him from ever supervising any white officers. In addition, the one black sergeant twice took and passed the test for lieutenant, but had not been promoted at the time of the trial.
As a particularly sad example, the trial judge found it necessary from the evidence presented to him to enjoin the use of the epithet "nigger" in the police force; in addition, he found that there was assignment of police beats by race in open defiance of this Court's decision in Baker v. St. Petersburg, supra, and various other overtly racial acts by the police department, all of which are delineated in the district court's opinion. We acknowledge also the district court's observation that a new administration appears to have decreased somewhat the more blatant discrimination in the department.
It is in the context of these findings that this Court must, in my opinion, view the testing issue. The record demonstrates that a significantly larger percentage of black applicants failed the test than did white applicants. Of 94 white applicants, over 60% passed; of 14 black applicants, about 14% passed. It is acknowledged by all parties that the test has a critical impact upon promotion and that failure to achieve a passing grade of 70 precludes promotion altogether. Statistics, of course, are usually not conclusive of a proposition of fact, but "[i]n the problem of racial discrimination, statistics tell much, and Courts listen." State of Alabama v. United States, 5 Cir. 1962, 304 F.2d 583, aff'd per curiam, 371 U.S. 37, 83 S.Ct. 145, 9 L.Ed.2d 112; see also Turner v. Fouche, 1970, 396 U.S. 346, 90 S.Ct. 532, 24 L.Ed.2d 567; Hawkins v. Town of Shaw, 5 Cir. 1971, 437 F.2d 1286, modified en banc, 5 Cir. 1972, 461 F.2d 1171. Although it is not noted in the district court's opinion, the record also shows that the police department's promotion sheets record the race of the applicant alongside the test scores. The district court and this panel are in agreement that the appellants produced during the trial a prima facie case that there was clear racial discrimination in the Mobile police department.
The test in question is described by the district court. Allen v. City of Mobile, 331 F. Supp. at 1141. It is prepared by the National Public Personnel Association of Chicago, a cooperative organization of local and state civil service officers. However, I have examined the record, and I must conclude that the district judge was incorrect when he stated that Dean O. W. Wilson of the University of California, a renowned expert in the field, aided in the preparation of the test in question. The record demonstrates only that Dean Wilson's materials were read by those preparing the test.
The last scores available for use at trial were those recorded in 1968.
The other circuits have found racial discrimination in testing situations where there is not the history and existing milieu of racial discrimination that there is in the instant case. These circuits place more emphasis upon the bare statistics regarding substantial racial difference in rates of passing and promotion than I would find necessary in this case. See Castro v. Beecher, supra; Chance v. Board of Examiners, supra; Carter v. Gallagher, supra. Given the pronounced racial effect in the sergeants' test, accompanied by the findings of the district court that the majority now upholds regarding the long and deep history of racial practices in the police department, I would conclude that appellants have made a prima facie case that the sergeants' test is a part of the department's unconstitutional action. And if a prima facie constitutional violation is demonstrated, it is unnecessary, as a general proposition, that the plaintiff also establish a discriminatory intent on the part of the offending persons. See Whirl v. Kern, 5 Cir. 1969, 407 F.2d 781, cert. denied, 396 U.S. 901, 90 S.Ct. 210, 24 L.Ed.2d 177; Hawkins v. Town of Shaw, supra; Daniels v. Van de Venter, 10 Cir. 1967, 382 F.2d 29; Pierson v. Ray, 5 Cir. 1965, 352 F.2d 213, rev'd on other grounds, 386 U.S. 547, 87 S.Ct. 1213, 18 L.Ed.2d 288; cf. Burton v. Wilmington Parking Authority, 1961, 365 U.S. 715, 81 S.Ct. 856, 6 L.Ed.2d 45. I would also argue, however, that the record and the district court's opinion and decree evince some conviction that the police department's procedures were, at least in part, discriminatory by intent. See, e.g., Allen v. City of Mobile, 331 F. Supp. at 1138. I do not, however, base my dissent upon a specific finding of so-called "intentional" discrimination by the department.
