UNITED STATES DISTRICT COURT FOR THE SOUTHERN DISTRICT OF NEW YORK
March 17, 1977
Guardians Association of New York City Police Department, Inc., et al., Plaintiffs
Civil Service Commission of the City of New York et al., Defendants
The opinion of the court was delivered by: CARTER
Statement of the Case
This action challenges the legality of the hiring and firing practices of the New York City Police Department ("NYCPD"). Jurisdiction is based on Title 28 U.S.C. §§ 1331 and 1343(3), (4) and 42 U.S.C. §§ 1981, 1983 and 2000e-5(f)(3).
Plaintiffs allege that defendants' practices violate the Fourteenth Amendment, 42 U.S.C. §§ 1981 and 1983 and Title VII of the Civil Rights Act of 1964 ("Title VII"), 42 U.S.C. § 2000e et seq.2 Plaintiffs have moved for a preliminary injunction restraining the NYCPD from firing or recalling any police officers until seniority lists are reordered to accord plaintiffs the seniority they would have had but for defendants' discriminatory practices. Plaintiffs also seek certification of a class consisting of black and Hispanic officers who were subject to layoff in June, 1975. For the reasons set forth below, plaintiffs' motions are granted.
This action was commenced May 3, 1976, by an order to show cause seeking a temporary restraining order and a preliminary injunction. The application for a temporary restraining order was denied. A hearing on the application for a preliminary injunction was held on May 12, 1976. At the hearing, defendants moved to dismiss the complaint on the grounds of statute of limitations, laches, and lack of subject matter jurisdiction. Their motion was denied, and the hearing commenced. It continued intermittently until July 29, 1976.
Having heard that the NYCPD intended to recall previously laid off police officers, plaintiffs in the early part of July renewed their application for a temporary restraining order barring such recall pending determination of the rights of the discharged black and Hispanic police officers in this litigation. The court denied this application on the basis of an affidavit of Michael J. Codd, Police Commissioner, which recited that all recalled officers were to be notified that their rights to employment were subject to any determination that might be made in this pending litigation.
Plaintiffs are black and Hispanic police officers on layoff since June, 1975. They allege that the NYCPD's use of its pre-1973 entry level examinations and its former requirement that all police officers be 67 inches or taller discriminated against them, and that but for this discrimination they would have accrued sufficient seniority to withstand being fired. Defendants are the NYCPD, the New York City Department of Personnel, whose staff prepare and administer the examinations, and the Civil Service Commission, which prescribes rules for administering the civil service laws.
Because of the fiscal crisis affecting the entire city, the NYCPD was required to cut its budget by three percent. As a result, 2,864 police officers were discharged in June, 1975, pursuant to the last-hired, first-fired seniority system existing in the NYCPD. Plaintiffs complain that these terminations, 21.5 percent of which were minorities,
perpetuate defendants' past discriminatory practices and violate the strictures of the Fourteenth Amendment, Title VII and 42 U.S.C. §§ 1981 and 1983.
Class Action Certification
For class action treatment to be appropriate, all the requirements of subdivision (a), and one of the alternative requirements of subdivision (b), of Rule 23, Fed. R. Civ. P. must be satisfied.
There is little doubt that plaintiffs have met the first three requirements of Rule 23(a).
1. The Numerosity Requirement.
The proposed class consists of all black and Hispanic New York City policemen currently on layoff who would not have been furloughed but for defendants' allegedly discriminatory employment practices. Plaintiffs make the uncontroverted allegation that the class numbers between 200 and 600 blacks and Hispanics. Given these figures, the numerosity requirement is clearly met. Korn v. Franchard Corp., 456 F.2d 1206, 1209 (2d Cir. 1972); Ste. Marie v. Eastern R.R. Ass'n, 72 F.R.D. 443 (S.D.N.Y.1976) (Carter, J.).
2. Common Questions of Law or Fact.
Race discrimination cases, by their very nature, involve behavior that affects a group of people. Rodriguez v. East Texas Motor Freight Co., 505 F.2d 40, 50 (5th Cir. 1974); Oatis v. Crown Zellerbach Corp., 398 F.2d 496, 499 (5th Cir. 1968). Plaintiffs here claim that defendants' practice of discharging police officers based on their seniority violates Title VII and 42 U.S.C. §§ 1981 and 1983 by perpetuating discrimination effected by the use of racially biased entry level examinations and a discriminatory height requirement. It is clear that these claims:
not only can, but most appropriately should be decided as a class action since "the evil sought to be ended is discrimination on the basis of a class characteristic, i.e., race, sex, religion, or national origin." . . .
Hecht v. Cooperative for Am. Relief Everywhere, Inc., 351 F. Supp. 305, 312 (S.D.N.Y. 1972) (Lasker, J.) (citations omitted).
3. Representative Parties' Claims Typical of the Claims of the Class.
The claims of the representative parties are typical of most of the class. The individual plaintiffs claim to be on layoff either as a result of defendants' discriminatory use of entry level examinations or as a result of defendants' discriminatory height requirement. Plaintiff organizations include members who have been victims of both these practices. Together, their claims range over the totality of defendants' challenged employment practices. Thus, the typicality standard is met.
4. Fair and Adequate Protection of the Class' Interests.
In Mersay v. First Rep. Corp. of Am., 43 F.R.D. 465, 469 (S.D.N.Y. 1968) Judge Metzner held that:
the requirement of adequate representation comprises only two elements: (a) that the interests of the representative party must coincide with those of the class; and (b) that the representative party and his attorney can be expected to prosecute the action vigorously. Dolgow v. Anderson, 43 F.R.D. 472 (E.D.N.Y. 1968).
Defendants presumably contest the adequacy of plaintiffs' representation by contending that in Guardians Ass'n of New York City Police Dep't v. Civil Serv. Comm'n, 72 Civ. 928 (S.D.N.Y., July 16, 1973) (Ryan, J.) ("Guardians I"), a suit brought under 42 U.S.C. § 1983, plaintiff associations took a position incompatible with the one taken here. Defendants argue that in the previous action these associations sought a permanent injunction against the use of lists predating 1972 for appointments to the Police Department. Therefore, it is claimed, since the individual plaintiffs in the instant case obtained appointments from pre-1972 lists "precisely because the relief sought by these associations was denied in Guardians I. . . [the] associations represented interests directly adverse to [those of] the individual plaintiffs herein." Def's. Post-Trial Memo. at 5. This argument is patently without merit.
In the earlier suit, plaintiff sought to enjoin the use of pre-1972 officer eligibility lists out of the fear that a disproportionately low number of minorities would be hired due to the discriminatory impact of the examinations from which the lists were prepared. This is made clear from plaintiffs' alternative request to establish a quota of one minority appointment for every white appointment. The relief sought was denied since it appeared that the lists were being exhausted and all the plaintiffs would be hired.
The premise of that suit is consistent with that asserted here. In both cases, the thrust of plaintiffs' attack is directed toward the use of employment requirements which discriminate against minorities. Accordingly, I am convinced that the interests of the plaintiff associations coincide with those of the class, and based on the extensive work already conducted by plaintiffs, they can be expected to prosecute the action vigorously.
Defendants also argue that plaintiff Santos could not have been appointed as a police officer until he was a United States citizen and that he did not become a citizen until the time of a hiring freeze; therefore he is unable to show that he could have been appointed earlier than he actually was. Plaintiff Perez, defendants assert, challenges the height requirement but has never failed a police officer examination and does not claim the written examination discriminated against him. To the extent these assertions suggest inadequate representation or atypicality of claims, they pose no bar to certification of the class. They may present individual questions which will determine each individual's right to recovery, but Rule 23 gives the court the power to consider individualized problems if and when they appear. Ste. Marie v. Eastern R.R. Ass'n, supra. In any event, since plaintiff organizations are suitable representatives, the potential disqualification of Santos and Perez poses no threat to the viability of this class suit.
5. Action or Inaction on Grounds Generally Applicable to Class -- Appropriateness of Injunctive or Declaratory Relief (Rule 23(b)(2)).
Plaintiffs have asked for declaratory and injunctive relief that would prevent defendants from perpetuating allegedly discriminatory hiring practices. Plaintiffs also seek damages. This situation is well-suited for class action treatment under subdivision (b)(2) of Fed. R. Civ. P. 23, since a finding that defendants' practices are unlawful will necessarily affect all minority police officers. See Advisory Committee Notes to Fed. R. Civ. P. 23, and cases cited therein; Leisner v. New York Tel. Co., 358 F. Supp. 359, 373 (S.D.N.Y. 1973); Rosario v. New York Times, 10 E.P.D. P 10,450 p. 5944 (S.D.N.Y. 1975). A claim for damages may properly be considered in a (b)(2) action as long as the monetary relief is not predominant. Almenares v. Wyman, 334 F. Supp. 512, 519 (S.D.N.Y.), modified 453 F.2d 1075 (2d Cir. 1971), cert. denied, 405 U.S. 944, 92 S. Ct. 962, 30 L. Ed. 2d 815 (1972); Lynch v. Sperry Rand Corp., 62 F.R.D. 78, 85 (S.D.N.Y. 1973).
