GENERAL TELEPHONE COMPANY OF THE NORTHWEST, INC., PETITIONER V. EQUAL EMPLOYMENT OPPORTUNITY COMMISSION No. 89-1900 In The Supreme Court Of The United States October Term, 1990 On Petition For A Writ Of Certiorari To The United States Court Of Appeals For The Ninth Circuit Brief For The Equal Employment Opportunity Commission In Opposition TABLE OF CONTENTS Questions presented Opinions below Jurisdiction Statement Argument Conclusion OPINIONS BELOW The opinion of the court of appeals, Pet. App. 1a-17a, is reported at 885 F.2d 575. The opinion of the district court, Pet. App. 27a-39a, is reported at 40 Fair Empl. Prac. Cas. (BNA) 1533. JURISDICTION The judgment of the court of appeals was entered September 12, 1989. The order of the court of appeals denying the petition for rehearing en banc was entered March 6, 1990. Pet. App. 45a. The petition for certiorari was filed June 4, 1990. The jurisdiction of this Court is invoked under 28 U.S.C. 1254(1). QUESTIONS PRESENTED 1. Whether petitioner defeated the inference of discrimination raised by a regression analysis accounting for the major factors in the employment decision merely by pointing to a variable that it argued was not sufficiently accounted for in that analysis. 2. Whether petitioner should have been allowed to introduce self-congratulatory evidence of its equal employment opportunity efforts without disclosing self-critical portions of that material. STATEMENT Following a bench trial in the United States District Court for the Western District of Washington, petitioner was found not to have violated Title VII of the Civil Rights Act of 1964, 42 U.S.C. 2000e et seq. The court of appeals unanimously reversed and remanded for a new trial. 1. Petitioner provides telephone and telecommunications services in Washington, Oregon, Idaho, California, and Montana. In 1977, the Equal Employment Opportunity Commission (EEOC) brought this pattern and practice sex discrimination suit based on numerous individual charges filed against petitioner. Among other practices, EEOC charged that petitioner systematically excluded women from higher-paying hourly jobs (including craft, technician, and operative positions) and from higher-paying salaried jobs. Pet. App. 2a, 19a, 25a, 29a-30a. /1/ 2. In 1985, after years of extensive discovery, the case went to a four-week trial. Pet. App. 19a, 27a. EEOC presented statistical evidence, including regression analyses, which showed that women received significantly lower starting wages than men who had the same job preferences, experience, education, and personal characteristics (aside from gender). Id. at 14a, 19a. EEOC also presented testimony of witnesses who described company practices that deterred women from requesting, obtaining, and remaining in higher-paid hourly and management jobs. Id. at 2a, 19a, 24a, 27a, 33a. In rebuttal, petitioner presented a regression analysis of its own which showed a correlation between job interest and job placement. Pet. App. 15a, 34a. Petitioner also presented, over EEOC's objection, extensive evidence of its equal opportunity efforts -- evidence that included the testimony of management employees regarding the company's affirmative action programs and two hundred pages of exhibits (including policy statements, in-house newspaper articles, and letters) lauding the company's commitment to equal opportunity. Id. at 4a. Petitioner also compared its female representation in broad occupational categories to female representation in the labor force using regional and national census data. Id. at 37a-38a. 3. The district court held that petitioner had not engaged in a pattern and practice of excluding women from higher-paid hourly and salaried jobs. Applying a disparate treatment analysis, the district court concluded that petitioner rebutted any inference of discrimination created by EEOC's statistical and anecdotal evidence. Pet. App. 39a. The district court rested its conclusion on findings (1) that EEOC's statistical studies failed to analyze "in a meaningful way the extent to which career interests differed between males and females, and the effect of these differences upon the placement of employees," id. at 34a; (2) that petitioner had an active commitment to equal employment opportunity, id. at 28a-33a; and (3) that the distribution of women in petitioner's work force compared favorably with that found in public utilities and in the regional and national labor force as a whole according to census data, id. at 37a-38a. Although the district court found EEOC's evidence of individual instances of discrimination to be "uncontroverted," it found those examples insufficient by themselves to establish class-based discrimination. Id. at 38a. 4. A unanimous panel of the Ninth Circuit reversed. Pet. App. 17a. It held that the district court abused its discretion in admitting petitioner's "voluminous" evidence of its "equal employment opportunity" efforts where it had earlier precluded the