# Opposition Brief — Sheehan v. Purolator, Inc. > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/brief%3Amicro_IA40385019_3008%3A2 ## Record - **Collection:** Supreme Court brief - **Document type:** Opposition Brief - **Published:** January 1, 1988 - **Citation:** 488 U.S. 891 ## Text IN THE Supreme Court of the United States OCTOBER TERM, 1988 > PATRICIA SHEEHAN, ELIZABETH HENOCH, and KAYHAN HELLRIEGEL, on Behalf of Themselves and All Others Similarly Situated, Petitioners, —VvV.— PUROLATOR, INC. and PUROLATOR COURIER CORP., Respondents. ON PETITION FOR A WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT BRIEF IN OPPOSITION COLLEEN MCMAHON 1285 Avenue of the Americas New York, New York 10019 Counsel of Record for Respondents Of Counsel: Morris B. Abram Jay Cohen Diana Hassel PAUL, WEISS, RIFKIND, WHARTON & GARRISON wth PNRM TE SN LTB D7 - : oe aaesaeh s 7 » eal baat 14 — a Re ae ee eT Ee sani : » 2 ia - 3 7 ? - os oe or aes : | ) : e } re / | : j | : \ | | | | | } : j E . 5 : | © , ¢ , : ‘ : ae ; Y 3 : De be i FEE 3 . . : | 3 : : | me : 7 ‘ J ( ' 3 . 2 4 | 3 6 . i ‘ J Ra é | , 3 | : ~ ; | S | > | Y ba / 4 ; : : . | * i QUESTIONS PRESENTED 1. Should this Court review the Second Circuit’s affirmance of the discretionary decision not to certify a class in this action, where that decision was based on a failure of proof— specifically, petitioners’ failure to satisfy the court that their application for class certification complied with the require- ments of Fed. R. Civ. P. 23(a), as required by General Tele- phone Co. of the Southwest v. Falcon, 457 U.S. 147 (1982)? 2. Should this Court review the denial of class certification in light of Watson v. Fort Worth Bank & Trust, LS. ‘ 108 S. Ct. 2777 (1988), even though (a) Watson did not change the standard for obtaining class certification articulated in Fal- con, and (b) the sole issue addressed in Watson—namely, whether claims of subjective discrimination can be analyzed under the so-called ‘‘disparate impact’? model—was raised by petitioner in the District Court but abandoned on appeal? il List of Parties and Rule 28.1 List The parties are listed in the petition. Respondent Purolator Courier Corp. (‘‘Courier’’) is a wholly owned subsidiary of Emery Air Freight Corp. ill TABLE OF CONTENTS EES eee Lest Of Parties and Rule 28.1 List. .........cccccscces cc. esas dgesdeesbhenenenneces es ko ccd owes en ec.evdecebanec ns chad daadsdcesssesescbaeces Reasons for Denying the Writ.............cccccccees I. This Case Presents No Question of General Importance Warranting Review by this Court.. A. Petitioners Are Seeking Review of Affirmed Findings of Fact, Which Is Not The Prov- EE i. eo Son cdeabeneaecesee B. Petitioners’ Attucks on the Decisions Below Under Eisen and Bazemore Do Not Raise Issues Worthy of Certiorari.............. Il. Watson v. Fort Worth Bank & Trust Does Not Require that this Court Review the Denial of et ede daweeeeeeene A. Falcon, Not Watson, Controls The Propri- ety of Class Certification ................ B. Petitioners Cannot Raise The Disparate Impact Issue Here Because They Abandoned tis cote sagt babies sue aues és ee ee aaneaee i ss eee widened ethecéckeesecnsobe ee PAGE il 12 12 15 17 la iv TABLE OF AUTHORITIES Cases: PAGE Adickes v. S.H. Kress & Co., 398 U.S. 144 (1970)..... 1S Bazemore v. Friday, 478 U.S. 385 (1986)........... 9, 10, 11 Berenji v. District Director, Immigration and Naturaliza- OR BOVIS, SES U.S. GRO CHOY cn cccccccccscscces i) Branti vy. Finkel, 445 U.S. 507 (1980). ................ y Coopers & Lybrand v. Livesay, 437 U.S. 463 (1978) ... i] Delta Air Lines, Inc. v. August, 450 U.S. 346 (1981)... 15 Eisen v. Carlisle & Jacquelin, 417 U.S. 156 (1974) ..... 9, 10 General Telephone Co. of the Southwest v. Falcon, 457 Se ee ES Cc tokyo ncuu areas vb bases eens passim Graver Tank & Mfg. Co. vy. Linde Air Products Co., 336 Ser ee es Sa wsn asd as VaR E 6s 64k sakenee 2 Helvering y. Clifford, 309 U.S. 331 (1940)............ 16 Helvering v. Wood, 309 U.S. 344 (1940). ............. 16 International Brotherhood of Teamsters v. United ee. Se Se Te 6c edn deb ceeeesesetaencs 4, 10 Layne & Bowler y. Western Well Works, 261 U.S. 387 DN cciaceeruseonss cere eeneetkeieeerssaesekes 7 Neely v. Martin 4. by Construction Co., Inc., 386 U.S. ee te des ch ee nese aaa a's ag wh bg 5's 15 Rogers v. Lodge, 458 U.S. 613 (1982) ................ 9 Rossini v. Ogilvy & Mather, Inc., 798 F. 2d 590 (2d Cir. PAGE Watson v. Fort Worth Bank & Trust, __. U.S. ___, SOP S. Ch. SIRF CUBR s voce cicccrevenccccssasasane 16 Watson v. Fort Worth Bank & Trust, __. U.S. ___., ' § Se Mrrel|\,. PRPPPPrrererer ieee er passim Statutes and Regulations: ro Rik Foe >. rr rrr rr rrr rer eo. 3 Bod, RR. Clu. BP. ZOD occ cccccccccasdccestnaccaneces passim Sep. Ct. Ru 17 onc cccccccncesscccscescccecceasecses 1] IN THE Supreme Court of the United States OCTOBER TERM, 1988 No. 88-281 PATRICIA SHEEHAN, ELIZABETH HENOCH, and KAYHAN HELLRIEGEL, on Behalf of Themselves and All Others Similarly Situated, Petitioners, — PUROLATOR, INC. and PUROLATOR COURIER CORP., Respondents. ON PETITION FOR A WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT > BRIEF IN OPPOSITION Preliminary Statement In Generai Telephone Co. of the Southwest v. Falcon, 457 U.S. 147 (1982), this Court held that it was not appropriate to certify a class in a Title VII case unless the named plaintiffs sat- isfied the trial court, after rigorous analysis, that they shared common claims of discrimination with an identifiable class of similarly situated persons. In this Title VII case, the United States District Court for the Eastern District of New York, after reviewing the meager and 2 unpersuasive evidence presented by petitioners in support of their motion for class certification, ruled that petitioners had not met their burden under Falcon and declined to certify a class. On appeal, a panel of the United States Court of Appeals for the Second Circuit examined the same evidence and decided that the District Court's assessment of that evidence was not clearly erroneous. The Court of Appeals therefore ruled— unanimously—that the District Court had not erred in denying the class certification motion. Petitioners now ask this Court to review—for a third time-— whether their evidence was sufficient to sustain their burden under Falcon. To put it in petitioners’ own words, they ask this Court to find that, ‘*[t}he District Court committed clear error and abused its discretion’’ by not certifying a class, and that, ““[t}he Second Circuit erroneously disregarded . . . highly probative evidence”’ in affirming that decision. (Pet. 6) This Court, however, sits to decide questions of federal law that raise novel, general or important issues. It does not exercise its jurisdiction to review findings of fact that relate solely to an individual case. Indeed, as described in Graver Tank & Mfg. Co. v. Linde Air Products Co., 336 U.S. 271, 275 (1949), this Court is “‘a court of law. . . rather than a court for correction of errors in fact finding, [and] cannot undertake to review the concurrent findings of fact by two courts below. . . .** Under this long-established principle, nothing presented by this case is worthy of review by this Court. Petitioners also assert that certiorari should issue because the denial of class certification is inconsistent with this Court's recent decision in Watson v. Fort Worth Bank & Trust U.S. _ , 108 S. Ct. 2777 (1988), in which allegations that subjective employment practices are discriminatory were deemed amenable to analysis under the “‘disparate impact”’ model. But the propriety of certifying a class in this particular case presents no issue that warrants further review in light of Watson, because Watson did not modify the standards for class certification established by this Court in Falcon for all class actions, whether brought under a disparate impact or disparate treatment theory. Moreover, while the District Court rejected petitioners’ argu- ment that subjective employment practices could be analyzed under the disparate impact model, petitioners abandoned that argument when they took their appeal, as the Court of Appeals specifically noted. (App. 6)' This Court customarily declines to review questions that were not presented to the Court of Appeals; the fact that the issue was raised and then abandoned only reinforces the propriety of not making an exception to that rule in this case. Statement of the Case In February 1982, Patricia Sheehan and Elizabeth Henoch, along with a third plaintiff,’ filed a class action under Title VII, 42 U.S.C. § 2000(e). They alleged that Courier had discrimi- nated against them and other women on the basis of their sex in job assignment, pay, fringe benefits, promotion and transfer and had retaliated against them for filing discrimination charges with the EEOC. Each of the three plaintiffs had unique jobs at Courier. (App. $8-59, 64) They were all highly successful, well-paid employees. (App. 53, 59, 64) Two of the plaintiffs—Sheehan and Henoch—were officers of the company. The third was the sen- ior regional manager of Courier’s Chicago terminal and one of the highest paid managers—male or female—in Courier’s field operations. (App. 64) Proceedings Below District Court Decisions On March 23, 1983, petitioners moved for class certification. After extensive discovery and briefing, the motion was denied by the District Court on December 26, 1984. (App. 35) l Appendix references are to the petition appendix unless otherwise noted. 