Accordingly, plaintiffs' motion for certification of a plaintiff class is granted.
Entry Level Examinations
In order to be appointed a police officer, an applicant must, among other things, achieve a 75 percent score on a written examination. Individuals who pass this examination are ranked on an eligible list in order based on their scores.
The highest ranked are the first to be given available jobs. The lowest ranked are the last to get jobs, and consequently are the first to be fired when there are layoffs. Plaintiffs contend that the use of these examinations prior to 1973 violated both their constitutional and statutory rights.
1. Constitutional Claims
In Washington v. Davis, 426 U.S. 229, 48 L. Ed. 2d 597, 96 S. Ct. 2040 (1976), the Supreme Court upheld the validity of a qualifying test administered to applicants for positions as police officers in the District of Columbia Metropolitan Police Department even though the test had a discriminatory impact in screening out black candidates. The Court of Appeals had declared a lack of discriminatory intent irrelevant to establishing a constitutional violation. The Supreme Court reversed, holding that the circuit court "erroneously applied the legal standards applicable to Title VII cases
in resolving the Constitutional issue before it," viz., whether the use of the qualifying examination invidiously discriminated against blacks and hence denied them their due process rights contrary to the commands of the Fifth Amendment. The Supreme Court continued:
The central purpose of the Equal Protection Clause of the Fourteenth Amendment is the prevention of official conduct discriminating on the basis of race. It is also true that the Due Process Clause of the Fifth Amendment contains an equal protection component prohibiting the United States from invidiously discriminating between individuals or groups. . . . But our cases have not embraced the proposition that a law or other official act, without regard to whether it reflects a racially discriminatory purpose, is unconstitutional solely because it has a racially disproportionate impact.
Id. at 239. The rule in other equal protection cases is likewise a requirement that plaintiffs show intentional discrimination, not merely discriminatory impact. Id. at 239-41. It follows that §§ 1981
require the same showing.
While an invidious discriminatory purpose may often be inferred from a practice which affects a definable minority race or group more than the majority, disproportionate impact "is not the sole touchstone of an invidious racial discrimination forbidden by the Constitution." Id. at 242. In the instant case, while the discriminatory impact of entry level examinations seems clear enough, the tests used by the NYCPD, like the examination upheld in Washington v. Davis, appear to be facially neutral and "rationally may be said to serve a purpose the Government is constitutionally empowered to pursue." Id. at 246.
Plaintiffs attempt to show discriminatory intent primarily from the fact that instead of administering a general qualifying examination, the NYCPD used entry level tests which it knew had a discriminatory effect in screening out minority candidates. The proof of intent offered by plaintiffs here differs little from that before the Supreme Court in Washington v. Davis and accordingly must be deemed insufficient.
Defendants here were constrained by the New York State Constitution, Article 5 § 6, and the Civil Service Law § 50 et seq. to select employees for civil service positions based upon individual merit and fitness, to be determined, where practical, by competitive examination. The position of police officer is a competitive title, and appointments are made from eligibility lists compiled on the basis of ratings received by the candidates on the competitive portions of the examinations. Civil Service §§ 50(6) and 61. Absent a showing that the entry level examinations were purposely designed to keep minorities out of the police force, Washington v. Davis requires that such procedures be sustained even if they have the result of screening out minorities.
Defendants have proffered evidence of recruitment efforts on their part to demonstrate their good faith, including the creation of a recruitment squad within the department's Personnel Bureau in 1967; participation in, and adoption of recommendations made by, a nine month Rand Institute Study in 1968-69; the use of a federally funded "Community Service Office" program, employing mostly minorities in quasi-police positions for the purpose of absorbing them into the police officer ranks; a cadet program begun in 1966 which paid minority youths receiving training in the Police Academy; and the establishment of an Equal Employment Opportunity Office. As a result of their commitment to increase minority representation, defendants claim that 23 percent of all new police recruits hired since 1973 were minority members.
Plaintiffs challenge the defendants' assertion of bona fides with respect to minority recruitment. They argue that prior to 1970 the NYCPD made no real efforts to increase the number of minority police, nor was support given to recruitment efforts made by plaintiffs. Moreover, plaintiffs charge, the police department's recruitment program began in earnest in 1972, only after Title VII became applicable to municipalities and litigation was threatened. Finally, it is argued, defendants' recruitment efforts began after a freeze was placed on hiring, and whatever gains had been made were later wiped out by the layoffs.
Even were these allegations enough to controvert defendants' evidence of good faith with respect to recruitment, they are nonetheless insufficient to meet the intentional discrimination requirements of Washington v. Davis. Mere absence of recruitment efforts, by itself, is not equivalent to an intent to discriminate. Also, even if the police department's stepped-up recruiting program can be directly linked to the threat of litigation after Title VII became applicable to municipalities, this alone is still no evidence of discriminatory intent. The applicability of Title VII to municipalities would establish prima facie liability on the part of the police department upon a showing that its hiring practices had a discriminatory impact regardless of intent. Taking steps to avoid such liability cannot justifiably be characterized as tantamount to admitting discriminatory intent. Nor is it reasonable to impugn the police department's recruitment efforts after 1972 because the City's financial problems required budget cuts and the imposition of a job freeze. The fact that subsequent layoffs by seniority eliminated recruitment gains may make out a Title VII cause of action, but it cannot stand for proof of invidious discrimination proscribed by the 14th Amendment to the Constitution of the United States under the Washington v. Davis rationale.
The strongest evidence of defendants' discriminatory intent was the testimony of Police Commissioner Murphy and Deputy Commissioner Ward. Murphy testified that the NYCPD and the Department of Personnel did not try as hard as he would have liked to increase minority representation on the police force, but he never stated that either actively discriminated against blacks and Hispanics. In fact, he characterized the department's actions to recruit and retrieve minorities at least during his tenure (Oct., 1970 to May, 1973) as good faith efforts. Tr. 1130-32. He also stated that there was "a Civil Service effort to improve as much as it could the written examination." Tr. 1132.
The testimony of Deputy Commissioner Ward is a more unqualified condemnation of the defendants' motivations. What he has characterized as intentional discrimination, however, is the complacency of officials in the police department and the Department of Personnel with respect to the status quo and their failure to take more active steps toward increasing minority representation. Taken as a whole this testimony falls far short of establishing liability under the Constitution or §§ 1981 and 1983.
2. Title VII Claim12
a. Applicability of Title VII.
Plaintiffs challenge the job-relatedness of all pre-1973 entry level examinations.
However, the specific harm complained of in this action is not the decision to hire or not to hire based on the results of these pre-1973 examinations, but rather the decision to layoff police officers as a result of budget cuts. The layoffs, as indicated previously, were made in 1975 pursuant to a last-hired, first-fired seniority system which no one attacks as discriminatory in any way. Defendants claim that since the only acts complained of took place prior to the date Title VII was made applicable to municipalities, the standards of Title VII do not apply to this case.
Additionally, defendants contend that plaintiffs have not met the procedural requirements of Title VII, since neither any individual plaintiff nor the Guardians and Hispanic Associations filed timely charges with the Equal Employment Opportunity Commission ("EEOC").
Decisions in this circuit and elsewhere, however, do not support defendants' position. To the contrary, the layoff of police officers in 1975 brings the pre-1972 alleged discriminatory practices within the scope of Title VII. In Acha v. Beame, [11 EPD P 10,740] 531 F.2d 648 (2d Cir. 1976), two women who were formerly police officers brought suit under Title VII and under the Fourteenth Amendment, alleging that threatened NYCPD terminations unlawfully discriminated against women.
The plaintiffs in Acha alleged, inter alia, that the facially neutral seniority system
perpetuated the past discriminatory practices of the Police Department, and that the layoffs violated their Title VII statutory rights.
Acha did not specifically deal with the contention raised here by defendants that Title VII does not apply since the only post-Act practice complained of is the application of a neutral seniority system. Nevertheless that contention was implicitly rejected by the Court of Appeals.
Plaintiffs' complaint had been dismissed by this court (Duffy, J.) primarily on the ground that Section 703(h) of Title VII, 42 U.S.C. § 2000e-2(h) excepts from the sanctions of Title VII the application by an employer of "different terms, conditions or privileges of employment pursuant to a bona fide seniority . . . system." Judge Duffy had also held that section 703(j) of Title VII, 42 U.S.C. § 2000e-2(j), protected the defendants since to grant plaintiffs' requested relief would constitute unlawful preferential treatment on the basis of sex.