EEOC from discovering self-critical portions of that same material. Id. at 4a-5a. The court of appeals explained that even those courts that have held self-critical statements to be privileged -- on the theory that a qualified privilege would promote ameliorative efforts -- have found the privilege to be waived when the employer voluntarily introduces evidence of its equal opportunity efforts in order to prove nondiscrimination. Id. at 5a. Petitioner's reliance on such evidence in this case, while excluding the self-critical portions of it, perpetrated a "one-sided presentation of a defense" and left the EEOC "ill-equipped to effectively cross-examine those of (petitioner's) witnesses who testified concerning the implementation and efficacy of (petitioner's) equal opportunity efforts." Id. at 5a-6a. Since "(f)ifteen of the (district) court's fifty-three findings related to (petitioner's) equal opportunity programs and policies," and since that "court discussed this evidence in a vastly more detailed manner than any other genre of (petitioner's) evidence," the court of appeals found it "clear * * * that the EEOC was prejudiced by this error." Id. at 6a. The court of appeals held that the district court also erred when it held that the EEOC's regression analyses failed to establish discrimination. The "critical flaw" in those analyses, according to the district court, was that they "fail(ed) to adequately analyze differences in job interest between men and women," even though they "showed a disparity in hiring that correlates with sex, while controlling for the major legitimate factors." Pet. App. 14a. The court of appeals explained that the district court's reasoning was contrary to this Court's teaching in Bazemore v. Friday, 478 U.S. 385, 400 (1986), that "a regression analysis that includes less than 'all measurable variables' may serve to prove a plaintiff's case." Pet. App. 7a (quoting Bazemore, 478 U.S. at 400 (Brennan, J., joined by all other Members of the Court, concurring in part)). To rebut EEOC's regressions, petitioner had to produce "credible evidence" that "curing the alleged flaws would also cure the statistical disparity -- proof which (petitioner) did not offer." Pet. App. 14a. It was not enough "simply (to) point() out possible flaws in the EEOC's data." Ibid. The erroneous rejection of the EEOC's regression analyses was not harmless error, said the court of appeals, because petitioner's other evidence was insufficient to overcome the inference of discrimination raised by the EEOC's evidence. Petitioner's statistical data were recognized by the district court to be of "limited probative value" because the "most relevant data were based on skewed labor pools." Pet. App. 15a & n.11. Petitioner's comparisons of its work force with external labor markets did "little" to help one "understand() whether women were precluded from higher paying jobs." Id. at 16a. Since petitioner's only remaining evidence was the improperly admitted equal employment opportunity evidence, the court of appeals could not find that petitioner overcame the inference of discrimination raised by the EEOC's evidence. Id. at 16a. Because the court of appeals could not determine how the district court would have ruled had it excluded the "affirmative action" evidence and properly assessed the statistical evidence, it reversed and remanded for a new trial. Id. at 17a. /2/ ARGUMENT The court of appeals' decision simply paves the way for a fair consideration of the evidence in this 12-year old suit. Interlocutory review is not warranted. 1. The Ninth Circuit properly looked to this Court's decision in Bazemore for the correct standard by which to assess statistical evidence in a disparate treatment case. Like this case, Bazemore was a pattern and practice discrimination suit brought under Title VII of the Civil Rights Act of 1964, 42 U.S.C. 2000e et seq. 478 U.S. at 386 (per curiam). As here, the plaintiffs in Bazemore presented multiple regression analyses that "account(ed) for the major factors" in the employment decision. 478 U.S. at 400 (Brennan, J., joined by all other Members of the Court, concurring in part); see id. at 398 (regressions accounted for race, education, tenure, and job title). But like the district court here, the lower courts in Bazemore found plaintiffs' regression analyses "'unacceptable as evidence of discrimination,' because they did not include 'all measurable variables thought to have an effect on salary level.'" 