2 The third plaintiff, Kayhan Hellriegel, entered into a stipulation with Courier dismissing her complaint with prejudice. 4 The trial court articulated two bases for its decision. First, the District Court held that petitioners failed to show that an aggrieved class of similarly situated persons existed, as required by General Telephone Co. of the Southwest v. Falcon, 457 U.S. 147 (1982). Although petitioners introduced what they referred to as ‘‘significant proof that [Courier] operated under a general policy of discrimination,’’ Falcon, supra, 457 U.S. at 159, n.15, in support of their class certification motion, the Dis- trict Court concluded that the evidence did not amount to ‘‘sig- nificant proof’”’ of anything. The court ruled that the statistical evidence (including regres- sion analyses) submitted by petitioners in support of their motion was fundamentally flawed because it did not ‘‘offer the relevant comparison of similarly situated female and male employees (i.e., females and males with the same qualifications and experience). . . .’’ (App. 52) (Emphasis added) In light of the defects in petitioner’s statistical evidence, the non-statistical evidence about similarly aggrieved women that they introduced to bring ‘‘the cold numbers convincingly to life,’’ International Brotherhood of Teamsters v. United States, 431 U.S. 324, 339 (1977), assumed great importance. (App. 52) The District Court, however, found that evidence to be both inadequate and unpersuasive. It held that an affidavit from one member of the proposed class and a list of employees who had filed discrimination complaints against Courier—only three of whom would have been members of the proposed class—did not establish a class of aggrieved employees that bore ‘‘some Statistically significant relationship to the size of the relevant parts of the employer’s work force.’’ (App. 50, citing Falcon at 457 U.S. 157-158)’ In the alternative, the District Court concluded that class cer- tification was not warranted because petitioners, with their unique jobs and high ranking status, were atypical of the broad 3 The proposed class consisted of approximately 315 women, consti- tuting all women in jobs categorized as ‘‘exempt,’’ no matter what job they held. (App. 50) 2 class they sought to represent. (App. 54-59, 61-67)* The District Court judge acknowledged that a narrower class of high rank- ing women employees might have been appropriately certified, but noted, ‘‘[pji. intiffs have not articulated any basis for defin- ing such a class.’’ (App. 68-69) After class certification was denied, petitioners pressed their individual claims at an eight-day trial, following which the Dis- trict Court entered judgment in favor of Courier. (App. 22) The trial court, sitting as fact-finder, found that petitioners had failed to carry their burden of proof on any of their individual claims of discrimination. That court described petitioners’ evi- dence as ‘‘unpersuasive,’’ ‘‘lack[ing] merit’’ and ‘‘not credi- ble.’’ (App. 22-33) Petitioners introduced at trial the same flawed statistics that had been submitted and found wanting on the motion for class certification. They offered this evidence both in support of their individual claims and in support of a renewed request for class certification, which was made during trial. The District Court again held that the statistics proved nothing, this time remark- ing that the ‘‘statistical evidence is not persuasive because it does not take into account non-discriminatory variables that explain disparities in pay.’’ (App. 25) Court of Appeals Decision The Second Circuit Court of Appeals issued its opinion affirming the judgment on February 12, 1988. The panel unanimously affirmed the District Court’s denial of class certification. The Court of Appeals held that the Dis- trict Court had not erred in denying class certification, because petitioners failed to present ‘‘class-wide proof of an aggrieved class’’ as required by Falcon. (App. 7)° 4 Job categories within the proposed class included positions as diverse as officers and professionals, on the one hand, and customer service representatives and secretaries, on tiie other. (App. 57) 5 The Court of Appeals did not reach the alternate ground relied on by the District Court in denying class certification; namely, that peti- tioners were atypical of the class they wanted to represent. (App. 7) 6 In the view of the Court of Appeals, the District Court was not clearly erroneous in finding petitioners’ statistical evidence not probative of whether petitioners satisfied the requirements of Fed. R. Civ. P. 23(a), because ‘‘[t]he regression analysis did not take into account various relevant non-discriminatory fac- tors, such as education and prior work experience.’’ (App. 9) The Court of Appeals observed that Courier had presented evi- dence which showed that such factors could explain any dispari- ties between men and women to which petitioners pointed. (App. 9) The Second Circuit also affirmed the District Court’s conclu- sion that petitioners’ ‘‘limited amount’’ (App. 9) of anecdotal evidence about the existence of a class was insufficient to sup- port certification of a class. Reviewing the particular pieces of evidence, the Court of Appeals agreed with the District Court that a list of employees who had filed charges against Courier with the EEOC (only a few of whom were even members of the putative class) ‘‘had little probative value.’’ (App. 9) It also affirmed the District Court’s finding that a single affidavit from only one potential class member out of a proposed class of 315 members ‘‘was insufficient to establish a class of aggrieved individuals’’ under Rule 23(a). (App. 9) Finally, the Court of Appeals noted that statements from outdated personnel man- uals that petitioners had introduced did not establish that there existed a class of similarly situated women with grievances simi- lar to theirs because the manuals ‘‘were not in effect at corpo- rate headquarters,’’ where petitioners worked. (App. 9) By a 2-1 vote, the Court of Appeals also upheld the District Court’s decision to dismiss petitioners’ individual claims, ruling that the findings of fact made by the District Court with respect to those claims were also not clearly erroneous. (App. 16) Judge Kearse dissented from the decision to affirm the judgment dis- missing the individual claims, on the sole ground that she would have accorded more weight to certain non-statistical evidence presented at trial. (App. 17-21) On April 13, 1988, petitioners’ request for a rehearing was denied by the Court of Appeals. On August 15, 1988, respondents received the petition for a writ of certiorari. Petitioners sought review only of the Court of Appeals’ unanimous affirmance of the decision denying class certification. They did not ask this Court to review the dismissai of their individual claims. Reasons for Denying the Writ I. THIS CASE PRESENTS NO QUESTIONS OF GENERAL IMPORTANCE WARRANTING REVIEW BY THIS COURT A. Petitioners Are Seeking Review of Affirmed Findings of Fact, Which is Not the Province of This Court The Court of Appeals did not decide in this case any impor- tant question of federal law that requires further review. There is no conflict among the circuits on any point that needs to be resolved. Nor does this matter raise any novel, general or important legal issue. The Court of Appeals held only that the District Court did not clearly err when it found that petitioners had failed to meet their evidentiary burden of proving that a class of similarly aggrieved women existed, and, hence, that class certification was unwarranted. That ruling, which is of no moment to any case but this one, hardly raises an issue worthy of certiorari. Layne & Bowler v. Western Weli Works, 261 U.S. 387, 393 (1923) (Supreme Court will not review a case unless it involves principles of public importance). That the named plaintiffs in a Title VII case must convince a court that they are proper representatives of a defined class as a prerequisite to obtaining class certification has been settled ever since this Court recognized that . . the allegation that such discrimination has occurred neither determines whether a class action may be main- tained in accordance with Rule 23 nor defines the class that may be certified. Conceptually, there is a wide gap between (a) an individual’s claim that he has been denied a promotion on discriminatory grounds, and his otherwise unsupported allegation that the Company has a policy of discrimination, and (b) the existence of a class of