The Court of Appeals reversed, holding that neither section 703(h) nor 703(j) of Title VII insulated the defendants from a valid Title VII claim. The court stated:
If a female police officer can show that, except for her sex, she would have been hired early enough to accumulate sufficient seniority to withstand the current layoffs, then her layoff violates section 703(a)(1) of Title VII, 42 U.S.C. § 2000e-2(a)(1), since it is based on sexual discrimination. . . .
Until the past discrimination against these particular plaintiffs is remedied by according them the seniority position to which they are entitled, the system cannot be considered "bona fide" and in fact represents a continuation of past intentionally discriminatory practices, and this falls outside the terms of section 703(h). 531 F.2d at 654-55 (emphasis added).
The court also disagreed with Judge Duffy's holding as to Section 703(j) and held instead that an award of seniority to those who had been actually discriminated against would not constitute a "preference", but would rather be a remedial action well within the power delegated the courts by Title VII. Acha, supra, at 656.
Although defendants' argument here is a variation of that used in Acha, the point remains the same. If defendants' position in either case were correct, Title VII would be unavailable to individuals seeking to remedy past discrimination unless they could show discrimination in hiring by a municipality which occurred after the passage of the March 24, 1972 amendment to Title VII. This position has been rejected by federal courts on numerous occasions. In Griggs v. Duke Power Company, 420 F.2d 1225 (4th Cir. 1970), rev'd in part on other grounds 401 U.S. 424, 91 S. Ct. 849, 28 L. Ed. 2d 158 (1971), it was expressly held that while Title VII was intended to have prospective application only, it did provide relief based on pre-Act discrimination if there are present and continuing effects of such discrimination. 420 F.2d at 1230. See also, Local 53 of Int'l Ass'n of Heat and Frost Insulation and Asbestos Workers v. Vogler, 407 F.2d 1047 (5th Cir. 1969); United States v. Local 189, United Papermakers and Paperworkers, AFL-CIO, CLC, 282 F. Supp. 39 (E.D. La. 1968), aff'd, 416 F.2d 980 (5th Cir. 1969), cert. denied, 397 U.S. 919, 25 L. Ed. 2d 100, 90 S. Ct. 926 (1970); Cates v. Trans World Airlines, F. Supp. , Civ. No. 73-5487 (S.D.N.Y., July 22, 1976) (Brieant, J.).
The Supreme Court provided support for the lower court's position on the question. The Chief Justice in his opinion in Griggs stated:
The objective of Congress in the enactment of Title VII is plain from the language of the statute. It was to achieve equality of employment opportunities and remove barriers that have operated in the past to favor an identifiable group of white employees over other employees. Under the Act, practices, procedures, or tests neutral on their face, and even neutral in terms of intent, cannot be maintained if they operate to "freeze" the status quo of prior discriminatory practices. 401 U.S. at 429-30.
Griggs and the cognate cases all involved the application of the pre-1972 version of Title VII. Nevertheless, their rule applies by analogy to the instant case. As this court has recently held:
[If] the plaintiffs allege that [defendants] engaged in activities prior to the effective date [of the 1972 Amendment] which had a continuing discriminatory effect past the effective date, those continuing effects constitute a violation of the Act and give rise to a cause of action. Essentially, it was the obligation of governmental employers after the effective date of the 1972 amendments to examine their past activities to ensure that any discriminatory effects of those activities did not continue in violation of the Act.
Melani v. Board of Higher Educ. of the City of New York, 12 EPD P 11,068 at 4962 (S.D.N.Y. 1976) (Gagliardi, J.).
Defendants correctly point to Monell v. Department of Soc. Servs. of the City of New York, 532 F.2d 259 (2d Cir. 1976) as establishing the circuit rule in respect of the "retroactivity" of Title VII. Monell involved a claim by female employees of the New York City Department of Social Services and of the New York City Board of Education that rules and regulations of the city agencies which formerly compelled pregnant employees to take unpaid leaves of absence before medical reasons required any absences, were, inter alia, violative of Title VII. The court (Metzner, J.) had dismissed the Title VII claims for back pay covering the periods which plaintiffs alleged they could have been working but for the challenged rules. The dismissal was based on the court's determination that the 1972 amendment to Title VII could not be applied retroactively, and the fact that all the acts of discrimination complained of had occurred prior to the 1972 amendment.
The Court of Appeals affirmed, holding:
[The] 1972 amendment to Title VII does not apply retroactively so as to permit an award of back pay against the city in this case. The same law, Publ. L. No. 92-261, section 2(1) of which amended the definition of "person" to include "governments, governmental agencies [and] political subdivisions," also struck out in section 3 the exemption for the employment of individuals engaged in educational activities of nonreligious educational institutions. In Weise v. Syracuse University, 522 F.2d 397, 410-11 (2d Cir. 1975), we met the question whether Title VII was applicable to discrimination occurring before the lifting of the exemption for educational institutions. We held that it was not retroactive, for it was "a statute creating new rights where none had previously existed." . . . We see no valid ground for distinguishing the sex-discrimination claim in Weise [here . .. where] entirely new substantive rights were created by the amendment to Title VII.
532 F.2d at 261-62. See also Cleveland Bd. of Educ. v. LeFleur, 414 U.S. 632, 639 n.8, 94 S. Ct. 791, 39 L. Ed. 2d 52 (1974); Melani v. Board of Higher Educ. of the City of New York, supra.
Monell, however, does not deal with the issue of equitable relief where an employer's post-amendment practice perpetuates and continues pre-amendment discrimination, as is alleged here. It merely limits the Title VII of a municipality to damages sustained by victims of discrimination after the 1972 Title VII amendments.
Monell precludes the plaintiffs from recovery of money damages under Title VII which resulted from defendants' unlawful hiring practices. It, does not, however, bar equitable relief made necessary because of pre-amendment discrimination. Indeed, such a bar to equitable relief would run totally counter to the congressional purpose not to effectuate a freeze of employees in positions to which they were relegated by past discrimination predating enactment of the fair employment practices legislation. Griggs v. Duke Power Co., supra, 401 U.S. at 429-30; Quarles v. Philip Morris, Inc., 279 F. Supp. 505, 516 (E.D. Va. 1968).
b. Title VII Standard.
In Griggs v. Duke Power Co., supra, the Supreme Court held that Title VII forbids the use of employment selection practices that have a racially discriminatory impact unless the employer meets "the burden of showing that [the selection practices have] . . . a manifest relationship to the employment in question." Id. at 432. This burden arises only after the complaining party or class has made out a prima facie case of discrimination. The Court held in Albemarle Paper Company v. Moody, 422 U.S. 405, 45 L. Ed. 2d 280, 95 S. Ct. 2362 (1975) that a prima facie case is made out if the complainants can show "that the tests in question select applicants for hire or promotion in a racial pattern significantly different from that of the pool of applicants." Id. at 425. See also McDonnell Douglas Corp. v. Green, 411 U.S. 792, 36 L. Ed. 2d 668, 93 S. Ct. 1817 (1973). A showing of discriminatory purpose need not be made under Title VII. Griggs, supra; Albemarle, supra. Based on the evidence adduced at the hearing I find that while the plaintiffs have made a prima facie showing of discrimination, the defendants have not met their burden of showing job-relatedness.
c. Disproportionate Impact.
At the May 24, 1976, hearing the court ruled that plaintiffs had sustained their burden of establishing that the examinations had a racially discriminatory impact, and that the burden therefore shifted to defendants to show job-relatedness. The following constitutes the court's basis for so ruling.
In the wake of Guardians I, the Rand Institute ("Rand") was authorized by all parties to that lawsuit to conduct an independent study to determine the distribution by race of civil service scores on NYCPD entry level examinations #8046, given July 20, 1968, and/or #0013, given May 9, 1970. The study was paid for by the City of New York and the Fund of the City of New York.
Rand attempted to determine the respective passing rates of whites, blacks and Hispanics
on NYCPD entry level examination #8046, the earliest examination with an eligibility list still in use, and examination #0013, the latest examination then in use. Both parties to Guardians I were aware that these were the only examinations to be studied.
Before Rand could complete such a study it had to determine the race of those applicants who took these examinations. Because the NYCPD did not keep systematic records of the race of the applicants, Rand was forced to rely upon indirect methods of identification. Five different sources of data were used: (1) NYCPD records giving the race of about 30 percent of the applicants taking each examination; (2) results of a questionnaire inquiring into race which was mailed to nearly all persons who took the examinations; (3) a compilation of approximately 8,000 surnames compiled by the U.S. Bureau of Census with which the subjects' names were compared; (4) census tract data, evidencing racial residential concentration with which the subjects' addresses were matched; and (5) a telephone sample of the subjects whose race remained unknown after the NYCPD data and the questionnaire data were on hand.