I realize that the "compelling state interest" test in all of its ramifications has not yet been applied to situations involving so-called "unintentional" discrimination, and I do not analytically approach this dissent with the idea that this is a so-called "intent" case. Nevertheless, I do not fully agree with distinctions often drawn in similar cases between "intentional" and "unintentional" racial discrimination. See, e. g., Chance v. Board of Examiners, supra; Castro v. Beecher, supra. It appears to me that "motive" is often simply another way of stating that the statistical evidence and the context in which the statistics are set are sufficient to allow, if not compel, a prima facie inference of "intent." See Swain v. Alabama, 1965, 380 U.S. 202, 85 S.Ct. 824, 13 L.Ed.2d 759. Similarly, one who employs a test that unerringly produces greatly divergent results among applicants of different races and who makes no attempt whatsoever to study or to justify the reasons for that divergence can reasonably be said to employ a discriminatory test with "intent." His "intent" need not necessarily be the less for purposes of enforcing the Constitution simply because he continues to use a device with known discriminatory effect rather than choosing to announce openly his discriminatory employment devices or to couch such devices in methods less subtle than testing:
"[W]e now firmly recognize that the arbitrary quality of thoughtlessness can be as disastrous and unfair to private rights and the public interest as the perversity of a willful scheme."
Hobson v. Hansen, D.D.C. 1967, 269 F. Supp. 401, 497.
The degree of "justification" by the state to maintain a process or device discriminatory in fact cannot turn simply upon the fact that one practice might have been transcribed into statute and another practice followed unerringly in fact. See Johnson v. State of Virginia, 1963, 373 U.S. 61, 83 S.Ct. 1053, 10 L.Ed.2d 195; Lombard v. State of Louisiana, 1963, 373 U.S. 267, 83 S.Ct. 1122, 10 L.Ed.2d 338; Cisneros v. Corpus Christi Independent School Dist., 5 Cir. 1972, 459 F.2d 13 [1972] (en banc); United States v. Texas Education Agency, 5 Cir. 1972, 467 F.2d 848 [1972] (en banc); cf. Hawkins v. North Carolina Dental Society, 4 Cir. 1965, 355 F.2d 718; Cypress v. Newport News Gen. and Nonsectarian Hosp. Ass'n, 4 Cir. 1967, 375 F.2d 648. Enforcement of the Fourteenth Amendment's prohibitions against racial discrimination is not a matter of "punishing" those "guilty" of discrimination, and accordingly the degree of justification required in discrimination cases should not turn upon the relative degree of "offensiveness" among perpetrators of racially discriminatory acts. If the prohibition of racially discriminatory acts is far from "punishment" but is rather the enforcement of constitutional rights and responsibilities under the Fourteenth Amendment, then perhaps "intent" should mean nothing more than the knowing perpetration of a racially discriminatory act or practice. To attempt to differentiate the burden of proof that is required to justify a discriminatory act upon the existence or degree of bad motive on the part of the perpetrator of the act seems to me to focus upon an unworkable issue and to ignore the entire thrust and purpose of the Fourteenth Amendment. However, I note again that the lack of a specific finding of so-called "intent" either by the trial judge or by this panel does not reflect in any way the substance of my dissent. I have approached the department's action as "unintentional," in the previously discussed meaning of that term.
The real issue of this case with regard to testing becomes one of establishing a standard of review to be applied to a test when the issue of racial discrimination is adequately and substantively raised. I am guided in my analysis by cases decided under Title VII of the Civil Rights Act of 1964, 42 U.S.C.A. § 2000a et seq., although Title VII is not specifically applicable to a local police department, 42 U.S.C.A. § 2000e. It appears to me, however, that the rationale of Title VII, as elucidated in Griggs v. Duke Power, supra, provides a strong analogy to similar issues that are raised under the aegis of the Constitution and 42 U.S.C.A. § 1983 with regard to actions by state or local employees. Accord, Allen v. City of Mobile, supra; Chance v. Board of Examiners, supra; Carter v. Gallagher, supra; Castro v. Beecher, supra. I cannot conclude that constitutional rights litigable under section 1983 would be entitled to significantly less thorough examination than rights founded upon congressional statute. In addition, I note that the Equal Employment Opportunity Act of 1972 has amended Title VII to include, among others, precisely the employees in question in the instant case.