persons who have suffered the same injury as that individual, such that the individual’s claim and the class claims will share common questions of law or fact and that the individual’s claim will be typical of the class claims. For respondent to bridge that gap, he must prove much more than the valid- ity of his own claim. Falcon, supra, 457 U.S. at 157. To bridge that gap, any putative class representative in a Title VII case must demonstrate that he or she and others share grievances against their employer that arise out of common questions of law and fact. Jd. at 158. Only ‘‘if the trial court is satisfied, after rigorous analysis, that the prerequisites of Rule 23(a) have been satisfied’’ may it certify a Title VII class. Jd. at 161. Petitioners do not contest the proposition that the District Court and Court of Appeals were bound to follow Falcon in this case. What petitioners do challenge is the District Court’s holding, affirmed by a unanimous Court of Appeals, that class certification was appropriately denied because the evidence they presented in support of their motion for class certification did not suffice to bridge the gap identified in Falcon. Before denying class certification, the District Court engaged in the rigorous analysis required by Falcon. It rested its decision squarely on factual findings that were thereafter reviewed by the Court of Appeals and found not to be clearly erroneous. Petitioners recognize this. The petition frames the issue raised by the case as follows: ‘‘The District Court committed clear error and abused its discretion in failing to recognize that sub- jective decisions at Purolator were being made in a highly sex- biased atmosphere and by officers and supervisors with company mandated sex-biased directives. The Second Circuit erroneously disregarded this highly probative evidence.’’ (Pet. 6) In page after page of the petition, petitioners catalogue the 9 evidence that allegedly proves their point—never once pointing to the rebuttal evidence that the trial court found more convinc- ing, and never once alluding to the actual findings of fact (including credibility findings) that were made and affirmed below. But this Court, unlike petitioners, does defer to factual deter- minations in which the District Court and Court of Appeals have concurred. It has repeatedly held that it will not review such findings. Rogers v. Lodge, 458 U.S. 613, 623 (1982); Branti v. Finkel, 445 U.S. 507, 512 n.6 (1980); Berenji v. Dis- trict Director, Immigration and Naturalization Service, 385 U.S. 630, 635 (1967). And where, as here, there is no dispute that the lower courts correctly followed controlling legal prece- dent in making and reviewing those factual findings, there is no basis for further review by this Court. B. Petitioners’ Attacks on the Decisions Below Under Eisen and Bazemore Do Not Raise Issues Worthy of Certiorari In an attempt to mask their request for further review of the District Court’s factual findings, petitioners also assert that the lower courts committed two errors of law which require certio- rari. The assertion is baseless. Petitioners first argue that the decisions below run afoul of this Court’s opinion in Eisen v. Carlisle & Jacquelin, 417 U.S. 156 (1974). According to petitioners, the lower courts, in con- travention of Eisen, required them to prove the merits of their claims of discrimination in order to obtain class certification. That is simply not the case. Under Eisen, a court cannot hold a hearing on the ultimate merits of the claim to determine whether to certify a class. But as this Court has observed in several post-Eisen decisions, ‘*it]he class determination generally involves considerations that are ‘enmeshed in the factual and legal issues comprising the plaintiff’s cause of action.’ ’’ Coopers & Lybrand v. Livesay, 437 U.S. 463, 469 (1978), cited in Falcon, supra, 457 U.S. at 160. The Court has never suggested that this enmeshing of 10 issues either contravenes Eisen or places an impossible burden on putative class plaintiffs. Here, neither the District Court nor the Court of Appeals held, or even suggested, that petitioners were required to prove the merits of their class claims in order to obtain certification. The lower courts only required, in accordance with Falcon, that petitioners satisfy the court that the prerequisites of Rule 23(a) had been met. That petitioners chose to do so by offering what they contended was ‘‘significant proof’’ of Courier’s discrimi- natory practices—evidence that was necessarily ‘‘enmeshed in the factual and legal issues’’ raised by their complaint—is of no moment. The District Court evaluated that evidence and found it insufficient to bridge the Fa/con gap. That does not run afoul of Eisen; it does not even implicate Eisen. Second, petitioners allege that the Court of Appeals erred in not overturning the denial of class certification because the Dis- trict Court failed to consider petitioners’ statistical evidence, as required by Bazemore v. Friday, 478 U.S. 385 (1986). As the Second Circuit held on appeal, this argument has no merit. As petitioners read Bazemore, the trial court was required to find that they had discharged their burden under Fa/con simply by offering statistical evidence to support their claim that other women had grievances similar to theirs—no matter how defec- tive those statistics might be. That, of course, is not what Baze- more says. All Bazemore holds is that a court may not ignore statistics simply because they do not control for all relevant var- iables. This Court also recognized in Bazemore that the failure to include all relevant variables would necessarily affect the weight accorded to statistical evidence. /d., 478 U.S. at 400. And in a post-Bazemore Title VII decision, Justice O’Connor observed, ‘‘courts or defendants [are not] obliged to assume that plaintiffs’ statistical evidence is reliable. ‘If the employer discerns fallacies or deficiencies in the data offered by the plain- tiff, he is free to adduce countervailing evidence of his own.’ ”’ Watson, supra, 108 S. Ct. at 2789. See also Teamsters, supra, 431 U.S. at 340 (“‘[S]tatistics, are not irrefutable, they come in infinite variety and, like any other kind of evidence, they may 1] be rebutted. In short, their usefulness depends on all of the sur- rounding facts and circumstances.’’) That is precisely what happened here. The District Court did not ignore petitioners’ statistics. That evidence was part of the record, both on the motion for class certification (App. 50-51) and at trial (App. 25, 30), and it was discussed in both lower court opinions. Contrary to petitioners’ assertion (Pet. 13), the statistical evidence was controverted; Courier introduced unre- butted evidence which showed that elementary non- discriminatory factors, such as education, prior job experience and job level, could (and did) explain any disparities in pay between men and women employees at Courier. (App. 9) Peti- tioners failed to control for those basic factors in their statisti- cal analyses. (App. 8-9) Obviously, the District Court, after rigorous analysis, found Courier’s countervailing evidence more persuasive than peti- tioners’ seriously flawed statistics, and its conclusion was affirmed by the Court of Appeals as not clearly erroneous. Such a finding, as the Court of Appeals held, does not run afoul of Bazemore. (App. 8-10) Thus, petitioners cannot identify any error of law that was committed by the Court of Appeals, let alone an error of law that qualifies as worthy of certiorari under Rule 17 of the Supreme Court Rules. There is, therefore, no ground for this Court to review the decision below. 12 il. WATSON V. FORT WORTH BANK & TRUST DOES NOT REQUIRE THAT THIS COURT REVIEW THE DENIAL OF CLASS CERTIFICATION A. Falcon, Not Watson, Controls The Propriety of Class Certification Petitioners also argue that certiorari should issue to review the decision denying class certification because the District Court held that the purely subjective employment practices at issue in this case could not be analyzed under the disparate impact model as well as the disparate treatment model. That holding, petitioners argue, does not accord with this Court’s recent decision in Watson v. Fort Worth Bank & Trust, U.S. , 108 S. Ct. 2777 (1988). The flaw in petitioners’ argument is that Watson did not con- cern class certification or anything related to class certification. It is Falcon, not Watson, that outlines the burden a putative class plaintiff must assume in order to obtain class certification. That burden is no different in disparate impact cases than in disparate treatment cases—to bridge the gap between individual and class claims by satisfying the court, after rigorous analysis, that the prerequisites of Rule 23(a) have been complied with. Nothing said in Watson alters that settled principle of law. Indeed, the Watson Court specifically declined to address any questions other than whether disparate impact analysis could apply to a subjective promotion system. Jd. at 