Brenda DAVIS
v.
CITY OF DALLAS et al.
Cynthia Jane DURBIN
v.
CITY OF DALLAS et al.
Nos. CA 3-76-0834-G, CA 3-76-1593-G.
United States District Court, N. D. Texas, Dallas Division.
March 17, 1980.*390 *391 Kenneth Molberg, Dallas, Tex., for plaintiffs.
Lee E. Holt, City Atty., Ronald E. Deutsch, Lois C. Bacon, Asst. City Attys., by Ronald E. Deutsch, Asst. City Atty., Dallas, Tex., for defendants.
PATRICK E. HIGGINBOTHAM, District Judge.
MEMORANDUM ORDER
In its order of December 28, 1979, the court set a 30-day deadline for any motion to reconsider its finding of phase I liability as to black applicants. On January 25, 1980, the City moved for reconsideration of the court's December 18, 1979, order establishing class-wide liability, and for reopening of the case for additional testimony.
I. Reconsideration of the December 18 Order
In its motion, the City presents additional data refining some of the raw statistics previously presented by the parties as a joint exhibit. This additional evidence consists of a detailed breakdown of the reasons for rejection of all rejected applicants during the period from July 15, 1977, through January 19, 1979. For the purpose of demonstrating that the additional evidence now tendered would not have changed the result, the evidence will in this order be treated as part of the record.
The tables accompanying this order summarize the results of the applicant screening process for four relevant periods. Table 1 lists the asserted reasons for rejection, grouped roughly by subject matter. Tables 2 and 3 reflect hiring statistics for 1973-76 and 1974-76, respectively, and are taken from page 1 of Joint Exhibit 1. Table 4 covers the period from July, 1977, through November, 1978, and is based on page 2 of Joint Exhibit 1. Table 5 involves the period from July 15, 1977, through January 19, 1979, and is drawn from page 3 of Joint Exhibit 1 as supplemented by the City's proffered breakdown.
The City urges availability figures reflecting only applicants who have satisfied *392 most of the City's employment criteria, and in particular its more objective criteria. It argues that any step in the process which impacts adversely on black applicants "is job-related and essential to the operation of the police department." The City also points to the fact that many of its criteria are state-mandated, and to the large number of applicants who voluntarily withdrew from consideration.
This argument is flawed by the City's failure to validate its facially neutral criteria. It may intuitively seem that college training, a good driving record, and absence of criminal activity or drug usage are related to good police performance, but the court cannot take judicial notice that they are manifestly related to the hiring of quality law enforcement personnel. The relationship is no more obvious than that between high school education and performance as an industrial worker, as in Griggs v. Duke Power Co., 401 U.S. 424, 91 S. Ct. 849, 28 L. Ed. 2d 158 (1971), and Albemarle Paper Co. v. Moody, 422 U.S. 405, 95 S. Ct. 2362, 45 L. Ed. 2d 280 (1975). In the words of Griggs, the City has "the burden of showing that any given requirement [has] a manifest relationship to the employment in question." 401 U.S. at 432, 91 S.Ct. at 854.
The court is mindful that the City's burden of validation ordinarily arises only after plaintiff has established a prima facie case of "gross disparity" in the impact of the challenged requirements. In a case such as this, however, where the court must choose among competing sets of raw and refined applicant flow figures in deciding whether plaintiffs have met this initial burden, it is incumbent upon the City to justify the application of other than the raw applicant flow data.[1] The court cannot presume that the City's recruitment efforts have generated a minority applicant pool less well qualified than the general minority population. The latter group is ordinarily that with which a defendant's hiring is compared, without any reduction for those who do not meet the employer's minimum standards. See Dothard v. Rawlinson, 433 U.S. 321, 97 S. Ct. 2720, 53 L. Ed. 2d 786 (1977). This being the case, any refinement of the applicant availability figures to exclude those obviously disqualified must depend on validation of the disqualifying factors.