Using these data, Rand estimated the numbers of whites and blacks who took the examinations in two ways, and the number of Hispanics in three.
These estimates, which Rand found to fall well within the range of acceptable statistical error, indicated that about 71 percent of the examinees of both examinations were white, and 29 percent were minority-group members (a larger proportion of this 29 percent were Hispanics for examination #0013 than for #8046).
Similarly, Rand used three methods to estimate the distribution by race of the scores on examination #8046 and #0013.
Under each method the result was substantially the same. Whites consistently passed at higher rates than minorities and consistently comprised the overwhelming majority of individuals in the top deciles. Minorities consistently comprised the overwhelming majority of individuals in the bottom deciles. In summarizing its findings, Rand stated that:
[we] have found substantial and statistically significant differences by race in the distributions of exam scores for whites as compared to blacks and Hispanics for both exam #8046 and exam #0013. Blacks and Hispanics scored lower than whites and failed at higher rates -- No passing grade could have been set to result in equal proportion of whites and minority group members passing. These findings were confirmed by three different methods of data analysis. Rand Report at 46.
At the hearing, Dr. Cohen testified that given a racially neutral text [test] the chances of this result occurring were statistically less than one in a billion. Tr. 45-46. Rand Report, Pl's. exh. 1 at vi.
Because of the similarity of results for the studies of the two examinations, which were given two years apart, Rand concluded that it was
confident that the patterns observed for exam #0013 are not unique to that exam but are most likely typical of the patterns for all recent civil service patrolmen's exams.
Rand Report at 38; Testimony of Dr. Cohen, Tr. 47, 48.
Plaintiffs assert that there is additional evidence which supports their claims of racially discriminatory impact.
Because the court believes that the results of the Rand Report are sufficient to make out a prima facie case, however, this other evidence need not be considered.
Furthermore, it is unnecessary to determine whether plaintiffs or defendants are correct in their assertions as to the proper makeup of the applicant pool. Plaintiffs seek to prove that minorities are grossly underrepresented in the NYCPD. Defendants contend that the statistics demonstrate otherwise. However, even if the result were overrepresentation of minorities, that would have little relevance to the Title VII issue presented here. All that is challenged in this case is the selection practice that occurs at the point that members of the applicant pool actually show up and take the entry level examinations. Even if, as defendants contend, the percentage of minority policemen is nearly equivalent to the percentage of minorities in the relevant applicant pool, the fact remains that utilization of the examinations challenged has had the effect of significantly decreasing what might otherwise be the number of minorities who would, under NYCPD procedures, be eligible for appointment. The key question that must be faced is given the number of whites, blacks and Hispanics who took the NYCPD entry level examinations, whether the number who passed or failed and the scores they achieved were to a legally significant degree determined by their race.
Defendants have at all times contested the statistical basis for the finding of a prima facie case of discrimination. At the May, 1976 hearing they called Professor Alan Gross, an associate professor in the department of educational psychology at the graduate center of City University. Dr. Gross testified that he did not "think any conclusions at all can be drawn from [the Rand] study because of the methodology used. . . ." Tr. 578. Dr. Gross proceeded to analyze each of the methods used by Rand to reach its conclusions and found each severely flawed. In analyzing Rand's use of the census records, for example, Dr. Gross stated:
[Rand] looked at a tract, a region. Supposing this region, you find according to census records that 90 percent of all individuals are black and 10 percent are not -- this is known from census records -- and therefore what the authors of the report tried to do was, they looked at one of those people they couldn't contact in the telephone survey, but they knew what tract, census tract this person lived in. They said if this person lives in a census tract that is 90 percent Hispanic, then we can assume that this person is Hispanic and although we might be wrong, 10 percent, that's not an unreasonable rate and we'll call anybody in this census tract, will assume is Spanish.
There is a big problem with that and a flaw in that kind of reasoning. . . . A person from a tract 90 percent Spanish, is Spanish. If a person comes from the tract and is taking the police examination, what are the chances he is Spanish? If a greater proportion of whites apply than do Spanish individuals, then the probability is not .90 more. It goes down considerably. Tr. 591-93.
Dr. Gross was similarly critical of the other methods used by Rand in its study. As another example, Dr. Gross stated that there was no theoretical basis to extrapolate from the results of the study of the two examinations a conclusion relating to the results of all the examinations that were given in between. In short, Dr. Gross testified that the Rand report was based on "[assumptions] that there is no basis for defending. They may be right and they may be wrong." Tr. 619.
Dr. Gross' opinions are of course quite relevant. I am not persuaded, however, by his arguments. Although the Rand study may be based in minor part on certain unproved assumptions, viewed as a whole its findings are clearly trustworthy. As the report indicates, its authors recognized that they were confronted with certain methodological problems. Accordingly, they used three different statistical methods in order to ensure that their findings would be accurate. Dr. Gross' criticisms might have been telling had Rand's conclusions been based on only one of the methods used. However, regardless of which of the three statistical methods the authors used, they came up with substantially equivalent results. I have no hesitancy in concluding that this consistency is more than sufficient to remove any doubt as to the reliability of Rand's methods and findings.
My faith in the accuracy of Rand's findings is based additionally on the fact that the Rand study was conducted entirely independently of the parties to this litigation, and that there is nothing whatsoever to indicate that the authors of that study had any desire or predilection to reach a particular conclusion. Moreover, there can be little dispute that the Rand study was conducted in a thoroughly professional manner by experts in the field. Indeed, although the defendants are not estopped from challenging the methodology used by Rand,
it is significant that the defendants trusted Rand's technical capabilities sufficiently to authorize its study of the problem in the first place.
In sum, I find that the results of the Rand study are reliable and by themselves are sufficient to establish a prima facie case.
d. Defendants Have Not Shown the Tests to be Job-Related.
Since plaintiffs have established a prima facie case of discrimination, defendants must show that the tests in question are job-related. Griggs v. Duke Power, supra; Albemarle Paper Co. v. Moody, supra; Chance v. Board of Examiners, 458 F.2d 1167 (2d Cir. 1972). To do so, defendants must show that the entry level examinations "are demonstrably a reasonable measure of job performance." Griggs, supra, at 436.
The parties agree that there are three generally accepted measures of test validation -- content, construct and criteria-related validity.
Defendants contend that the NYCPD exams are demonstrably validated according to both the content and criteria-related standards.
[i] Content Validity.
In Vulcan Soc'y of the New City Fire Dep't, Inc. v. Civil Serv. Comm'n, 360 F. Supp. 1265 (S.D.N.Y.), aff'd, 490 F.2d 387 (2d Cir. 1973), Judge Weinfeld described content validity:
An examination has content validity if the content of the examination matches the content of the job. For a test to be content valid, the aptitudes and skills required for successful examination performance must be those aptitudes and skills required for successful job performance. It is essential that the examination test these attributes both in proportion to their relative importance on the job and at the level of difficulty demanded by the job.
360 F. Supp. at 1274 (footnotes omitted). See Jones v. New York City Human Resources Administration, 391 F. Supp. 1064 (S.D.N.Y. 1975); testimony of Solomon Weiner, Tr. 373, testimony of Dr. Barrett, Tr. 866.
Although generally considered less reliable than criterion-related validation, content validation will satisfy the employer's burden of showing job-relatedness. Vulcan, supra; Jones v. New York City Human Resources Administration, supra ; EEOC Guidelines, supra, 29 C.F.R. at § 1607.5(a).
The EEOC has stated that where an attempt is made to demonstrate the content validity of an examination, evidence of validity should be
accompanied by sufficient information from job analysis to demonstrate the relevance of the content (in the case of job knowledge or proficiency tests) . . .. Evidence of content validity alone may be acceptable for well-developed tests that consist of suitable samples of the essential knowledge, skills or behaviors composing the job in question. The types of knowledge, skills or behaviors contemplated here do not include those which can be acquired in a brief orientation to the job.
EEOC Guidelines, supra, 29 C.F.R. at § 1607.5(a) (emphasis added). See American Psychological Association Standards ("APA Standards"), E 12.4, Pl's, exh. 3. As defined by Judge Weinfeld, a job analysis is a "thorough survey of the relative importance of the various skills involved in the job in question and the degree of competency required in regard to each skill. It is conducted by interviewing workers, supervisors and administrators; consulting training manuals; and closely observing the actual performance of the job." Vulcan, supra, 360 F. Supp. at 1274; Kirkland v. New York State Dep't of Correctional Services, 374 F. Supp. 1361, 1373 (S.D.N.Y. 1974); Jones v. New York City Human Resources Administration, supra, 391 F. Supp. at 1077. Where an examination is prepared on the basis of a valid job analysis it should be content valid -- i.e., test in proper relative measure the key areas of knowledge and ability involved in the actual job for which the examinees are applying. Conversely, where an examination is prepared without benefit of a job analysis, or on the basis of an inaccurate job analysis, it will no doubt lack content validity. Only chance will provide otherwise, and in the absence of a job analysis the burden of persuasion in demonstrating the job-relatedness of a challenged examination becomes even greater than is normally the case. Vulcan, supra, 490 F.2d at 395-96; Jones, supra, 391 F. Supp. at 1076-77 (unless the job analysis accurately describes the content of the job, the content of the examination is likely to be seriously distorted).