2791. Thus, there is no support for petitioners’ assertion that Watson rendered the decision denying class certification erroneous in any respect. Further misreading Watson, petitioners suggest that Falcon has minimal or no relevance in disparate impact cases. (Pet. 9) However, that is clearly not correct. In Falcon, the class certi- fied contained members who had disparate impact claims as well as disparate treatment claims. Falcon, supra, 457 U.S. at 154. This Court overturned certification of the class on the ground that the lower courts had failed to require fhe named 13 plaintiff to prove that the case satisfied the requirements of Rule 23(a). At no point did the Court suggest that its discussion of the burdens imposed on putative class plaintiffs by Rule 23(a) applied only when disparate treatment complaints, rather than claims of disparate impact (whether objective or subjec- tive), were involved. And no subsequent decision of this Court—including Watson—has so held. But even assuming that Watson had some relevance on the question of class certification, certiorari would still be inappro- priate, because the findings of the District Court, as affirmed by the Court of Appeals, demonstrate conclusively that no class could be certified under any theory. As the Watson plurality noted, the burden on a plaintiff seek- ing to prove that subjective employment practices had a dispar- ate impact on a class of employees is not met as easily as in a case involving facially neutral practices, such as height require- ments or intelligence tests. The plaintiff must both identify the specific employment practice that is challenged and then offer statistics ‘‘of a kind and degree sufficient to show that the prac- tice in question’’ has resulted in the exclusion of those in a pro- tected group from some job benefit, such as employment or promotion. Watson at 2777-78. Further, courts and defendants are not ‘“‘obliged to assume that plaintiffs’ statistical evidence is reliable. ‘If the employer discerns fallacies or deficiencies in the data offered by the plaintiff, he is free to adduce countervailing evidence of his own.’ ’’ Watson at 2789. That is precisely what happened here. Respondents estab- lished, to the satisfaction of both courts below, that the statis- tics offered in support of petitioners’ motion for class certification were so fallacious and deficient as to prove nothing because: 1) they failed to offer comparisons between similarly situ- ated men and women (App. 9, 52); and 2) they failed to control for the most elementary non- discriminatory factors that can account for differences in treat- 14 ment among employees, such as education and prior work experience. (App. 9, 25) Nothing in Watson suggests that such patently defective sta- tistics would be any more acceptable as evidence that a dispar- ate impact class exists than they were under the disparate treatment model. Indeed, the plurality’s recognition of the need for rigorous statistical proof in a disparate impact case involv- ing subjective decisions compels the conclusion that the class certification decision would have been the same if this had been analyzed as a disparate impact case from the beginning. Relying on extensive quotations from Judge Kearse’s dissent, petitioners also assert that their non-statistical evidence alone would be sufficient to guarantee certification of a class under the disparate impact theory. But nothing in Watson suggests that the minimal and largely irrelevant anecdotal evidence rejected by the courts below would be sufficient, in and of itself, to certify a class under the disparate impact model. Moreover, petitioners’ reliance on Judge Kearse’s dissent is misplaced, inasmuch as she was concerned solely with evidence she thought should have been given more weight by the District Court in deciding petitioners’ individual claims. As petitioners recognize, (Pet. 5) Judge Kearse joined the majority in affirm- ing the District Court's denial of class certification. Even Judge Kearse did not believe petitioners’ anecdotal evidence was suffi- cient to warrant class certification. In sum, Watson neither vitiates a plaintiff's burden of prov- ing the existence of a class under Falcon nor endows evidence as flimsy as that introduced by petitioners with sufficient proba- tive value to sustain that burden. Thus, petitioners’ reliance on Watson is misplaced. 15 B. Petitioners Cannot Raise The Disparate Impact Issue Here Because They Abandoned It On Appeal In any event, petitioners are in no position to invoke Watson here because, while they raised the disparate impact issue in their motion for class certification, they abandoned it on appeal. (App. 46) Petitioners assert that the District Court's ruling on the dis- parate impact question was affirmed by the Court of Appeals. (Pet. 9) It was not. The briefs submitted to the appellate court by petitioners make no mention of the issue (Courier’s App. la- 88a), and the Court of Appeals took pains to note that the Dis- trict Court’s ruling on the question “‘has not been challenged on appeal.’’ (App. 6)° This Court has repeatedly held that it will not normally con- sider an issue that was not considered by the Court of Appeals. Adickes v. S.H. Kress & Co., 398 U.S. 144, 147 n.2 (1970) (‘‘Where issues are neither raised before nor considered by the Court of Appeals, this Court will not ordinarily consider them.’’); Delta Air Lines, Inc. v. August, 450 U.S. 346, 362 (1981); Neely v. Martin K. Eby Construction Co., Inc., 386 U.S. 317, 330 (1967). Petitioners obviously believe that, in light of Watson, this Court should invoke an exception to this well-established rule However, there is no reason to depart from customary practice in this case. Only in exceptional cases will this Court consider an issue that was not raised in the Court of Appeals; this case hardly qualifies as exceptional. And where the petitioner (like 6 The Court of Appeals stated: **The court analyzed the case as one of disparate impact, requiring proof of discriminatory motive. This find ing has not been challenged on appeal."’ (App. 6) It is obvious, Dorh from reading the District Court's decision and from the Court of Appeals’ reference to discriminatory motive (which only comes into play in disparate treatment cases) that the word “‘impact™’ should be ““treatment."* Petitioners recognize this. (Pet. 17 n.6) It is equally clear from context that this error is far more likely to be the result of care- less proofreading than some misunderstanding of the law, as petition- ers suggest. 16 petitioners here) has expressly waived reliance on a particular issue On appeal, this Court will not consider that issue on certio- rari, even in cases where the Court has addressed a previously unsettled point of law in a supervening opinion. Helvering v. Wood, 309 U.S. 344, 349 (1940). (Supreme Court declined to consider an issue when petitioner specifically waived reliance on that issue in the Court of Appeals, even though, under opinion announced in a companion case, Helvering v. Clifford, 309 U.S. 331 (1940), petitioner would have prevailed on the merits.) Attempting to avoid the impact of their waiver, petitioners apparently suggest that they should be excused from their fail- ure to press their disparate impact argument on appeal in this case because the Second Circuit had already rejected the identi- cal argument in a prior case. Rossini v. Ogilvy & Mather, Inc., 798 F. 2d 590 (2d Cir. 1986). (Pet. 5 n.2, 9 n.3) Such a conten- tion is unpersuasive for two reasons. First, the Second Circuit in Rossini did not even consider the question that petitioners failed to preserve here—whether dis- parate impact treatment for subjective employment practices would affect the propriety of certifying a class in a Title VII case. Thus, petitioners could not rely on the decision in Rossini to excuse their failure to raise the issue, even if the pendency of an adverse decision would otherwise have vitiated their waiver—which it would not. Second, petitioners’ failure to preserve their disparate impact argument is particularly inexplicable in light of the fact that the issue was the subject of conflict between the circuits; indeed, petitioners’ counsel (who were, coincidentally, counsel in Ros- sini ) made that very point to the Second Circuit in Rossini. /d., 798 F.2d at 605. And this court granted certiorari in Watson to resolve that conflict a mere three weeks after the District Court announced its decision in this case, which was well before the instant appeal was perfected. Watson v. Fort Worth Bank & Trust Co., cert. granted, _U.S. , 107 S. Ct. 3227 (June 22, 1987). Petitioners could have apprised the Second Circuit of the pendency of Watson and, if they deemed the case important or relevant here, they could even have moved for a stay of pro- 17 ceedings in the Court of Appeals pending the decision in Wat- son. That they chose not to do so, thereby depriving this Court of the Second Circuit’s views on the question they now pro- pound, speaks volumes