In the present case, the foregoing tables reveal that statistically significant "gross disparity" remains even after the application of many of the City's challenged and unvalidated hiring criteria. Tables 2 and 3 show Z-scores[2] of -3.25 and -4.24, *393 respectively, using as an availability figure the pool of applicants who have passed all but the partially subjective "background investigation." Otherwise stated, the background investigation itself has a grossly disparate impact on blacks who have passed all other steps. Absent validation of the qualifications found wanting in the background check, this disparate impact is sufficient to establish liability through December 31, 1976.[3]
A similar result is reached when the more detailed data of Tables 4 and 5 are considered. Assuming (without of course deciding) that criteria such as criminal activity, falsification, physical agility, drug usage, and age are valid, the applicant pool which meets these requirements remain sufficient to produce gross disparity when compared with the City's actual hires.[4] Indeed, statistically significant results are obtained even assuming that all job requirements except education are valid.[5]
The City cites Moore v. Southwestern Bell Telephone Co., 593 F.2d 607 (5th Cir. 1979), for the proposition that a small differential (e. g., 7%) between hire-to-application ratios for whites and for blacks is insufficient to constitute gross disparity. The corresponding differences in the present case range from 5.5% to 10.8%.[6] Such a comparison may be misleading, however, in that the Moore case involved success ratios of 97% and 90%. A 7% difference between 97% and 90% ought not to be treated the same as a 7% difference between, e. g., 14% and 7%, since the latter figures evince a much larger degree of disparity. Furthermore, there is no indication that the record in Moore contained expert testimony on the calculation of statistical significance levels, which constitute a considerably more reliable measure of disparity than the comparison of raw percentages urged by the City. See Davis v. City of Dallas, 483 F. Supp. 54, 57 n.3 (N.D.Tex. 1979); Quigley v. Braniff Airways, Inc., 85 F.R.D. 74, 81 n.6 (N.D.Tex.1979); Cooper v. University of Texas at Dallas, 482 F. Supp. 187 (N.D.Tex.1979). Cf. Hazelwood School District v. United States, 433 U.S. 299, 309 n.14, 97 S. Ct. 2736, 2742 n.14, 53 L. Ed. 2d 768 (1977) (difference of two to three standard deviations triggers suspicion of nondiscriminatory hypothesis); Castaneda v. Partida, 430 U.S. 482, 497 n.17, 97 S. Ct. 1272, 1281 n.17, 51 L. Ed. 2d 498 (1977) (same with respect to jury selection). As the court stated in Bilingual Bicultural Coalition on Mass Media, Inc. v. F.C.C., 595 F.2d 621, 625 n.7 (D.C.Cir.1978), "percentage statistical comparisons, while often useful, are not always dispositive or even reliable."
For these reasons, the motion to reconsider the court's order of December 18, 1979, is DENIED.
II. Reopening of the Phase I Liability Trial
The City has suggested that evidence at least partially validating its requirements may be available within a short time. This allegation confirms the fact that no validation evidence was presented at trial. As the Sixth Circuit stated in Ramsey v. United Mine Workers, 481 F.2d 742, 753 (6th Cir.), cert. denied, 414 U.S. 1067, 94 S. Ct. 576, 38 L. Ed. 2d 473 (1973): *394 Generally, of course, parties to litigation must present their evidence when the case is called for trial. It is normal to wish to present additional evidence after once having lost a dispute of fact. But, of course, such a practice would lead to never-ending litigation.
Use of newly developed validation evidence in a reopened phase I trial would require substantial additional time for the preparation and presentation of revised statistical evidence related to the disparities remaining after applicant flow data have been refined to allow for validated exclusions, in addition to the delay occasioned by the validation process itself. A reopening of the case as to race discrimination would also raise questions concerning whether, in fairness to plaintiffs, the sex discrimination case ought also to be reopened. In light of these considerations, and in view of the fact that these cases have been pending for over three years with their end now only barely in sight, the court is of the opinion that defendant's motion, insofar as it seeks to reopen the case for the purpose of defeating the phase I presumption of discrimination established by the court's earlier findings, should be DENIED.