478 U.S. at 400. This Court held that view of the evidentiary value of the regression analyses to be "plainly incorrect": "a regression analysis that includes less than 'all measurable variables,' may serve to prove a plaintiff's case." Ibid. Although there may be "some regressions so incomplete as to be inadmissible as irrelevant," id. at 400 n.10, and others flawed because they fail to "account() for the major factors" in the employment decision, id. at 400, it is not sufficient for an employer "to declare simply that many factors go into" the employment decision. Instead, an employer confronted with a regression analysis that establishes discrimination after taking into account the major factors in the employment decision must make some attempt -- "statistical or otherwise -- to demonstrate that when these factors were properly organized and accounted for there was no significant disparity" remaining. Id. at 403-404 n. 14. a. In light of the legal standard established by this Court in Bazemore and followed by the Ninth Circuit in this case, petitioner sets up a straw man when it reads the court of appeals to say that "the defendant cannot rebut an inference of discrimination by * * * pointing to flaws in the plaintiff's statistics." Pet. 9-10 (ellipses in original) (quoting Pet. App. 12a). /3/ As petitioner quickly notes in the margin, the court of appeals did not say that. It said -- in the context of regression analyses which "controll(ed) for the major legitimate factors" in the employment decision, id. at 14a -- that petitioner could not rebut an inference of discrimination "merely" by pointing to flaws in the EEOC's statistics. Pet. 10 n.8. That single adjective makes all the difference. Of course petitioner could discredit a regression analysis that is "so incomplete as to be * * * irrelevant" by pointing out that fact. Bazemore, 478 U.S. at 400 n.10. Petitioner could also undermine a regression analysis that failed to "account() for the major factors" in the employment decision by pointing out that flaw, id. at 400; thus, the court of appeals noted that in Penk v. Oregon State Bd. of Higher Educ., 816 F.2d 458 (9th Cir.), cert. denied, 484 U.S. 853 (1987), a case where plaintiff's regressions omitted the most critical factors in the promotion and tenure decisions at issue there (teaching quality, service, research, and scholarship), the defendant "could defeat any inference of discrimination" "by merely pointing out such omissions." Pet. App. 13a. What petitioner could not do, and what it attempted to do in this case, was fault EEOC's otherwise rigorous regression analyses for taking insufficient account of one variable without "demonstrat(ing) that when th(is) factor() (is) properly organized and accounted for there was no significant disparity" remaining. Bazemore, 478 U.S. at 403-404 n.14; see Pet. App. 14a (petitioner "had to produce credible evidence that curing the alleged flaws would also cure the statistical disparity."). Hence, although the district court found fault with EEOC's "failure to analyze in a meaningful way" the effect of job interest on job placement, id. at 14a n.10, 34a (emphasis added), /4/ it did not find that the job interest factor was a major factor in the employment decision that had been omitted altogether. It follows that the court of appeals properly required petitioner to bear the burden of producing "credible evidence" in response to EEOC's regression analyses to defeat the "strong inference of discrimination" they raised. Pet. App. 14a. Since the district court failed to require such a showing, the Ninth Circuit was compelled to remand for a proper assessment of the evidence. Contrary to petitioner's intimation, Pet. 9-11, the court of appeals did not shift the burden of proof or impose on a defendant the requirement of presenting countervailing regression analyses. Following Bazemore, which recognizes that the burden of proof remains on the plaintiff, 478 U.S. at 400, the court of appeals merely placed upon petitioner the burden of producing "credible evidence" that legitimate factors explain the disparity. Nowhere did the court require that this evidence take the form of a regression analysis. Indeed, such a requirement would be inconsistent with Bazemore's observation that the defendant in that case failed to make an attempt, "statistical or otherwise," to defeat plaintiff's showing of a disparity. 478 U.S. at 403 n.14 (emphasis added). b. By its decision in this case, the Ninth Circuit has joined the Second, Eighth, and District of Columbia Circuits in holding that Bazemore requires the defendant to "do more than simply point out possible flaws in the proponent's statistical analyses in order to rebut the inference of discrimination raised by the statistical evidence." Pet. App. 8a. See Sobel v. Yeshiva University, 839 F.2d 18, 34 (2d Cir. 1988) ("We read Bazemore to require a defendant challenging the validity of a multiple regression analysis to make a showing that the factors it contends ought to have been included would weaken the showing of a salary disparity made by the analysis."), cert. denied, 109 S. Ct. 3154 (1989); Catlett v. Missouri Highway & Transp. Comm'n, 828 F.2d 1260, 1266 (8th Cir. 1987) (defendant claiming flaws in plaintiff's definition of relevant work force "bore the burden of introducing evidence to show this failure was significant"), cert. denied, 485 U.S. 1021 (1988); Palmer v. Shultz, 815 F.2d 84, 