about the irrelevance of Watson to the issue of class certification in this case. Conclusion The petition for a writ of certiorari should be denied. Dated: New York, New York September 13, 1988 Respectfully submitied, COLLEEN MCMAHON 1285 Avenue of the Americas New York, New York 10019 (212) 373-3000 Counsel of Record for Respondents Of Counsel: Morris B. Abram Jay Cohen Diana Hassel la No. 87-7540 United States Court of Appeals FOR THE SECOND CIRCUIT te PATRICIA SHEEHAN, ELIZABETH HENOCH, and KAYHAN HELLRIEGEL, on Behalf of Themselves and All Others Similarly Situated, Plaintiffs-Appellants, —against— PUROLATOR, INC. and PUROLATOR COURIER CORP., Defendants-Appellees. ON APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE EASTERN DISTRICT OF NEW YORK PLAINTIFFS-APPELLANTS’ BRIEF VLADECK, WALDMAN, ELIAS & ENGELHARD, P.C. Attorneys for Plaintiffs-Appellants 1501 Broadway, Suite 800 New York, New York 10036 (212) 354-8330 Of Counsel: Judith P. Vladeck Joseph J. Garcia Laura S. Schnell 2a TABLE OF CONTENTS TAREE CD ALR SUD ccc ckssccssescnsasecneen FIED WEG NMED oc cn cece vecsencenseaneeouapes POINT | DENIAL OF CLASS CERTIFICATION WAS AN Pe AOE CEC hb 6c nc nkg ns ackadkan donne A. The District Court Erroneously Required Affi- davits From Class Members When Plaintiffs Presented Substantial Evidence of Defendants’ Discriminatory Policies .................eee- B. The District Court’s Refusal To Certify The Plaintiffs As Class Representatives Was An Pa SE TORR «no vk Se a ex dvkeus POINT II THE DISTRICT COURT’S FINDINGS ON SHEEHAN’S AND HENOCH’S DISCRIMINA- TION CLAIMS ARE CLEARLY ERRONEOUS.. A. The District Court’s Failure To Consider The Discriminatory Atmosphere And The Sex- Biased Attitude Of Purolator’s Decisionmakers Constitutes Reversible Error PAGE 19 19 19 26 36 36 ean oan 3a PAGE B. The District Court Improperly Evaluated the Statistical Evidence........----++e essere eee: 39 POINT Ill THE DISTRICT COURT’S FAILURE TO EXAM- INE FULLY PLAINTIFFS’ RETALIATION CLAIMS REQUIRES REVERSAL .........----- 44 Pe a os 50 4a TABLE OF AUTHORITIES Cases: PAGE AFSCME y. County of Nassau, fe FEP Casts SB3 C.07.00. 8. PTs ohaktacncsaadencen 4 Anderson v. City of Bessemer City, 470 U.S. 564 (1985) 46 Avagliqano v. Sumitomo Shoki America, Inc., 104 ie ROR ae | ee rr re 28 Bauman vy. U.S. District Court, 557 F.2d 650 (9th Cir. oe) A er On BAPE Em eet y rer ENE t See u ee Met 21 Bazemore vy. Friday, aah , 106 S.Ct. 3000, 92 | Me BEd CUES ci ee Lao ae secu hw da 3, 7, 25, 42, 43, 44 Bell v. Birmingham Linen Service, 715 F.2d 1552 (11th Cees Woe ba ke kobe Ree ahd aneene sarees 36 Benson vy. Little Rock Hilton Inn, 30 Empl. Prac. Dec. Ser DS oe ee, Ps RE eka aan aeeeuuden eens 23 Carpenter v. Stephen F. Austin State University, 706 oe fe Ee oe. er er eee 28 Castaneda v. Partida, 430 U.S. 482 (1977)............ 8 Cox v. American Cast Iron Pipe Co., 784 F.2d 1546 (11th Cir. 1986), cert. denied, U.S. , 107 S.Ct. 274 CURR a 5 60 05 ioe 40 oP EES Oh ahha ois AGES 20, 21, 28 Davis v. State University of New York, 802 F.2d 638 (2d Cs Sd sine has Ws eae aR eee ae 44 DeCintio v. Westchester County Medical Center, F.2d , 44 FEP Cases 32 (2d Cir. 1987)...... 44, 45, 49 De La Fuente v. Stokley-Van Camp, Inc., 713 F.2d 225 (7th Cir. 1983) SCH ESL ECHABHHRSEBEA HEC CCE ALH OD TAB E ODE SE 6 0 ig. anon Sa PAGE Dominic v. Consolidated Edison Co., F.2d ? eo et UU Re ef rere 45 Duke v. University of Texas at El Paso 729 F.2d 994 (Sth Cir.), cert. denied, U.S. , 105 S.Ct. 386 RE Wns WHO's oe ke RES Ee CNT REWER Seas 7 Eastland v. TVA, 704 F.2d 613 (11th Cir. 1983), cert. Ge ia FO CR evans cen donccaakadnes 43 Eisen v. Carlisle, 417 U.S. 1567 (1974) ............... 28, 35 Equal Employment Opportunity Commission v. Sage Realty Corp., 87 F.R.D. 365 (S.D.N.Y. 1980)....... 39 Francis vy. American Telephone & Telegraph Co., 55 Fee. ee NEED ci iio cb wlkn saab eda Raa des 47 General Telephone Company of the Southwest v. Falcon, ee Pe I io hes oe cdg oe uaale eee oe 3, 4, 20, 27 Grant v. Bethlehem Steel Corp., 622 F.2d 43 (2d Cir. SN ceed acn each Ke ed Cte bb hae eae eee 45 Grant v. Morgan Guaranty, 548 F.Supp. 1189 (S.D.N.Y. 264k CERES CLEAN R VEES RGN a OREN 22 Griffin v. City of Omaha, 785 F.2d 620 (8th Cir. 1985) 36 Hansen v. Webster, 41 FEP Cases 223 (D.D.C. 1986).. 29 Hazelwood School District v. United States, 433 U.S. 299 SE See Doce ie teat ade bug ease sloaede ae 8 Herbst v. Able, 46 F.R.D. 11 (S.D.N.Y. 1969) ........ 35 Holsey v. Armour & Co., 743 F.2d 199 (4th Cir. 1984). 35 Hopkins v. Price Waterhouse, F.2d ___, 44 FEP eee eee ee, ED oh oh ec ba waa ke eee ncn ae 38 6a PAGE Jordan v. Wilson, 649 F.Supp. 1038 (M.D. Ala. 1986) . 39 Lilly v. Harris Teeter Supermarket, 720 F.2d 326 (4th Cir. 1983), cert. denied, 466 U.S. 951 (1984)........ 27 Lo Re v. Chase Manhattan Corp., 431 F.Supp. 189 (S.D.N.Y. TS7F bs ik cceceeakeea eres 33 Martin vy. Eastern Pub. Co., 73 F.R.D. 678 (E.D. Pa. IDTT). oc cccrivccncsdawiae pase ee eee ee 34 McDonnell Douglas v. Green, 411 U.S. 792 (1973)..... 45 McKenna v. Weinberger, 729 F.2d 789 (D.C. Cir. 1984) 38 McKenzie v. Sawyer, 684 F.2d 62 (D.C. Cir. 1982) .... 28 Mersey v. First Republic Corporation, 43 F.R.D. 465 (S.D.N. 7. YOR. i oscsca va cane eens 35 Meyer v. MacMillan Publishing Co., Inc., 95 F.R.D. 411 (S.D.N.Y . FO) 66nd dense 33 Moffett v. Gene B. Glick Co., 621 F. Supp. 244 (N.D. ind. 96S) 2 occa kceec¢asaae eee eee 46 Mosley vy. General Motors Corp., 497 F. Supp. 583, aff'd, 691 F.2d S04 (Gth Cir, JOBE)... occ ncccacesc 49 Odom vy. U.S. Homes Corp., 76 F.R.D. 381 (S.D. Tex. W975)... ucccsthans desde eee eee 34 Richardson v. Restaurant Marketing Associates, Inc., 83 F.B.D. 268 GN.0D. Cae. Biles see ee ee 24 Rossini v. Ogilvy & Mather, 798 F.2d 590 (2d Cir. | POPPER ere ry rere el he ery ra oe passim Rowe v. Bailar, 26 FEP Cases 1145 (D.D.C. 1981)..... 31 Rule v. Iron Workers, Local 296, 568 F.2d 558 (8th Cir. + cna lll 7a PAGE Sheehan v. Purolator, 103 F.R.D. 641 (E.D.N.Y. Sai penis PPPereeee reer erereee errr e eee passim Sheehan v. Purolator, 676 F.2d 877 (2d Cir. 1981)..... 2 Sheehan v. Purolator, 36 FEP Cases 1449 (E.D.N.Y. eee ia sua seaseceaescasesecss 2 Society for Good Will to Retarded Children v. Cuomo, ee 19 Sperry Rand Corp. v. Larson, 554 F.2d 868 (8th Cir. Os 35 Steffin v. First Charter Financial Corporation, 77 F.R.D. es dei aheacecseceeadicdusss 26 Teamsters v. United States, 431 U.S. 324 (1977)....... 7, 25 Thompkins v. Morris Brown College, 752 F.2d 558 (11th Cs eel beedesaceuneeenewes errr. 36 Warren v. ITT World Communications, 95 F.R.D. 425 Ee a 23 Wheeler v. City Of Columbus, 686 F.2d 1144 (Sth Cir. ee eC aaiddesbscecewans 38 Womack v. Munson, 619 F.2d 1292 (8th Cir.), cert. Se ee EO, PW COMED caso cect cesecsscascecs 45 Williams v. Boorstin, 451 F. Supp. 1117 (D. D.C. 1978), rev’d on other grounds, 663 F.2d 109 (D.C. Cir. 1980), ey SS EE UP. DOS CEDUED ccc cece ticscesees 34 Wright v. Stone Container Corp. , 524 F.2d 1058 (8th Cir. 8a PAGE Statutes and Rules: ra ery passim Ge SPs NEED 6 645d Saks cee tescccdseducawene 2 Rule 23, Federal Rules of Civil Procedure............ passim Treatise: Manual for Complex Litigation ..................... 20, 21 9a ISSUES PRESENTED 1. Whether the District Court erred in its class certification decision in requiring affidavits from class members when plain- tiffs presented substantial evidence of defendants’ discrimina- tory policies, discriminatory attitude and statistical evidence of discrimination. 2. Whether the District Court erred in holding that the plain- tiffs were inadequate class representatives. 3. Whether the District Court clearly erred in finding that plaintiffs did not suffer from discrimination, when it failed to consider the discriminatory atmosphere at Purolator and the sex-biased attitude of defendants’ officials, and when it improperly rejected plaintiffs’ statistical evidence. 4. Whether the District Court’s findings of no retaliation against Sheehan and Henoch are clearly erroneous. 