III. Use of Validation Evidence as a Defense to Injunctive Relief
Part of the relief sought in the Davis complaint includes injunctive relief against the use of racially discriminatory practices. In addition to an injunction prohibiting racially motivated disparate treatment, such relief may potentially include an injunction prohibiting the use of various unvalidated facially neutral criteria having a racially discriminatory impact. Thus the question of validation recurs, in the form of an inquiry as to whether the City may now validate previously unvalidated employment criteria in order to exclude such criteria from the reach of possible injunctive relief.
A balancing of the interests of the parties to this suit, together with the goal of judicial economy, has led the court to conclude that presentation of validation evidence to combat plaintiffs' prima facie case would be both unwise and inequitable, yet an additional set of interests those of the people of the City of Dallas must be weighed in formulating appropriate injunctive relief. The court noted in its December 28, 1979, order the public interest in a competent and effective police force. To enjoin use of validated hiring criteria because they were unvalidated at an earlier stage of the proceeding, would disserve that interest. The Supreme Court has recognized the public interest as an important factor to be considered in formulating Title VII injunctive relief:
In devising and implementing remedies under Title VII, no less than in formulating any equitable decree, a court must draw on the "qualities of mercy and practicality [that] have made equity the instrument for nice adjustment and reconciliation between the public interest and private needs as well as between competing private claims."
International Brotherhood of Teamsters v. United States, 431 U.S. 324, 375, 97 S. Ct. 1843, 52 L. Ed. 2d 396 (1977), quoting from Hecht Co. v. Bowles, 321 U.S. 321, 329-30, 64 S. Ct. 587, 591-92, 88 L. Ed. 754 (1944). Consideration of the public interest in the injunctive relief, if any, to be formulated in this case mandates that validation evidence, if now available, be admitted in defense of claimed equitable relief. Cf. James v. Stockham Valves & Fittings Co., 559 F.2d 310, 355 (5th Cir. 1977), cert. denied, 434 U.S. 1034, 98 S. Ct. 767, 54 L. Ed. 2d 781 (1978) (practices with disparate impact enjoined "until validated"); Johnson v. Goodyear Tire & Rubber Co., 491 F.2d 1364, 1372 (5th Cir. 1974) (same).
IV. Future Course of the Litigation
The court in its December 28, 1979, order invited the parties to address the question of whether validation evidence might be used in individual back pay claims to rebut the prima facie presumption of discrimination created by the liability finding. Briefing on this question should proceed expeditiously. The parties in their briefs should consider whether, if validation *395 is to be considered in deciding entitlement to back pay claims, validation evidence presented at the injunctive relief stage should be applied to back pay claims on a group-by-group basis so as to expedite the processing of what will surely be a large number of such claims. See section II of the court's December 28, 1979, order. The preparation of a notice and claim form should also be undertaken as rapidly as time permits.
Within 30 days, plaintiffs should submit to the court a proposed injunction, supported by an accompanying brief. The City may respond within 20 days. A hearing will then be scheduled for any additional evidence. The court will be available for an informal conference to discuss any changes in schedule.
In preparing for the hearing on injunctive relief, it must not be forgotten that although the award of such relief is discretionary, the court's discretion must be exercised in light of the state of the record consistent with the purposes of Title VII. Cf. Albemarle Paper Co. v. Moody, 422 U.S. 405, 415-16, 95 S. Ct. 2362, 2370, 45 L. Ed. 2d 280 (1975) (backpay). Plaintiffs are entitled to an injunction against the use of those criteria which have a racially disparate impact and are not shown by the City, through competent evidence, to have a significant and demonstrable relationship to performance as a police officer. That this task may be a difficult one is no doubt true; that it is essential under the law is equally plain.