The district judge should first examine the passing spread of the test, as the able district judge in this case did. If there is a substantial difference between black and white applicants in the rates of passing in the context of other evidence of racial discrimination, then the offending person or persons should be required to establish reasons for utilizing the tests. Put another way, when a prima facie case for unconstitutional action on the basis of race has been made, the burden should be upon the police department to justify its reasons for continuing actions that have been adequately called into constitutional question. See, e. g., Chance v. Board of Examiners, supra; Castro v. Beecher, supra; Carter v. Gallagher, supra. A peculiar result would ensue if a private defendant in a case alleging racial discrimination in employment or promotional practices were required to assume the burden of demonstrating the validity of suspect practices, and a public defendant allegedly engaging in precisely the same practices were not. See Title VII of the Civil Rights Act of 1964, 42 U.S.C.A. § 2000a et seq., and amendments to Title VII by the Equal Employment Opportunity Act of 1972. In addition, it is the police department that, presumably, readily has the necessary information to justify its own procedures.
It is at this point in the review of a test that my divergence with the majority is greatest. I am of the opinion that the district court and the majority applied a standard of "justification" that required much too little of the police department. The district court required only that the sergeants' test be rationally "job related," citing Griggs v. Duke Power Co., supra. Allen v. City of Mobile, 331 F. Supp. at 1146. I too am of the opinion that the rationale of Griggs should apply to a discrimination case brought under section 1983 with force at least equal to its application to Title VII cases, but I believe that the district court and the majority have misconstrued the thrust of that seminal decision and, by analogy, the constitutional requirements regarding promotional practices. If ". . . the jobs in question formerly had been filled only by white employees as part of a long-standing practice of giving preference to whites . . ." and the test operates ". . . to disqualify Negroes at a substantially higher rate than white applicants," Griggs v. Duke Power, 401 U.S. at 426, 91 S.Ct. at 850, 28 L.Ed.2d at 161, then the police department should have to prove that the test bears ". . . a manifest relationship . . ." to the position of police sergeant. Griggs v. Duke Power, 401 U.S. at 432, 91 S.Ct. at 849, 28 L.Ed. 2d at 165 [emphasis added].
Like the district court and the majority, I entertain no doubt that the sergeants' test was "rationally related" to the job of being a sergeant. In fact, I would find it difficult to envision a test that was not somehow "rationally related" to the task of being a police sergeant. But to stop at this low denominator seems to me to ignore both any sort of purposive analysis of testing and the thrust of Griggs. Because the racially discriminatory context and effect of the instant test have been established, I am of the opinion that the state must demonstrate a substantial interest in maintaining the use of such a test. The First and Second Circuits have recently adopted standards very similar to those I propose here. See Castro v. Beecher, supra, (also a police department case); Chance v. Board of Examiners, supra (supervisory positions in a school system); see also Carter v. Gallagher, supra, (firemen). While the standard I propose approaches the "compelling state interest" test often appropriate in cases of racial discrimination, compare McGowan v. Maryland, 1960, 366 U.S. 420, 81 S.Ct. 1101, 6 L.Ed.2d 393, with Loving v. Virginia, 1967, 388 U.S. 1, 87 S. Ct. 1817, 18 L.Ed.2d 1010 (intentional discrimination case), in terms of the degree of proof required to justify continuing the discriminatory impact of a test, I am in agreement with the First Circuit that the test is inappropriate "[t]o the extent . . . that ["compelling state interest"] connote[s] a lack of alternative means" under the facts of this case. Castro v. Beecher, 459 F.2d at 733; see also Carter v. Gallagher, supra; Chance v. Board of Examiners, supra; Penn v. Stumpf, N.D. Cal. 1970, 308 F. Supp. 1238; supra note 4. The police department should not be required to select an "optimal" test so as to erect some form of racial balance in its sergeants' staff. See 42 U.S.C.A. § 2000e-2(j). However, the department should be required to make a substantial showing of job-relatedness, which it was clearly not required to do by the district judge or by the majority.
It appears that there are a number of methods available by which to evaluate substantial job-relatedness, see generally, Cooper Sobol, "Seniority and Testing Under Fair Employment Laws: A General Approach to Objective Criteria of Hiring and Promotion," 82 Harv.L. Rev. 1598 (1969); Chance v. Board of Examiners, supra, but there appear to be two major methods:
"One is `content validation,' which requires the examiners to demonstrate that they have formulated examination questions and procedures based on an analysis of the job's requirements, usually determined through empirical studies conducted by experts. . . . The other method of evaluating job-relatedness is `predictive validation,' which requires a showing that there is a correlation between a candidate's performance on the test and his actual performance on the job."