10a UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT Docket No. 87-7540 ~~ PATRICIA SHEBHAN, ELIZABETH HENOCH, and KAYHAN HELLRIEGEL, on behalf of themselves and all others simi- larly situated, Plaintiffs-Appellants, —against — PUROLATOR, INC. and PUROLATOR COURIER CORP., Defendants-Appellees. STATEMENT OF THE CASE This action was brought pursuant to Title VII of the Civil Rights Act of 1964, as amended, 42 U.S.C. § 2000e ef seq. (‘Tithe VII'"), by plaintiffs Patricia Sheehan (‘‘Sheehan’’), Elizabeth Henoch (‘‘Henoch’’) and Kayhan Hellriegel (‘*Hellriegel’’), on behalf of themselves and a proposed class of female exempt (salaried) employees and former employees of Purolator Courier Corporation and Purolator Inc. (‘‘defendants’’/**Purolator’’). The Complaint generally alleged that defendants unlawfully discriminated against plaintiffs, and the class of women they sought to represent, in assigning them to lower paying, lower status positions, in promotion and trans- fer opportunities and in the payment to them of lower salaries and other benefits than those given to similarly-situated males. In addition, plaintiffs Sheehan and Henoch asserted that they were retaliated against for having filed charges of discrimina- tion. See Tithe VII, 42 U.S.C. [2] § 2000e-3(a). Plaintiffs requested injunctive relief and monetary compensation for back pay and other benefits lost by them and by members of the proposed class as a result of the defendants’ unlawful practices. a In March, 1983 plaintiffs moved, pursuant to Rule 23(b)(2) of the Federal Rules of Civil Procedure, for class certification. ' That motion was denied by the Hon. |. Leo Glasser in Decem- ber, 1984. The Court’s decision is reported at 103 F.R.D. 641. After the class certification motion was denied, Hellriege! and Purolator entered into a settlement agreement resolving all of her claims. The claims of discrimination and retaliation’ alleged by plain- tiffs Sheehan and Henoch were tried to the Court from June 24 through July 3, 1985. The District Court issued its decision on May 27, 1987, dismissing all of plaintiffs’ claims. That decision is unreported. (See J.A. 3224-38). Plaintiffs renewed their motion for class certification at the conclusion of the trial on the merits (J.A. 3152-53) and, although in its decision on the merits the District Court acknowledged that [3] plaintiffs’ renewed motion was pending, it did not comment on it further. (See J.A. 3224-38). Judgment was entered on June 2, 1987. On June 30, 1987, plaintiffs Sheehan and Henoch appealed from all adverse opinions and orders of the Court below. (J.A. 3240- 41). SUMMARY OF ARGUMENT Plaintiffs appeal from the denial of their motion for class cer- tification pursuant to Rule 23 of the Federal Rules of Civil Pro- cedure and from the dismissal of their individual claims, which alleged sex discrimination in the terms and conditions of their ! Many months of discovery, at large expense to plaintiffs, were wasted in resisting defendants’ motion to dismiss the complaint as against Purolator, Inc. Well after all papers were filed on the class certification motion, Purolator, Inc. and Purolator Courier announced consolidation of their headquarters and that the executive officers of Purolator, Inc. were assuming responsibility for their respective functions at Courier. This made de jure what plaintiffs had asserted was the fact at an earlier time. (See Joint Appendix 1199-1200; “J.A.”*). 2 = This District Court earlier denied defendants’ motion for summary judgment with respect to Sheehan's retaliation complaint. 6 FEP Cases 1449 (1982). This case was before this Court in 1981 when the Court reversed the District Court's holding that it was without jurisdiction to hear Sheehan's retaliation claim. 676 F.2d 877 (1981) l2a employment, and retaliation for having complained of discrimi- nation. In its analysis of plaintiffs’ application for class certification, the Court below ignored plaintiffs’ factual showing and made fundamental errors of law in applying General Telephone Com- pany of the Southwest v. Falcon, 457 U.S. 147 (1982). These errors resulted not only in the incorrect initial denial of the motion, but also permeated the Court's erroneous assessment of the merits of the individual plaintiffs’ claims. The District Court’s failure to consider or rule on the renewed motion for class certification in its final opinion is inexplicable and in disre- gard of the additional evidentiary support for the motion devel- oped on trial. It also is contrary to controlling decisions of this Court and the United States Supreme Court issued after the trial, but before issuance of the Court’s opinion and order of May 27, 1987. See Bazemore vy. [4] Friday, 478 U.S. , 106 S.Ct. 3000, 92 L.Ed.2d 315 (1986); Rossini v. Ogilvy & Mather, 798 F.2d 590 (2d Cir. 1986). The Court’s rulings not only disregard the requirements of Rule 23(c)(1), but also reject sub silentio the express admonition of this Court in Rossini that ‘‘the primary thrust of Falcon was that satisfaction of Rule 23a) requirements may not be pre- sumed’’ and that the Supreme Court in Falcon ‘‘did not hold that a plaintiff asserting one particular type of claim can never represent a class of employees asserting other types."’ 798 F.2d at 597 (emphasis in original). In addition, had the District Court applied Bazemore, 92 L.Ed.2d at 329-33, to the statistical evidence, it would have been obliged to credit plaintiffs’ proof both for purposes of class certification and liability.’ 3 The Court's failure to address Rossini is even more baffling in view of the citation to it in an opinion issued by the same Court less than two months after it issued its final opinion in this case, in which the Court granted class certification on allegations similar to those herein. See AFSCME v. County of Nassau, F.Supp. , 44 FEP Cases 583, $86 (E.D.N.Y. 1987). Applying the law as it was set forth in AFSCME, id. at 585-86, the Court could not have denied plaintiffs’ motion for class certifi- cation here. There, the Court, in what it deemed a disparate treatment case l3a The Court below misapprehended the law, and its obligations under Rule 23(c)(1), in the following critical respects: 1. The Court ruled that, since Falcon, class representatives are required to supply affidavits from class members to demon- strate not only the ‘‘existence of an aggrieved class’’, but also that other class members ‘‘feel aggrieved’’ in order to show typ- icality. 103 F.R.D. at 648-49. In so doing, the Court errone- ously required what in effect was an opt-in procedure. [5] 2. The Court, ignoring the facts and prevailing law, found that Sheehan, Henoch, and Hellriegel could not represent a class of all female exempt employees because they occupied ‘‘relatively high level’’ positions. 103 F.R.D. at 650, 653, 655. 3. The Court ruled that Sheehan, Henoch and Hellriegel could not be proper class representatives because they asserted that they had suffered unlawful retaliation, in addition to their class wide claims of discrimination in compensation, promo- tion, and other conditions of employment. 103 F.R.D. 652, 653-54, 656. The District Court also made far-reaching and fundamental errors in evaluating the evidence before it, at both the class cer- tification and trial stages of the litigation. The Court committed clear error and abused its discretion in failing to recognize that decisions were being made in a highly sex-biased atmosphere and by officers and supervisors with company mandated sex- biased directives. Thus, 1. The District Court never mentioned Purolator’s blatantly sex-biased written policies, which were distributed to all officers of the company and ‘‘to all managers within the field organiza- tion, terminal mangers, [and] those people having direct day to day responsibility for running the operations,’’ and which were issued in 1971, renewed in 1975, and not withdrawn until of salary discrimination, certified a class of all female Nassau County civil service employees who were employed in traditionally female jobs (those hav- ing 70% or more females). l4a shortly before trial in 1983 or 1984. (J.A. 2101-02, 2106; E.A. 1). Those policies and interperations included: (a) An express belief that ‘‘{t]he sex of the individual is more often related to job success than one might expect.”’ (Exhibit Appendix 1; ‘‘E.A. ___’’). ‘‘The sex of the employee is often related to turnover.’’ (/d.) (b) An admonition that the amount of responsibility an employee may be allowed to assume should be related to the number of dependents. Specifically, ‘‘[i]n the case of women applicants, the number and ages of children are important.’’ Managers and supervisors also are directed to determine how the children are to be cared for while the ‘*mother’’ works. (/d.) (c) An express preference for promoting men by instructing managers and supervisors that, ‘‘before decid- ing on a man for promotion’”’ consider such highly subjec- tive criteria such as the candidate’s ‘‘heat [6] resistance’’ and loyalty because ‘‘being loyaltothe Company. . . . is a quality of men who are responsible for the success of its operation.’’ (J.A. 726) (emphasis added). The District Court also ignored testimony that showed the sex-biased attitude of Purolator officials: 2. The District Court erroneously ignored an admission by Sheehan’s supervisor that he did not take her request for trans- fer seriously because he knew she had children. (J.A. 2203). This testimony not only demonstrated the biased attitude of her supervisor, but also showed that he took seriously and imple- mented the express policy that women with children were not to be given responsibility within Purolator. 3. It was erroneous for the District Court to ignore the testi- mony of a former male manager that he was told to pay women low salaries because they had husbands, and that based upon his management experience within Purolator, he observed women were paid less than their male counterparts, and that it was more difficult to get promotions or salary increases for women than for men. (E.A. 362-63, 365-66, 394, 399). lSa 4. The District Court erroneously refused to allow a witness to testify at trial that John Delany (‘‘Delany’’) (both Henoch’s and Sheehan’s supervisor) told her she would not be considered for a position because no one would be respectful of a woman in that position. (J.A. 2524-25). 