Defendants have never introduced in evidence any formal job analysis, and as I stated at the June 30, 1976 hearing, their failure to do so shall be taken as evidence that no written job analysis was ever prepared. Tr. 465. Nevertheless, defendants contend that the "job-relatedness of [the examination in question] is supported . .. by the method of preparation . . . ." Def's. Mem. at 126. It appears that by this statement defendants are making two assertions: (1) in light of the test preparation procedures they followed, their failure to conduct a job analysis is irrelevant in respect of the burden of persuasion they must meet; and (2) the test preparation procedures taken were so professional as to constitute proof of job-relatedness, or at least to lessen their burden of persuasion in showing job-relatedness. See also Def's. Mem. at 43.
The evidence, however, does not support defendant's position. Solomon Weiner, former director of examinations for the Department of Personnel, testified that the actual preparation of the examinations in issue encompassed review of files of previous tests; review of materials available in the library of the Bureau of Examinations; review of the Patrol Guide rules giving the procedures of the Police Department; contact with the Police Department concerning any special conditions warranting particular attention; the drafting of a notice of examination; and finally the construction of the actual test itself, using the various material described above.
While the use of such procedures no doubt indicates a good faith effort to construct a valid test, it cannot compensate in any way for a failure to construct the test based on a proper analysis of a police officer's duties. It is particularly noteworthy that defendants have presented no evidence that the test preparations followed included any regular program of field visitations or observations. Given the absence of such a program, it is impossible for me to understand how even the most conscientious procedures imaginable could, except fortuitously, supply the crucial information otherwise available through a job analysis, and on which a content valid test must be based. In Vulcan, supra, 360 F. Supp. at 1274-76, Judge Weinfeld held a similar test preparation procedure inadequate, and the record in this case provides no basis for a different finding here.
Yet even apart from the issue of inadequate preparation, defendants have failed to present any convincing evidence that the challenged examinations are content valid. Indeed, about the only direct evidence of content validity put forth by defendants at trial was the testimony of Mr. Weiner, who defined content validity and testified that the tests in issue met that standard, but who gave no reasons in support of his conclusion. Tr. 372 et seq.
In contrast, plaintiffs' witness Dr. Richard S. Barrett, a nationally recognized testing expert, was able to point with specificity to serious deficiencies in the content of the examinations. It appeared to the court that Dr. Barrett's analysis was both better reasoned than Mr. Weiner's and based on a greater expertise in the science of testing. I am aware that Dr. Barrett sets a high standard for test validation;
nevertheless, I am convinced by this testimony, certain other evidence in the record and common sense that the challenged examinations not only have not been shown to be content valid, but are in fact lacking in content validity.
In addition to severely criticizing the method of test preparation used by defendants, Dr. Barrett stated that the actual content of the examinations had little relation to the content of the job of police officer.
For example, Dr. Barrett testified that the "judgment" questions on the examinations were inappropriate since:
there is certainly no reason to believe that a person taking this test will experience the same emotions and really relive the same experience if he were actually confronted with the situation. So what he may say he would do under these circumstances may be vastly different from what he would actually do. Tr. 871.
Yet examinations #9080, 9049 and 0013 all contained between 20-25 percent or more of such questions (involving hypothetical police situations). See Def's. exhs. N.O.P.I note that former NYCPD Commissioner Patrick Murphy also testified that he did not believe such questions tested for the same sort of judgments that a police officer would actually have to make.
Examinations #0013 and 9049 had approximately 20 percent "civic awareness" questions. Examination #9080 had approximately 8 percent of such questions. Yet Dr. Barrett testified that "certainly there is no police related behavior that stems directly from the knowledge or the lack of knowledge of that kind of question." Tr. 874. That, no doubt, is completely correct. Indeed, Dr. David Hall, defendants' witness who prepared the 1973 examination #3014, testified that civic awareness questions were not found important by his job analysis. Tr. 777-78. See Vulcan, supra, 360 F. Supp. at 1276.
Examinations #9049, 9080 and 0013 had 4 percent, 12 percent and 20 percent of their questions in the area of space and figure relations. Dr. Barrett testified that these questions had little to do with what a police officer does on his job. Tr. 876-77. Dr. Hall testified that he "would not say [these areas were] unrelated to the job, but in [Psychological Corporation's] opinion [they] did not warrant being included in the test." Tr. 781. Dr. Hall also testified that the same was true as far as "coding" questions were concerned. Ibid. Dr. Barrett again testified that such questions did not test for the content of a police officer's job. Tr. 877. Yet examination #9080 included 6 percent "coding" questions. See Def's. exh. O.
It is clear that this evidence falls short of demonstrating content validity, which is the burden defendants must meet.
Defendants contend, however, that they have shown indirectly what they have failed to demonstrate directly. They assert that the content of the pre-1973 examinations is substantially equivalent to the content of examination #3014,
and since the latter examination is content valid, the earlier examinations must be as well. Assuming arguendo the content validity of #3014, however, defendants' argument remains unpersuasive.
No doubt in part because of the significant change in test preparation procedure,
the content of the later test varies substantially from that of the earlier tests. For example, the test plan for examination #3014 prepared by the Psychological Corporation in its Phase I Report suggests the following weighting in the written test:
Test Content Number of Items
Judgment in Police Situations 40-45
Understanding of Written 35-40
Materials & Diagrams
Completion of forms
Observational Skills 15-20
Recognition of faces
Recall of Significant Details
Basic Arithmetic Skills 5-10
Counting and Dial Reading
Def's. exh. R. On the other hand, the test plan for examination #0013, for example, gives a totally different picture of the areas in which applicants should be tested:
Test Content Relative Emphasis
General Aptitude 40%
Social Sensitivity 20%
Awareness of Urban Problems 20%
Pl's. exh. 25.
Inspection of other of the challenged examinations and their test plans shows not only a similar difference in testing content between them and examination #3014, but also a substantial difference in testing content among themselves. See, e.g., Pl's. exhs. 27A-N, 23, 24. This evidence is insufficient under Title VII standards. Accordingly, I find that the defendants have not demonstrated that the challenged examinations are content valid.
ii. Criteria-Related Validity.
Defendants argue that they have submitted sufficient evidence to validate the challenged examinations according to the criteria-related (or "predictive") method of test validation.
Criteria-related validity studies correlate test performance with job performance. Tr. Barrett at 889; Weiner at 379; Hall at 761. When a test that is validated by a criteria-related method is employed, the examinees who achieve the highest (or lowest) scores should also be those who will be the better (or worse) employees. See APA Standard, supra, at 26 et seq.
Defendants point to a validity study prepared by Dr. William Weichun ("Weichun study"), a researcher at the New York City Board of Education. The purpose of this study was to determine the relationship between the NYCPD entry level examinations and academic success at the NYC Police Academy ("the Academy"). Tr. 835. Dr. Weichun found that the examinations employed between 1968 and 1970 were "very good [predictors] of success at the police academy." Tr. 839-40. Dr. Weichun also found that the tests were equally predictive for whites, blacks and Hispanics. Ibid.
Plaintiffs do not contest the findings of Dr. Weichun.
Rather, plaintiffs contend that the study does not constitute evidence of criteria-related validity. Had Dr. Weichun's study revealed a high correlation between results on the challenged examinations and job performance, then presumably the Title VII claim would fail, as predictive validity would have been proved.
The correlation shown here, however, was not between test scores and job performance, but between test scores and performance at the Academy. There was no attempt to correlate performance at the Academy and performance on the job.
Relying on the Supreme Court's recent decision in Washington v. Davis, 426 U.S. 229, 48 L. Ed. 2d 597, 96 S. Ct. 2040 (1976), defendants assert that such a showing is unnecessary, and that the Weichun study is sufficient under the law of that case to show job-relatedness. I disagree.
In Vulcan, supra, 490 F.2d at 396, the Second Circuit disapproved of a contention similar to that asserted here by defendants. The court stated there that without a showing that performance at training school correlated with performance on the job, demonstration of a high correlation between entry level test scores and training school performance meant little.