Chance v. Board of Examiners, 458 F.2d at 1174. It appears that the district court attempted to utilize a "content validity" examination, for the opinion looks only to the questions themselves. The court below was faced with contrasting expert testimony, as is becoming usual in cases requiring any sort of expert opinion. Neither expert witness had any actual knowledge of the Mobile police department itself. The state's expert witness, however, was also the originator of the test. Allen v. City of Mobile, 331 F. Supp. at 1141. Under cross-examination he testified that no studies whatsoever had been made by his firm with regard to the possible racially discriminatory effect of the sergeants' test. In the context of demonstrated racial prejudice in other facets of the department's procedures and a prima facie case that the test itself is racially discriminatory, I would require that the authority offering the test at least conduct studies regarding racial impact before that authority could convincingly insist that the test has "content validity." In addition, it appears that the test was based entirely upon a so-called "job description" prepared in 1959 by an outside consulting agency that was not called to testify at the trial below. I fail to understand how a test that is based upon a "description" of thirteen years ago, which description the testing agency itself neither prepared nor updated, can have "content validity" sufficient to pass muster under the facts of this case in the absence of much more substantial "content" analysis than appears in the transcript. Moreover, appellants dispute whether the "job description" itself was adequate, even independent of the fact that it was made quite some time ago. There is no finding with regard to the adequacy of that description upon which the questioned test was based, nor does it appear that any examination of the "job description" ever took place at the trial below. Finally, the testing agency that prepared the tests has not itself made any studies regarding the comparative performance of officers with high and low test scores so as to examine the efficacy of their own test. In sum, I contend that the analysis of "content" in the instant case was completely superficial and wholly insufficient as a matter of law to constitute "content validity."
The state also presented the testimony of the executive director of the personnel board that administered the test, but it appears from the record that the executive director was not conversant with testing analysis and had himself made no personnel studies or analysis regarding the predictive or content validity of the test in question.
I do not intimate in any way that validation must be particularized to the local community by an expert organization. Such a requirement would be a great financial and time burden, impossible to fulfill in some smaller communities. But see NAACP v. Allen, M.D. Ala. 1972, 340 F. Supp. 703, where the district judge himself "localized" a challenged test to some extent by removing some questions and altering others. I do urge, however, that, in addition to those factors that I have just outlined, any so-called "content validation" should be demonstrated by some organization other than that which drew up the test. Without disparaging the expertise or the opinions of test originators in any way whatsoever, the originators do have an interest in maintaining the public integrity of their own inventions. After all, if a reputable organization proffers a test, it is obviously convinced in its own mind that the test is sufficiently valid as to content. That opinion, however well-intentioned, should not provide the sole evidence of validity to a court, for the question presented is simply too important and subtle for such an examination. See Castro v. Beecher, supra; Chance v. Board of Education, supra.
I should note in passing a few of the questions in the test.
"31. The most important rule to remember when questioning children and low-intelligence adults is to
(1) speak clearly. (2) treat them as any other suspect. (3) allow such suspects wide freedom of narration. (4) avoid suggestions.
"32. The success of a patient, well-planned interrogation of a presumed guilty party pleading innocent is based on the assumption that it is
(1) impossible to commit the perfect crime. (2) possible to detect the veracity of the suspect by observing him. (3) difficult to lie consecutively and logically. (4) impossible for the suspect to live with his guilt very long.
"33. Boys aged 10 to 15 can provide reliable testimony and are especially keen observers in areas relating to
(1) phenomena of nature. (2) intimate occurrences. (3) girls of the same age. (4) moral matters.
"39. As a general rule, the first approach to questioning a suspect should be
(1) emotionally confusing. (2) direct and friendly. (3) Stern and authoritarian. (4) indifferent.
"40. To inspire full confidence on the part of his subject it is vital that the interrogator establish that his attitude is one of
(1) dignity and objectivity. (2) belligerence and intimidation. (3) efficiency and aloofness. (4) sympathy and understanding.
"46. Experience has shown that several types of motives predominate in arson cases. That one of the following occurs more frequently than all of the others is
(1) revenge. (2) pyromania. (3) economic gain. (4) intimidation.
"67. In deciding whether a case involving a juvenile delinquent should be referred to a casework agency or to juvenile court, which of the following factors would likely be the last to be considered?
(1) the parents' desire for help. (2) the juvenile's school record. (3) the emotional needs of the juvenile. (4) the number of offenses committed by the juvenile.