101 (D.C. Cir. 1987) ("Implicit in the Bazemore holding is the principle that a mere conjecture or assertion on the defendant's part that some missing factor would explain the existing disparities between men and women generally cannot defeat the inference of discrimination created by plaintiffs' statistics."). The Seventh Circuit's decision in EEOC v. Sears, Roebuck & Co., 839 F.2d 302 (1988), is no exception to the consensus interpretation of Bazemore, contrary to the view of petitioner, Pet. 12-14, and the court of appeals, Pet. App. 10a-12a. The Seventh Circuit in Sears did not hold that mere identification of alleged methodological flaws was enough to carry defendant's rebuttal burden. It addressed the entirely separate issue whether a defendant had to rebut a plaintiff's regression analysis with "a more probative statistical analysis" of its own. 839 F.2d at 313-314 (emphasis in original). The court of appeals answered in the negative, and found persuasive the defendant's substantial non-statistical evidence, including expert testimony, to show women's lack of interest in the jobs at issue in that suit. 839 F.2d at 312-313. This reading of Sears accords with other Seventh Circuit precedent. See Mister v. Illinois Cent. Gulf R.R., 832 F.2d 1427, 1431 (7th Cir. 1987) (courts will not discount plaintiff's statistics, despite their defects, where employer did not introduce facts sufficient to explain how omitted variable would account for hiring disparity), cert. denied, 485 U.S. 1035 (1988). 2. The court of appeals correctly held that the admission of petitioner's equal employment opportunity evidence was an abuse of discretion and prejudiced the EEOC. Moreover, the court's ruling that, when an employer voluntarily uses this kind of evidence to prove nondiscrimination, "it 'open(s) the door' and waives whatever qualified privilege may have existed" with respect to such evidence, is consistent with the only other circuit law on this question. Pet. App. 5a (quoting Coates v. Johnson & Johnson, 756 F.2d 524, 552 (7th Cir. 1985) (employer cannot offer its affirmative action policy as evidence of nondiscrimination and at the same time withhold self-critical evaluations that may undercut its favorable portrayal of its efforts)). Here, petitioner's one-sided presentation prejudiced the EEOC. While the Commission did what it could to dispute petitioner's self-congratulatory evidence regarding its equal opportunity efforts, it was denied access to basic data needed for effective cross-examination. For example, while petitioner's Equal Opportunity Director was permitted to describe her department's monitoring of employment decisions, Pet. App. 28a-33a, the EEOC was barred from discovering the availability figures and utilization standards used by petitioner in the monitoring process. Such evidence bears directly on the issue of intent and thus goes to the heart of this disparate treatment case. It is also clear that the equal employment opportunity evidence played a highly significant role in the district court's assessment of the case. More than one-quarter of the court's findings were devoted to this evidence, Pet. App. 6a, 28a-33a, and, as the court of appeals noted, this evidence was discussed by the district court "in a vastly more detailed manner" than any of the other evidence. Id. at 6a. See also id. at 24a (district court places "special emphasis" on affirmative action evidence). The emphasis placed on the equal employment opportunity evidence directly affected the outcome of the case. As the court of appeals observed, since the district court found petitioner's statistical data to be of limited probative value, it looked to the equal employment opportunity evidence to reinforce its conclusions. Id. at 15a. CONCLUSION The petition for a writ of certiori should be denied. Respectfully submitted. KENNETH W. STARR Solicitor General DONALD R. LIVINGSTON Acting General Counsel Equal Employment Opportunity Commission AUGUST 1990 /1/ Petitioner moved to dismiss the class aspects of the suit on the ground that EEOC had not been certified as a class representative in accordance with Federal Rule of Civil Procedure 23. The district court denied petitioner's motion, and its order was affirmed by the Ninth Circuit, 599 F.2d 322 (1979), which in turn was affirmed by this Court, General Telephone Co. of the Northwest, Inc. v. EEOC, 446 U.S. 318 (1980). /2/ The court of appeals also ruled that the record was inadequate to assess the timeliness of the underlying promotion charge and remanded for determination of this issue as well. Pet. App. 17a n.13. /3/ Accord, Pet. 11 ("The Court of Appeals holds that pointing up flaws in a plaintiff's statistics is insufficient as a matter of law to defeat a claim of discrimination based on those statistics."). /4/ In fact, the regression analyses presented by the EEOC at trial controlled for job interest, albeit not to the satisfaction of the district court. Pet. App. 22a.