Accord, Commonwealth of Pennsylvania v. O'Neill, 348 F. Supp. 1084, 1090 (E.D. Pa. 1972), aff'd in pertinent part, [5 EPD P 8448] 473 F.2d 1029 (3rd Cir. 1973); United States v. Chicago, 385 F. Supp. 543, 555-56 (N.D. Ill. 1974); Officers for Justice v. Civil Serv. Comm'n, 371 F. Supp. 1328, 1337 (N.D. Cal. 1973); Smith v. East Cleveland, 363 F. Supp. 1131, 1148-49 (N.D. Ohio 1973), aff'd in part and rev'd in part on other grounds, 520 F.2d 492 (6th Cir. 1976); Harper v. Mayor and City Council, 359 F. Supp. 1187, 1202-03 (D. Md.), modified and aff'd, 486 F.2d 1134 (4th Cir. 1973).
The reasoning of the Second Circuit in Vulcan is applicable to this case. Defendants' study demonstrates a high correlation between performance on the challenged examinations and performance at the Academy, but the task of correlating the examinations to actual performance as police officers remains unaccomplished. The respective positions of the parties would be unchanged if, for example, instead of giving these entry level examinations, the NYCPD had enrolled the same group of individuals in the Academy. Because of the established correlation, standing on completion of training at the Academy would, presumably, approximate standing achieved at the entry level examinations. That is to say, a consistently higher proportion of whites than minorities would achieve passing grades, and a large majority of the highest achievers would be white. Accordingly, the greatest number of Academy graduates first hired would be white, and the greatest number of those first fired would be Academy graduates who are black and Hispanic. In a Title VII attack in that situation, plaintiffs' prima facie case would be established by showing, as here, the disproportionate impact of Academy procedures and testing on minorities. Defendants would be required to demonstrate that those testing practices are job-related, as they are required here to do in respect of the entry level examination. By introducing Academy performance as a correlate to the examinations, defendants cannot avoid their obligation to show job-relatedness.
To provide that missing ingredient defendants would have this court find job-relatedness on the basis of a high correlation between the results of two separate testing practices, neither of which by itself has been validated according to accepted methods. We cannot accept this flawed argument.
Defendants seek support for their contention in the Supreme Court decision in Washington v. Davis, supra. Reliance on Washington, however, is surely misplaced.
In Washington, two blacks whose applications to become police officers in the District of Columbia had been rejected, were permitted to intervene in a suit that challenged the promotion policies of the D.C. Police Department. They asserted that the Department's recruiting practices, which included the giving of a written test ("Test 21"), violated their rights under the due process clause of the Fifth Amendment, 42 U.S.C. § 1981 and D.C. Code § 1-320. The intervenors (respondents in the Supreme Court but referred to here as plaintiffs) moved for partial summary judgment on the recruitment phase of the case, seeking a declaration that Test 21 unlawfully discriminated against minorities in violation of the Fifth Amendment. The defendants (petitioners in the Supreme Court) filed a counter motion for summary judgment, asserting that plaintiffs were entitled to relief on neither constitutional nor statutory grounds.
The district court granted the defendants' motion. Davis v. Washington, 348 F. Supp. 15 (D.C. 1972). The court found, inter alia : (1) there was no claim of a racially discriminatory purpose, but only a claim that Test 21 had a racially discriminatory impact; (2) a higher proportion of blacks than whites failed Test 21; (3) Test 21 had not been validated to establish its reliability for measuring job performance. Id. at 16. These findings were held sufficient to shift the burden to defendants to show the job-relatedness of Test 21. The court found that they had met that burden, as "the undisputable facts prove the test to be reasonably and directly related to the requirements of the police recruit training program and that it is neither designed nor operates to discriminate against otherwise qualified blacks." Id. at 17. Most importantly, for the purposes of this case at least, the court held that the defendants were entitled to summary judgment even though Test 21 had not been validated as to subsequent job performance. It was sufficient to show validation as to performance at recruit school. Id. at 18.
On appeal, the court of appeals reversed. 168 U.S. App. D.C. 42, 512 F.2d 956 (1975). That court held that in resolving the constitutional claim it would be guided by the Title VII principles established in Griggs v. Duke Power Co., supra, and determined that (1) it was irrelevant whether plaintiffs had shown any evidence of a racially discriminatory purpose; and (2) absent proof by defendants that Test 21 had been validated in regard to job performance, defendants had not rebutted the plaintiffs' showing of Test 21's racially discriminatory impact. 512 F.2d at 960.
The court of appeals' decision was in turn reversed by the Supreme Court. In an opinion written by Justice White, the Court held that the court of appeals had erred in applying Title VII standards to Fifth Amendment claim of discrimination. See discussion supra.
In part III of its opinion the Court considered the statutory issues raised by the summary judgment motion. Although Justice White does not explicitly state exactly what statutory issues the Court has in mind, and this does create some confusion, it is reasonably clear from a reading of the full opinion that the statutory holdings in Washington are limited to 42 U.S.C. § 1981 and § 1-320 of the District of Columbia Code.
No other interpretation makes sense. The Court explicitly states that no Title VII issue is raised by Washington, 426 U.S. at 238, n.10. See also Justice Stevens' concurring opinion, id. at 255. As far as the statutory issues were concerned, the Court held that respondents had no better claim under the statutes involved than they had under the Fifth Amendment. Justice White wrote:
The advisability of the police recruit training course informing the recruit about his upcoming job, acquainting him with its demands and attempting to impart a modicum of skills seems conceded. It is also apparent to us, as it was to the District Judge, that some minimum verbal and communicative skill would be very useful, if not essential, to satisfactory progress in the training regimen. Based on the evidence before him, the District Judge concluded that Test 21 was directly related to the requirements of the police training program and that a positive relationship between the test and training school performance was sufficient to validate the former, wholly aside from its possible relationship to actual performance as a police officer. This conclusion of the District Judge that training performance validation may itself be sufficient is supported by regulations of the Civil Service Commission, by the opinion evidence placed before the District Judge and by the current views of the Civil Service Commissioners who were parties to the case. Nor is the conclusion foreclosed by either Griggs or Albemarle Paper Co. v. Moody, 422 U.S. 405, 95 S. Ct. 2362, 45 L. Ed. 2d 280 (1975); and it seems to us the much more sensible construction of the job-relatedness requirement.
426 U.S. at 250-51. The Court was not convinced that the trial court's finding that Test 21 was directly related to the requirements of the police training program was erroneous. Id. at 251-52.
Contrary to defendants' assertions, Washington does not preclude this court from determining that they have not met the burden imposed on them by Title VII of showing the job-relatedness of the examinations at issue.
First, all the Court held in part III of its opinion was that, under the proper construction of the non-Title VII statutes applicable to the case, the lower court was entitled to find Test 21 to be a valid test once it found the test directly related to the requirements of the police training program. The Court's statement that this conclusion is not "foreclosed" by either Griggs or Albemarle, is fully consistent with this interpretation, since these cases involved the stricter validation standards of Title VII. Both Justice Stevens in his concurrence, 426 U.S. at 255-56, and Justice Brennan in his dissent, id. at 257, support my reading of the case.
Furthermore, as Justice Stevens points out, the EEOC Guidelines on Employment Procedures, 29 C.F.R. § 1607, see discussion supra, are not applicable to either 42 U.S.C. § 1981 or D.C. Code § 1-320, the statutes to which the Court's holding in part III was limited. These very guidelines, however, which require that a criteria-related validity study show a positive relationship between the challenged test and actual job performance, were an essential ingredient in Griggs and Albemarle, supra. Thus Washington does not in any way preclude a court in a purely Title VII case such as this, where the guidelines are applicable,
from holding that job performance validation is necessary.
Second, there is a crucial difference between Test 21 in the Washington case and the examinations challenged here. Test 21 is not used in an attempt to predict whether applicants will subsequently become good police officers. Rather, Test 21 is an examination administered generally throughout the federal service that tests an applicant's reading and verbal ability. Washington, supra, 426 U.S. at 234-35. It is used to insure that potential recruits for the police department (and indeed for other branches of the federal service) have the minimal amount of reading and verbal skills necessary for training purposes. Id. at 235; Memorandum in Support of defendants Hampton, Spain and Andolsek's motion for summary judgment, 426 U.S. Briefs 229, at 97 (1976).
This case involves a totally different type of test. The challenged NYCPD examinations are not used merely to screen out individuals whose verbal and reading proficiencies fall below those necessary for training purposes. They were clearly designed and used to predict which applicants would ultimately become the best police officers, and to eliminate from consideration those applicants who, based on the test results, would become the worst officers. In other words, they were designed to predict performance on the job, not trainability. Defendants attempt to skirt this fact by proffering evidence of the need for police officers of high intelligence. Yet, even if an applicant's score is indicative of the level of his general intelligence, that is clearly only an accidental and unplanned result of the use of the examinations. In any event, the plaintiffs in this case are all individuals who have passed the challenged examinations, and who therefore possess the minimum abilities that defendants contend are necessary for good police work and for which they contend the examinations test. Unlike Washington, plaintiffs' complaint is based not on the use of the NYCPD examinations for initial screening purposes, but rather on their use for ranking purposes, with the result that a disproportionately large percentage of those hired last and fired first are minorities.