"73. The use of narcotic drugs by juveniles seems to progress according to three definite steps. Generally, the first step ultimately to addiction is the use of
(1) alcohol. (2) marijuana. (3) opium. (4) codeine.
"74. The largest number of juvenile delinquents appearing before the juvenile court fall into which one of the following age groups?
(1) 12-14 year age group. (2) 16-18 year age group. (3) 14-16 year age group. (4) 18-20 year age group.
"76. Authorities in the field of criminal behavior know that nearly all confirmed adult criminals
(1) are sooner or later apprehended and punished for their crimes. (2) start their careers as juvenile offenders. (3) are substandard in intelligence. (4) develop as a result of no religious training.
"77. The more effective the police are in reducing the frequency of contact, the more effective they are in reducing exposure to venereal disease. Therefore, health authorities are in agreement that the most effective way to combat the spread of venereal disease is
(1) to suppress prostitution. (2) to legalize prostitution. (3) to require regular medical inspection of all prostitutes. (4) to encourage and sponsor sex education classes in the secondary schools."
These are just a few of the questions, of course. But one could argue convincingly, I believe, that the above questions are
(1) only very tangentially relevant. (2) subject to considerable disagreement among experts. (3) calling for very subjective judgments among close alternatives. (4) based on very specific knowledge not generally available or read. (5) all or any combination of the above.
Moreover, in the context of a showing of substantial racial bias in the police department and in the absence of sufficient "content validation" to explain obvious racial effect, I am of the opinion that any test used to exclude patrolmen from becoming sergeants must be validated by a comparison of performances under a "predictive validation" scheme. The testimony of police officials in the transcript reveals clearly that the department is seeking a predictive effect when it administers tests and other procedures for promotion. The department does not seek only or even primarily to reward past performance or seniority. The district judge, in fact, agreed with this general approach to testing. With regard to written tests the district court's decree ordered that
"[n]ot less than once each year hereafter from the date of this decree the defendants are to submit a written report to the court which consists of a statistical study of promoted officers which will show a comparison between their examination grade and their regular service or performance ratings."
Thus the district court concluded that there should be some empirical relationship between test scores and performance actually demonstrated, and the majority agrees with that conclusion. While I disagree with the district judge and the majority with regard to the quantum of "relationship" that should be required under the facts of this case, I am in agreement with the thrust of this approach. However, once again I am convinced that the district court and the majority have overlooked the nature of a cut-off test. The decree entered below and affirmed here acknowledges that there may be substantial inverse differences in performance between, for example, a sergeant promoted with a test score of 71 and a sergeant promoted with a test score of 90. By the same reasoning, I am compelled to conclude on the basis of this record that there might also be substantial inverse differences in performance between applicants who had test scores of 69 and 71, or for that matter between applicants with test scores of 50 and 90. Without some such empirical study of racial effect and/or performance, I am wholly unable to conclude that the test is "valid" under any validation theory and under the Fourteenth Amendment and the rationale of Griggs v. Duke Power, supra. See United States v. Jacksonville Terminal Co., 5 Cir. 1971, 451 F.2d 418, cert. denied, 406 U.S. 906,
92 S.Ct. 1607, 31 L.Ed.2d 815. There is a set of guidelines already in existence prepared by the Equal Employment Opportunity Commission. See 29 C.F.R. §§ 1607.1-1607.9. In addition, that same organization apparently evaluates large numbers of specific employment and promotion tests. This Court concluded in another employment context that ". . . the safest validation method is that which conforms with the EEOC Guidelines." United States v. Jacksonville Terminal Co., 451 F.2d at 456; see also Griggs v. Duke Power Co., 401 U.S. at 433, 91 S.Ct. 849, 28 L.Ed.2d at 165.