APPENDIX
"Withdrew" (incl. "Lack of Interest," "Failed to Return")
"No Action Noted" (incl. "Certification Expired," "Not Eligible for Rehire")
"Criminal Activity" (incl. "Thefts," "Employment Thefts," "Felony Convictions," "Buy/Sale Stolen Property," "Arrest Record," "Military Record")
"Police Officer Misconduct"
"Falsification"
"Physical Exam" (incl. "Pre-Physical")
"Inability to Understand English"
"Driving Record" (incl. "DWI")
"Physical Agility"
"Drugs" (incl. "Marijuana")
"Age"
"Education"
"Psychological" (incl. "Emotional Instability")
"Citizenship"
"Deviant Sex"
"Poor Employment History" (incl. "Poor Recommendations," "Unstable Employment")
"Poor Credit" (incl. "Delinquent Child Support," "Civil Suits")
Other ("Immaturity," "Could Not Enforce Law," "Poor Performance Under Pressure")
TABLE 1
GROUNDS FOR REJECTION
WHITE BLACK OTHER
No. % No. % No. % Total Z
Total Applications 2864 72.6 896 22.7 184 4.7 3944 - 4.24
Less:
Withdrew 1844 72.9 551 21.8 134 5.3 2529 - 3.77
No Action Noted 1799 73.1 532 21.6 131 5.3 2462 - 3.67
Physical Exam 1355 72.0 417 22.1 111 5.9 1883 - 3.93
Physical Agility 1333 71.8 415 22.3 109 5.9 1857 - 4.03
Preliminary Int. 974 72.6 279 20.8 90 6.7 1343 - 3.25
Background Inv. 437 76.0 88 15.3 50 8.7 575 0.00
Hires/Applications 15.3% 9.8% 27.2% 14.6%
TABLE 2
1973-1976 HIRING
*396
WHITE BLACK OTHER
No. % No. % No. % Total Z
Total Applications 2248 73.6 663 21.7 142 4.7 3053 - 5.75
Less:
Withdrew 1402 74.1 385 20.4 104 5.5 1891 - 5.24
No Action Noted 1366 74.3 371 20.2 101 5.5 1838 - 5.16
Physical Exam 1063 73.6 296 20.5 86 6.0 1445 - 5.28
Physical Agility 1041 73.4 294 20.1 84 5.9 1419 - 5.12
Preliminary Int. 706 74.9 170 18.0 66 7.0 942 - 4.24
Background Inv. 325 80.4 40 9.9 39 9.7 404 0.00
Hires/Applications 14.5% 6.0% 27.5% 13.2%
TABLE 3
1974-1976 HIRING
WHITE BLACK OTHER
No. % No. % No. % Total Z
Total Applications 1185 70.3 388 23.0 112 6.6 1685 - 4.37
Less:
Withdrew 827 72.0 245 21.3 77 6.7 1149 - 3.89
Criminal Activity 777 71.9 233 21.6 71 6.6 1081 - 3.98
Falsification 734 71.2 227 22.0 70 6.8 1031 - 4.09
Physical Exam 521 71.0 163 22.2 50 6.8 734 - 4.15
Driving Record 505 71.7 150 21.3 49 7.0 704 - 3.89
Physical Agility 489 71.8 147 21.6 45 6.6 681 - 3.98
Drugs 362 73.3 101 20.4 31 6.3 494 - 3.64
Age 341 73.2 95 20.4 30 6.4 466 - 3.64
Education 243 79.9 39 12.8 22 7.2 304 - 1.12
Psychological 243 80.5 37 12.3 22 7.3 302 - 0.92
Citizenship 243 81.3 37 12.4 19 6.4 299 - 0.96
Poor Employment History 224 80.3 37 13.3 18 6.5 279 - 1.31
Poor Credit 218 81.6 31 11.6 18 6.7 267 - 0.63
Oral Interview 171 83.0 21 10.2 14 6.8 206 0.00
Hires/Applications 14.4% 5.4% 12.5% 12.2%
TABLE 4
JULY 1977-NOVEMBER 1978 HIRING
WHITE BLACK OTHER
No. % No. % No. % Total Z
Total Applications 1016 71.2 312 21.9 99 6.9 1427 - 4.32
Less:
Withdrew 878 72.1 257 21.1 82 6.7 1217 - 4.09
Not Eligible Rehire 877 72.2 256 21.1 82 6.7 1215 - 4.09
Criminal Activity 809 72.4 235 21.0 74 6.6 1118 - 4.06
Police Officer Misconduct 794 72.1 235 21.3 72 6.5 1101 - 4.15
*397
Falsification 793 72.1 235 21.4 72 6.5 1100 - 4.17
Physical Exam 580 72.5 166 20.8 54 6.8 800 - 4.00
Inability to Understand
English 580 72.6 166 20.8 53 6.6 799 - 4.00
Driving Record 557 72.2 162 21.0 52 6.7 711 - 4.06
Physical Fitness 540 72.2 157 21.0 51 6.8 748 - 4.06
Drugs 368 73.0 100 19.8 36 7.1 504 - 3.70
Age 348 72.7 95 19.8 36 7.5 479 - 3.70
Education 245 79.8 38 12.4 24 7.8 307 - 1.13
Psychological 243 79.7 38 12.5 24 7.9 305 - 1.17