Defendants' invocation of Washington is, therefore, unavailing. Accordingly, the court holds that the defendants have failed to demonstrate that the challenged tests are job-related under the criteria-related method of validation.
The Height Requirement
Plaintiffs' challenge to the 67 inch height requirement focuses primarily on whether defendants have complied with Title VII. They make no showing that defendants employed a height requirement with an invidiously discriminatory purpose, as is required to make out a claim under the Fourteenth Amendment, § 1981 or § 1983. Washington v. Davis, supra. The use of height requirements is a well-established practice among law enforcement agencies throughout the country.
In a large sample of the nation's police departments, 97 percent had some minimum height requirements for male officers in 1973, with the average minimum requirement being 68 inches.
Indeed, in light of the traditional, accepted and widespread use of height requirements as requisites for law enforcement employment, the court would be hard-pressed to find an express discriminatory purpose in this case. Furthermore, during the examination periods under consideration, the contrary is demonstrated by the fact that the minimum height requirement for police officers was lowered from 5 feet 8 inch to 5 feet 7 inch with the hope that this would increase the number of Hispanics on the force.
Nor is the NYCPD's voluntary suspension of the height requirement [in] 1973 indicative of earlier discriminatory intent. In March, 1973, the Justice Department for the first time required police departments receiving federal funds to comply with EEOC Guidelines on employment selection procedures
and to demonstrate that minimum height requirements were an operative necessity for designated job categories. Because the defendant department concluded that it should conduct its own study
to show the correlation between height and job performance rather than rely on other studies and opinions, it eliminated the height requirement prospectively for the 1973 examination. In point of fact, the department's elimination of the height requirement in 1973 in order to conduct its own study may be more probative of good faith than of discriminatory intent.
However, with respect to the Title VII claim, plaintiffs have introduced evidence tending to show that most Hispanics could not meet the height requirement imposed by the NYCPD and that therefore, the requirement had a disproportionate impact on Hispanics.
Defendants have not met the burden of justification which under Title VII is required.
The pervasive use of height standards observed earlier bespeaks a widespread belief among law enforcers that the height of a police officer has an important effect on his job performance. These views, however, have recently come under attack and their validity has been seriously questioned. The fact that the use of height requirements are a time-honored or widely-accepted practice does not automatically save it from a claim of discrimination. As has been seen in the areas of race and sex discrimination, long-accepted stereotypes too often help perpetuate discriminatory practices even though they have no basis in fact. The federal government has demonstrated its concern that height requirements not be improperly used to exclude minority groups from employment as law enforcement personnel. EEOC Guidelines, 29 C.F.R. § 1606.1(b)
In order to defend a minimum height standard successfully, defendants have to show that it was actually job-related. This they have failed to do.
As conceded by defendants, an analysis of recent studies of height requirements conducted by the NYCPD revealed "major methodological weaknesses or insufficient data" which made the findings of these studies at best inconclusive. Def's. exh. E. See also testimony of Ingrid Balady, Tr. 212. One study referred to in defendants' post-trial memorandum concluded that height differences in two departments surveyed (Nassau County, New York and Dallas, Texas) have, with one exception, had no statistically significant effect upon performance.
Its findings were most persuasive.
Federal courts in other Title VII race and sex discrimination cases have found that height was not a valid occupational qualification. See Mieth v. Dothard, 418 F. Supp. 1169, 12 E.P.D. P 11,040 (M.D. Ala. 1976); League of United Latin Am. Citizens v. City of Santa Ana, 410 F. Supp. 873 (C.D. Cal. 1975); Officers for Justice v. Civil Serv. Comm'n, 395 F. Supp. 378 (N.D. Cal. 1975); Hail v. White, 8 E.P.D. P 9637 (N.D.Cal. 1973); but see Smith v. Troyan, 520 F.2d 492 (6th Cir. 1975).
We have been given no rational basis for validating such a requirement here. In sum, defendants have not satisfied Title VII yardsticks in respect of a 67-inch minimum height requirement; accordingly, they are held to have violated Title VII.
Plaintiffs' motions for certification of a plaintiff class and for a preliminary injunction with respect to both the entry level examinations and the height requirement are granted. The foregoing shall constitute the court's findings of fact and conclusions of law in accordance with Fed. R. Civ. P. 52(a).
It Is So Ordered.
A careful reading of the Rand Report, Pl's. exh. 1, will make clear the disproportionate racial impact of the challenged examinations. The following tables -- lifted directly from the Report -- are provided here for the convenience of those who do not have ready access to the record in this case.
ESTIMATED RACIAL COMPOSITION OF DECILES DATA FROM PD, QUESTIONNAIRE, TELEPHONE SAMPLE
White Black Hispanic Total
Num- Per- Num- Per- Num- Per- Num- Per-
Decile Score ber cent ber cent ber cent ber cent
1 94-100 574 12.8 34 3.5 19 2.3 627 10.0
2 92-94 548 12.2 44 4.6 34 4.2 626 10.0
3 89-92 514 11.5 54 5.6 59 7.2 627 10.0
4 87-89 535 11.9 67 6.9 24 2.9 626 10.0
5 84-87 484 10.8 74 7.7 69 8.4 627 10.0
6 80-84 473 10.6 101 10.4 52 6.4 626 10.0
7 76-80 426 9.5 126 13.0 75 9.2 627 10.0
8 70-76 376 8.4 131 13.5 119 14.5 626 10.0
9 61-70 335 7.5 130 13.4 162 19.8 627 10.0
10 0-61 216 4.8 206 21.3 205 25.1 627 10.0
Pass 75-100 3670 81.9 527 54.5 372 45.5 4569 72.9
Fail 0-74 811 18.1 440 45.5 446 54.5 1697 27.1
Total 4481 100.0 967 100.0 818 100.0 6266 100.0
Average score 84.9 75.3 73.0 81.8
NOTE: Counts and percents are subject to sampling errors as shown in Table 5.
95% CONFIDENCE INTERVALS FOR ESTIMATES OF EXAM SCORE DISTRIBUTION
Whites Blacks Hispanics
Decile Number Percent Number Percent Number Percent
1 552- 590 12.0-13.5 18- 56 1.7- 6.3 19-54 2.1- 7.3
2 522- 573 11.4-13.1 27- 65 2.6- 7.4 23-52 2.6- 7.0
3 480- 549 10.5-12.6 33- 78 3.1- 8.9 32-87 3.6-11.7
4 507- 564 11.1-12.9 37- 96 3.5-10.9 24-56 2.7- 7.5
5 457- 511 10.0-11.7 53- 94 5.0-10.7 50-90 5.6-12.1
6 442- 504 9.6-11.5 73- 129 6.9-14.7 41-68 4.6- 9.2
7 394- 457 8.6-10.4 98- 155 9.3-17.7 60-94 6.7-12.7
8 333- 418 7.3- 9.6 97- 166 9.2-18.9 84-153 9.4-20.6
9 290- 373 6.3- 8.5 98-162 9.3-18.5 126-199 14.1-26.8
10 174- 300 3.8- 6.8 167-244 15.8-27.8 136-274 15.2-36.9
Pass 3580-3760 78.1-86.0 447- 607 42.4-69.2 316-428 35.3-57.7
Total 4372-4586 100.0 877-1054 100.0 742-896 100.0
ESTIMATED RACIAL COMPOSITION OF DECILES DATA FROM PD, QUESTIONNAIRE, TELEPHONE SAMPLE
White Black Hispanic Total
Num- Per- Num- Per- Num- Per - Num- Per-
Decile Score ber cent ber cent ber cent ber cent
1 96-100 365 12.8 26 3.3 14 3.5 405 10.0
2 90-96 358 12.5 40 5.0 8 2.0 406 10.0
3 86-90 343 12.0 42 5.3 20 5.1 405 10.0
4 82-86 340 11.9 54 6.8 12 3.0 406 10.0
5 78-82 305 10.7 69 8.6 31 7.8 405 10.0
6 73-78 296 10.3 65 8.1 45 11.4 406 10.0
7 69-73 270 9.4 102 12.8 33 8.4 405 10.0
8 63-69 222 7.8 126 15.8 58 14.7 406 10.0
9 56-63 251 8.8 104 13.0 50 12.7 405 10.0
10 29-56 112 3.9 170 21.3 124 31.4 406 10.0
Pass 75 1932 67.5 280 35.1 119 30.1 2333 57.5
Total 2862 100.0 798 100.0 395 100.0 4055 100.0
Average score 80.0 68.9 65.9 76.4
NOTE: Counts and percents are subject to sampling errors as shown in Table 7.