5. The District Court completely ignored testimony support ing Purolator’s former affirmative action officer that Purolator only paid lip service to affirmative action (B.A. 125; J.A. 2174- 77), including the written admission that women in Purolator’s work force are underutilized as officials and managers (E.A. 3) and that Purolator’s affirmative action deficiencies disqualified it for certain government contracts. (J.A. 2277-79) 6. The District Court erroneously rejected plaintiffs’ statis- tics in foto on the ground that plaintiffs’ regression analysis, which included all exempt employees, did not have variables for education and prior experience, even though defendant’s expert testified that he could not state what effect, if any, including those variables would have on the results—and defendants offered no study of all exempt employees although it had the data available to do so. These errors, independently and cumulatively, so tainted the District Court’s evaluation of plaintiffs’ case as to require reversal and vacation of both the class certification [7] decision and the decision on the merits. Both Sheehan’s and Henoch’s individual claims were unfairly burdened by the Court’s errone- ous Class certification determination. As the Fifth Circuit noted in Duke v. University of Texas ai El Paso, 729 ¥.2d 994, 997 (Sth Cir.), cert. denied, 469 U.S. 982 (1984), ‘‘had [plaintiff’s} class claims prevailed, she would have faced a distinctly less onerous burden at the trial of her individual claim.’’ See also Rossini, 798 F.2d at 606. Findings in favor of the class would have simplified plaintiffs’ proof in the individual! claims and shifted the burden of proof to Purolator on their claims. See Teamsters v. United States, 431 U.S. 324, 362 (1977) l6a FACTS Work Force Distribution Purolator is engaged in the business of providing courier ser- vice over ground and air routes. Of the 1,113 salaried employ- ees in 1982, 798 were male and 315 female; women thus comprised approximately 28% of defendants’ exempt (salaried) work force. (J.A. 727).* In Corporate Headquarters, of 109 sal- aried positions, 34 were occupied by women (31%); in the East- ern division, of 482 salaried employees, 155 were women (32%); in the Western division, of 240 salaried employees, 67 were women (28%); and the Southern division, 57 of 264 sala- ried employees were women (22%). (J.A. 727-28) [8] The women represented in Purolator’s exempt work force were at the lowest levels and in a few job categories: in 1982, there were 39 officers, of whom only two were women; of 32 directors, only one was a female; there were no female senior managers, and of the 222 managers, only 31 were female. (J.A. 729). Almost one-half of the male exempt employees were in the high salaried, upper level ranks; about one-fourth of the women were at those levels. (See id.) Purolator admitted that it underutilized women in manage- rial, professional, and sales areas. (E.A. 3). It was also shown that from 1976 forward, 60% to 88% of all exempt job titles were sex segregated. (J.A. 730).° The facts were confirmed by plaintiffs’ unchallenged statistical analysis, which showed that there was a statistically significant underrepresentation of 4 Plaintiffs sought to represent a class comprised of defendants’ female exempt employees. The facts provided herein are derived from the record the District Court had before it for initial determination of the Rule 23 motion and after trial. 5 There are no jobs in Purolator that require male only/female only assignments. Purolator has a promote from within policy, and as Paul Wolfrum, a Senior Vice President testified, a high school education, ‘‘with experience’’ was sufficient qualification for most jobs. (J.A. 2860). hae TatSe 17a women in Purolator’s exempt work force, when compared with a census-derived nationally qualified work force. (E.A. 110).° Criteria Used In Personnel Decision Making The District Court was provided with, and totally ignored, direct evidence of highly subjective and gender-related [9] crite- ria applied by Purolator in all aspects of employment.’ The Supervisory Manual instructs personnel conducting initial inter- views of applicants for jobs that it is ‘‘especially important to detect’’ the applicant’s marital status and ages of any children. (J.A. 723). The Manual goes on to explain: Inahe case of women applicants, the number and ages of children are important. Follow-up in the interview should ascertain how the children are to be taken care of while the mother works and whether it will be necessary for her to leave her job in the event that they are sick. (E.A. 1). The Supervisory Manual explains that ‘‘the sex of the individual is more often related to job success than one might expect’? and further that the ‘‘sex of the employee is related to turnover.’’ (E.A. 1). These policies remained in effect at least until 1983 or 1984. (J.A. 2102). Two high level company officials, James Rapp (‘‘Rapp’’) and Paul Wolfrum (‘‘Wolfrum’’), confirmed that promotions are based on supervisors’ selections, for which no objective qualifications are established; nor is there job posting within Purolator. Both Rapp and Wolfrum testified that they never had to request a promotion to a higher position; rather, each 6 The results of this study were statistically significant at two standard deviations or greater. See Castaneda vy. Partida, 430 U.S. 482, 497 n.17 (1977). ‘‘As a general rule for such large samples, if the difference between the expected value and the observed value is greater than two or three stan- dard deviations,’’ then the hypothesis that exempt employees at Purolator were hired without regard to sex would be suspect. Jd. See also Hazelwood School District v. United States, 433 U.S. 299, 309 n.14 (1977). 7 The Court ignored this direct evidence both in the denial of class cer- tification and in its ultimate decision on liability. The evidence was uncon- troverted and clearly highly relevant; the Court simply made no reference to it. 18a was tapped for each promotion by another male superior. (See J.A. 668; 2862). In conformity with that practice, Purolator’s manuals instruct supervisors that, ‘‘before deciding on a man for promotion,’’ they should weigh subjective criteria, such as the candidates’ ‘‘heat resistance’’ or ability ‘‘to handle himself when [the] heat is on’’, [10] whether he is a ‘‘self-starter’’, or is ‘‘loyal’’ because ‘‘being loyal to the Company. . . is a quality required of men who are responsible for the success of its oper- ation.’’ (J.A. 668). The trial evidence also showed, without contradiction, that Sheehan’s supervisor refused to ‘‘take [ ] seriously’’ her request for transfer to a field position because she had children. (J.A. 2203).° A former male manager testified that he was told to pay women low salaries because they had husbands, that women were paid less than their male counterparts and that based upon his management experience within Purolator, it was more diffi- cult to get salary increases and promotions for women than for men. (E.A. 362-63, 365-66, 394, 399). Finally, Purolator’s former affirmative action officer believed that affirmative action at Purolator was given only lip service. (E.A. 125; J.A. 2174-77). This was borne out by the admission in Purolator’s affirmative action plan that women were underutilized at the official managerial levels. (E.A. 3-4). Another witness testified without contradiction that Purola- tor’s affirmative action deficiencies disqualified it for certain government contracts. (J.A. 2277-79). 8 Another female employee would have testified that Delany (Sheehan’s and Henoch’s supervisor) told her that she would not be consid- ered for a position because no one would be respectful of a woman in it. (J.A. 2524-25). The Court did not permit the employee to testify on the ground that her testimony went to the classwide issues. In so doing, the Court erroneously rejected the argument that the testimony was relevant to show Delany’s sex-biased attitude in personnel decisionmaking. (See infra, pp. 38-39). 