Citizenship 243 80.5 38 12.6 21 7.0 302 - 1.21
Deviant Sex 240 80.5 38 12.8 20 6.7 298 - 1.29
Poor Employment History 193 79.4 32 13.2 18 7.4 243 - 1.45
Poor Credit 187 82.4 22 9.7 18 7.9 227 + 0.13
Other 182 82.0 22 9.9 18 8.1 222 0.00
Hires/Applications 17.9% 7.1% 18.2% 15.6%
TABLE 5
JULY 15, 1977-JANUARY 19, 1979 HIRING
NOTES
[1] The term "raw" as applied to applicant flow figures is in one sense a misnomer. The law clearly envisions that availability estimates be refined geographically to encompass only the available minority population within the geographical area relevant to the employer's hiring. See Hazelwood School District v. United States, 433 U.S. 299, 310-12, 97 S. Ct. 2736, 2743, 53 L. Ed. 2d 768 (1977). Some refinement as to skill level may also be appropriate in cases where the jobs in question require a readily quantifiable skill whose validity is not open to serious question. Just as it makes no sense to judge a Maine employer by the minority population of Mississippi, it may make no sense to judge discrimination in the hiring of textbook writers against an availability pool including the illiterate. Such easily definable and virtually unchallengable skill requirements are not directly involved in the instant case.
[2] An excellent explanation of the use of Z-scores is given in Rich v. Martin Marietta Corp., 467 F. Supp. 587, 601 (D.Colo.1979):
The [statistical test] begins with the assumption that there is no expected difference in the anticipated performance among the population groups. The hypothesis is then tested by calculating the standard error and the Z statistic. A conclusion about the initial hypothesis may be drawn from the Z statistic. A Z statistic of less than 1.96 (roughly two standard deviations) is within the normal distribution and tends to confirm the initial hypothesis. In the instant context, such a value would indicate that the affected group is receiving an expected promotion rate. A Z statistic equal to or greater than 1.96 occurs less than 5% of the time in normal distribution and means that the initial hypothesis may be rejected. In the instant context, such a value would indicate that the proportions do not come from the same distribution. Such a statistically significant difference between the groups would be evidence of adverse impact.
See Caulfield v. Board of Education, 486 F. Supp. 862, 21 Empl.Prac.Dec. ¶ 30,389 (E.D. N.Y.1979), at 13,195. The Z-scores in Tables 2-5 are based on the binomial distribution.
[3] Hiring data is unavailable for the period from January through June, 1977.
[4] Z-scores are -3.64 and -3.70, respectively. While the sequence in which the criteria are applied by the City is slightly different from the sequence of the tables, these figures serve as rough measures of the disparate impact of the criteria which follow them in the tables. That is, although some applicants were rejected for lack of sufficient education early in the selection process and without measuring their fitness in other respects, there is no evidence that applicants rejected by reason of education were more or less likely than those which met the education requirement to be disqualified by other factors as well.
[5] Z-scores are -3.78 and -3.79 for Tables 4 and 5, respectively.
[6] Ratios for 1973-76 were 15.3% for whites and 9.8% for blacks; 1974-76 figures are 14.5% and 6.0%, respectively; July 1977-November 1978 figures are 14.4% and 5.4%; and July 1977-January 1979 figures are 17.9% and 7.1%.