In addition to my great doubts regarding the legal standard applied by the district court and the majority to determine a lack of unconstitutional discrimination in the test in question, I am also convinced that the district judge should have employed immediately effective injunctive relief. It is reasonably well settled at this point that so called "affirmative" hiring measures are sometimes required to offset the effects of past discrimination. See, e. g., United States v. Jacksonville Terminal Co., supra; Carter v. Gallagher, supra; United States v. Ironworkers Local 86, 9 Cir. 1971, 443 F.2d 544, cert. denied, 404 U.S. 984, 92 S.Ct. 447, 30 L.Ed.2d 367; Local 53 of Int. Ass'n of Heat and Frost Insulators and Asbestos Workers v. Vogler, 5 Cir. 1969, 407 F.2d 1047; United States v. Central Motor Lines, Inc., W. D.N.C. 1970, 325 F. Supp. 478; Contractors Ass'n of Eastern Pa. v. Secretary of Labor, 3 Cir. 1971, 442 F.2d 159, cert. denied, 404 U.S. 854, 92 S.Ct. 98, 30 L.Ed.2d 95. There are a number of ways to accomplish immediate relief. See, e. g., Cooper Sobol, 82 Harv.L.Rev. at 1132. I would leave the precise formulation of any methodology of promoting a reasonably flexible number of black patrolmen to the rank of sergeant to the district judge, who has demonstrated clearly both his ability and his objectivity in his determinations of this case, despite my disagreements on these two points of law. I would propose, however, that the immediate promotions that would take place during the interim period (prior to the time that a validated test and other corrective procedures ordered by the district judge could be established) should be made roughly in approximation to the percentage of black patrolmen on the force. I emphasize, however, that this would be only interim hiring and that my proposed rough percentages for promotion should be flexible to meet the circumstances and would be utilized only for purposes of interim reference. See Swann v. Charlotte-Mecklenburg Board of Education, 1971, 402 U.S. 1, 91 S.Ct. 1267, 28 L.Ed.2d 554. My conclusion that more immediate equitable relief is compelled in the instant case is based upon the district court's own findings regarding the Mobile police department, which has been generally immobile in matters of discrimination. Simply stated, no black man has ever been allowed to be placed in a supervisory position over white patrolmen, even the one black patrolman who managed to pass the sergeants' test.
The state argues that United States v. Jacksonville Terminal, supra, is inapposite to the instant case because the Jacksonville Terminal case involved an aptitude test for unskilled workers, while the instant case deals with admittedly very skilled work. I agree that police work is, of course, substantially more sensitive and skilled than unskilled baggage carrying, but I do not agree that this factor decreases the force of the excellent opinion in Jacksonville Terminal regarding immediate relief. The "labor pool" from which the Mobile police department draws for its sergeants consists of its own patrolmen, whose qualifications as police officers have never been questioned during the course of this case. I agree that the rationale of Jacksonville Terminal might be inapposite to the instant case if the potential "labor pool" from which police sergeants were to be promoted for purposes of immediate relief were only the general pool of available labor. However, the pool consists of men already very skilled in the task of being police officers, presumably very much the same skills required of a police sergeant. The rationale of Jacksonville Terminal is correct for the potential baggage-carriers of that case and for the potential police sergeants of this case.
For example, the department could be required to promote one black patrolman for every eight white patrolmen promoted (approximately the ratio of black to white patrolmen). See, e. g., NAACP v. Allen, supra; Carter v. Gallagher, supra; United States v. Ironworkers Local 86, supra; Contractors Ass'n of Eastern Pa. v. Secretary of Labor, supra. Or the district court could adjust any test employed by the department so as to equalize more appropriately the racial effect of the test. See Cooper Sobol, 82 Harv.L.Rev. supra. Such an adjustment would not amount to unequal treatment to the white applicants; rather, it would be a recognition of the effects of unsubstantiated racial or cultural orientation, and a corrective.
I would propose, finally, that this interim promotion scheme also be employed to test empirically the predictive validity of the test in question. The performance of officers promoted partially by means of the test could be compared with the performance of officers promoted on other factors and without reference to the questioned test. Nothing I have stated in dissent should be construed to forbid the department's employment of another test, one more appropriately and thoroughly validated as to predictive force and to content (including racial/cultural analysis). And, of course, any performance testing by the department is subject to review by the district court.
An attempt is now under way to equalize substantially educational opportunities for all the nation's children. If and until that effort reaches some reasonable degree of fruition, the Constitution cannot stand immobile while a generation of working police officers suffers from the continuing operations and effects of racially discriminatory procedures, however subtle. While the majority would require no immediate pruning of the foliated discrimination, and while I would not require immediate scything of every "root and branch," I would commit to the judicious husbandry of the able trial judge some immediate defoliation of the poisonous trees of discrimination so deeply rooted in the Mobile police department. I would reverse and remand to the district court on the two points of law that I have raised.