95% CONFIDENCE INTERVALS FOR EXAM SCORE DISTRIBUTION
Whites Blacks Hispenics
Decile Number Percent Number Percent Number Percent
1 344- 383 11.7-13.8 14- 44 1.6- 6.1 8- 28 1.8- 8.2
2 333- 378 11.3-13.6 19- 64 2.2- 8.8 8- 35 1.8-10.2
3 321- 367 10.9-13.2 24- 63 2.8- 8.7 14- 33 3.1- 9.6
4 322- 355 10.9-12.8 38- 71 4.2- 9.8 11- 38 2.5-11.1
5 287- 327 9.8-11.7 51- 86 5.8-11.9 23- 42 5.1-12.3
6 268- 321 9.1-11.5 45- 86 5.2-11.9 25- 65 5.6-19.0
7 241- 300 8.2-10.8 75-130 8.6-18.0 20- 48 4.5-14.0
8 189- 253 6.4- 9.1 96-155 11.0-21.4 36- 81 8.1-23.7
9 220- 282 7.5-10.1 76-132 8.7-18.2 30- 70 6.7-20.5
10 84- 139 2.9- 5.0 140-200 16.1-27.6 94-153 21.0-44.7
Pass 1886-1978 64.1-71.1 237-322 27.2-44.5 88-150 19.7-43.9
Total 2783-2942 100.0 724-872 100.0 342-447 100.0
ESTIMATED EXAM SCORE DISTRIBUTIONS BY RACE DATA FROM PD AND QUESTIONNAIRES ONLY
White Black Hispanic Total
Num- Per- Num- Per- Num- Per- Num- Per-
Decile Score ber cent ber cent ber cent be r cent
1 96-100 358 12.9 17 2.3 30 5.4 405 10.0
2 90-96 359 12.9 19 2.6 28 5.1 406 10.0
3 86-90 355 12.8 25 3.5 25 4.5 405 10.0
4 82-86 334 12.0 49 6.8 23 4.1 406 10.0
5 78-82 282 10.2 97 13.3 26 4.7 405 10.0
6 73-78 295 10.6 66 9.1 45 8.1 406 10.0
7 69-73 197 7.1 107 14.8 101 18.2 405 10.0
8 63-69 196 7.1 115 15.9 95 17.1 406 10.0
9 56-63 227 8.2 81 11.2 97 17.5 405 10.0
10 29-56 171 6.2 149 20.6 86 15.5 406 10.0
Pass 75 1910 68.9 257 35.4 166 29.9 2333 57.5
Total 2774 100.0 725 100.0 556 100.0 4055 100.0
Assumption used in deriving table:
The rate of response to the questionnaire by blacks was 91 percent higher than for whites, and the response rate for Hispanics was 11 percent higher than for whites.
HISPANIC COUNT BY SURNAME CHECK
of Total in Score
Decile Score Count Hispanics Range
1 96-100 10 2.7 2.4
2 90-96 11 3.0 2.7
3 86-90 25 6.8 6.2
4 82-86 16 4.4 3.9
5 78-82 32 8.8 7.9
6 73-78 30 8.2 7.4
7 69-73 45 12.3 11.1
8 63-69 58 15.9 14.3
9 56-63 61 16.7 15.1
10 29-56 77 21.1 19.0
Pass 75 115 31.5 4.9
Total 365 100.0 9.0
ESTIMATED EXAM SCORE DISTRIBUTIONS DATA FROM POLICE DEPARTMENT AND QUESTIONNAIRES ONLY (EXCEPT SURNAME MATCH USED FOR HISPANICS)
White Black Hispanic Total
Num- Per- Num- Per- Num- Per- Num- Per-
Decile Score ber cent ber cent ber cent ber cent
1 94-100 573 12.7 29 2.9 25 3.3 627 10.0
2 92-94 546 12.1 40 4.0 40 5.3 626 10.0
3 89-92 532 11.8 55 5.5 40 5.3 627 10.0
4 87-89 552 12.2 42 4.2 32 4.2 626 10.0
5 84-87 486 10.8 80 8.0 61 8.1 627 10.0
6 80-84 473 10.5 93 9.3 60 8.0 626 10.0
7 76-80 442 9.8 104 10.4 81 10.8 627 10.0
8 70-76 386 8.6 138 13.8 102 13.5 626 10.0
9 61-70 319 7.1 179 17.9 129 17.1 627 10.0
10 0-61 202 4.5 242 24.2 183 24.3 627 10.0
Pass 75 3712 82.3 482 48.1 375 49.8 4569 72.9
Total 4511 100.0 1002 100.0 753 100.0 6266 100.0
Assumption used in deriving table:
Response rate by blacks to questionnaire was 51 percent higher than for whites.
ESTIMATED EXAM SCORE DISTRIBUTIONS DATA FROM CENSUS TRACT CHARACTERISTICS (Percent of racial group in each decile) FOUR BOROUGHS COMBINED
Excluding Hispanic Surname Including All Subjects
Decile White Black White Black Hispanic
1 13.8 +- 1.3 2.3 +- 1.5 14.4 +- 1.4 2.7 +- 1.8 1.7 +- 2.5
2 11.7 +- 1.2 3.8 +- 1.7 12.3 +- 1.3 3.5 +- 1.9 3.8 +- 2.8
3 12.1 +- 1.2 4.3 +- 1.7 12.3 +- 1.3 5.7 +- 2.2 3.2 +- 2.8
4 12.4 +- 1.2 3.4 +- 1.6 12.0 +- 1.3 2.4 +- 1.8 7.2 +- 3.1
5 11.2 +- 1.2 7.0 +- 2.0 11.6 +- 1.3 8.1 +- 2.6 5.3 +- 3.1
6 10.5 +- 1.2 7.5 +- 2.1 10.7 +- 1.2 8.7 +- 2.6 6.0 +- 3.2
7 9.5 +- 1.1 12.8 +- 2.5 9.1 +- 1.2 10.2 +- 2.9 14.4 +- 4.0
8 8.0 +- 1.0 16.1 +- 2.8 7.8 +- 1.1 14.3 +- 3.3 17.0 +- 4.3
9 6.1 +- 0.9 21.1 +- 3.0 5.9 +- 1.0 22.1 +- 3.7 15.9 +- 4.4
10 4.7 +- 0.8 21.6 +- 3.0 4.1 +- 0.9 22.2 +- 3.8 25.6 +- 4.8
Pass 83.4 +- 1.4 45.7 +- 3.8 84.5 +- 1.6 45.5 +- 4.7 46.2 +- 5.9
Fail 16.6 +- 1.4 54.3 +- 3.8 15.5 +- 1.6 54.5 +- 4.7 53.8 +- 5.9
NOTE: Numbers after plus minus give 95% confidence intervals. Portions of range lying below zero may be ignored.
ESTIMATED EXAM SCORE DISTRIBUTIONS DATA FROM CENSUS TRACT CHARACTERISTICS (Percent of racial group in each decile) FOUR BOROUGHS COMBINED
Excluding Hispanic Surname Including All Subjects
Decile White Black White Black Hispanic
1 12.5 +- 1.5 3.6 +- 1.8 12.6 +- 1.6 2.8 +- 2.1 5.6 +- 3.8
2 12.2 +- 1.5 4.2 +- 1.8 12.1 +- 1.6 3.6 +- 2.2 6.8 +- 4.1
3 12.0 +- 1.5 4.4 +- 1.9 11.3 +- 1.6 3.6 +- 2.3 10.8 +- 4.3
4 11.9 +- 1.5 7.6 +- 2.2 12.6 +- 1.6 8.2 +- 2.7 3.5 +- 3.8
5 10.0 +- 1.4 10.1 +- 2.4 10.0 +- 1.5 12.3 +- 3.0 6.5 +- 4.2
6 9.7 +- 1.3 10.5 +- 2.4 9.2 +- 1.7 11.1 +- 3.4 11.1 +- 4.8
7 10.0 +- 1.5 7.7 +- 2.6 10.5 +- 1.8 6.9 +- 3.4 9.5 +- 5.0
8 8.4 +- 1.4 14.7 +- 3.2 9.1 +- 1.7 15.8 +- 4.8 10.9 +- 5.5
9 8.2 +- 1.4 13.7 +- 3.1 8.5 +- 1.6 13.0 +- 4.4 13.7 +- 5.9
10 4.9 +- 1.1 23.5 +- 3.7 4.2 +- 1.2 23.0 +- 5.8 21.1 +- 7.1
Pass 65.9 +- 2.8 37.8 +- 4.4 65.1 +- 3.5 38.0 +- 10.2 43.3 +- 12.1
Fail 34.1 +- 2.8 62.2 +- 4.4 34.9 +- 3.5 62.0 +- 10.2 56.7 +- 12.1
NOTE: Numbers after plus minus give 95% confidence intervals. Portions of ranges lying below zero may be ignored.