19a Salary Disparities [11] Plaintiffs established that women were not only relegated to a limited number of jobs, but that they were paid far less than men who occupied male only jobs. Thus, plaintiffs’ evi- dence showed: Average Weekly Average Weekly Salary of Female- Salary of Male- Difference Only Job Titles Only Job Titles Weekly Annually 1976 $248 $364 -$116 -$ 6,032 1977 293 437 - 144 - 7,488 1978 261 444 - 183 - 9,516 1979 277 476 - 199 - 10,348 1980 311 513 - 202 - 10,504 1981 31] 621 - 310 - 16,120 1982 362 711 - 349, - 18,148 (E.A. 108; J.A. 733). In 1982, the mean salary for all exempt employees was $435.41 per week. The mean salary of the 798 male exempts was $490.21 per week; the mean salary of the 315 female exempts was $296.58. That difference between male and female salaries translated into an annual disparity of $10,068.76 in favor of men. (E.A. 107; J.A. 733-34). Gross salary differences between Purolator’s male and female exempt employees were shown to exist even within job groupings. For example, in 1982, male managers earned $622 per week—female managers earned $387 (an annual difference of $12,220); similarly, in 1982, male supervisors earned $450 per week—female supervisors earned $322 (an annual differ- ence of $6,656). The male salary advantage persisted even where males and females with comparable tenure were com- pared. (j.A. 734-35). [12] Plaintiffs’ regression analysis, which remained unrebutted by defendants (see, infra, p. 41) showed that when sex, age of employee, tenure with Purolator, and job groupings were fac- tored and considered simultaneously, male salary domi- 20a nance persisted at a statistically significant level of two standard deviations or greater: Weekly Annual Male-Female Salary Male-Female Salary Difference Difference 1980 -$60 ~$3,120 1981 - 81 - 4,212 1982 - 99 - 5,148 (J.A. 735; E.A. 106).’ Promotion and Training Plaintiffs’ evidence established, through proof of longstand- ing sex-segregation of Purolator’s work force and gross male salary advantage, that defendants’ promote from within policy was not applied evenhandedly and that Purolator’s written preference for promoting men over women (see J.A. 726) was in fact applied to personnel decisions. Promotion discrimination also was demonstrated by the fact that men who achieve high-level, male-only positions are pro- moted out of job categories occupied by both sexes. For exam- ple, in 1982, men who were in the male-only job categories of Division Transportation Manager, Regional Operations Man- ager, [13] District Operations Manager, Divisional Vice Presi- dent, District Manager, and Account Executive had at one time occupied sex-integrated ‘‘feeder’’ positions of courier, courier Supervisor, dispatcher, and sales representative. (J.A. 735). Plaintiffs established, without contradiction, that during the years 1981 and 1982, ten men and no women were appointed to defendant’s Management Training program, and that since 1975, thirty-nine men and five women held the position of Man- agement Trainee. (E.A. 113). In 1982, a woman who had gone 9 When geographic location was factored in, the male-female salary difference increased; to wit, in 1980—$3,432; 1981—$4,524; and 1982— $5,356. (J.A. 1258). These disparities were statistically significant at two standard deviations or greater. Pa RAR ae 0 tt BRB OO 8! ont esta i PD nts: 2la through the Training Program was earning approximately $108.00 less per week than the male who had completed that same Program. (E.A. 113). Women also have been denied the right to certain important fringe benefits, including corporate automobiles, a not incon- siderable supplementary form of compensation. (See, infra, p. 41). Personnel Policies Purolator is highly centralized and all decisions affecting its employees at all locations emanate from Corporate Headquar- ters. All business decisions come from Headquarters; all personnel decisions, including job descriptions, lines of pro- gression, hire, discipline, termination and salary are dictated by Headquarters. Uniformity is achieved through periodic meet- ings with management personnel and issuance by Headquarters of standarized policies and procedures. (See J.A. 707-22). [14] Additionally, for many personnel matters, managers from headquarters go into the field to assure uniformity of response to particular problems. (See J.A. 712-22). Purolator’s efforts to portray itself as decentralized were put to rest during trial, when its former Senior Vice President for Human Resources explained his resignation in 1982 as arising from his philosophi- cal differences with Senior Management: He espoused decen- tralization; they insisted on strict centralization. (J.A. 2156). Claims of the Proposed Class Representatives The plaintiffs are women who were employed by Purolator in exempt positions.'° Plaintiffs’ claims were typical of those of the proposed class. Sheehan and Henoch asserted that, as other class members had been, they had been discriminated against in the terms and conditions of their employment because they were women; that they had been denied opportunities open to similarly qualified or less qualified men; and were paid low sal- aries and denied other benefits because they were women. 10 ~=At the time of the class certification motion, Sheehan’s employment had been terminated; Hellriegel had quit, claiming that harassment had made her continued employment impossible; Henoch remained employed. 22a Just before she was fired, Sheehan was classified as a Staff Vice President, Administration. She alleged that her employ- ment was unlawfully terminated on August 5, 1981 in retalia- tion for her complaints against defendant. (J.A. 34). Prior to joining Purolator in 1969, she had been employed as a [15] manager with another company for nine years. (J.A. 697-98, 1785). From early in her employment, Sheehan sought opportunities to move from a staff position at Corporate Headquarters into a line position in the field, where there were better opportunities for advancement and increased salary. (J.A. 698, 1819-22). She repeatedly expressed to top management her observation that women in the Corporation were being denied a chance to move into higher level positions and were being grossly under- compensated in comparison to similar males. (J.A. 698, 1819, 1838, 1840). She pointed out that a simple head count would confirm the longstanding practice of excluding women from upper-level jobs. (J.A. 698). John Nichols (‘‘Nichols’’), one of the supervisors, recorded for the file Sheehan’s discussions with him reflecting her efforts to obtain a fair salary and promotion opportunity. One of his memoranda quotes her as saying in 1980 that she would wait to see what the new President’s ‘‘pol- icy toward women and men in similar positions was.’’ (E.A. 8). Sheenan’s requests for equity for herself and other women in the Corporation first were rebuffed by either the excuse that there were no available positions, although she saw her male subordinates getting the jobs she had sought, or the promise to ‘*get back to her’’. (J.A. 698). Nichols, Sheehan’s supervisor at the time. testified at trial that he did not take seriously Sheeha, . request for transfer because she had children (J.A. 2203). [16] Henoch received a B.S. in Marketing from New York Uni- versity in 1947. (J.A. 700, 2233). At the time Henoch was hired as a secretary by defendant in 1967, she had acquired business experience. (J.A. 700, 2233-34). Shortly after her hire, Henoch became responsible for handling applications to the Interstate Commerce Commission (‘‘ICC’’) for Purolator’s tariff approv- als. (J.A. 700, 2239-43). On her own time and at her own expense, Henoch pursued a course to become licensed as an ICC practitioner. (J.A. 700). — a 23a Despite the responsibilities she had assumed in the handling of tariffs, and despite the title changes which resulted in her ulti- mately acquiring the title of Staff Vice President, Henoch’s sal- ary did not keep pace with that of men who were promoted faster or paid more than she, although their responsibilities and qualifications were comparable to hers. (J.A. 700). Like Sheehan, Henoch was denied use of a company automobile while her male counterparts enjoyed this perquisite. (J.A. 700- 01). Henoch’s requests for consideration for transfers and for promotions were ignored by Purolator. During the seventies, she applied for a position in the Marketing Department, because she had a marketing degree and was seeking opportuni- ties to advance. (J.A. 2247-48). Her application, in her words, was ‘‘not taken seriously.’’ (J.A. 2248). Henoch spoke on a number of occasions to Delany, Nichols and to Executive Vice President, James Perry (‘‘Perry’’), about the slowness of her progress in the Company. (J.A. 2311-12). In those conversa- tions, Henoch [17] emphasized that she was seeking a line posi- tion with decision-making functions. (J.A. 2312). She was never offered such an opportunity. Like Sheehan, Henoch was denied salary and bonuses equiv- alent to those granted similarly qualified males. As early as 1975, Henoch began to complain about her salary, after she saw less qualified and lower ranking males being hired at higher wages than she was making. (J.A. 2308). Retaliation Within months of filing her EEOC charge Sheehan was fired. Almost immediately after she filed the charge, her responsibili- ties were changed drastically. (See J.A. 1908-11, 1894-95). Her long-time subordinates were called into a meeting to witness her demotion. She was the target of insults; her ability and integrity were impugned. Her hours were montiored, she was yelled at by her supervisors and a memo-writing campaign was designed to keep her on the defensive. (See J.A. 1902, 1904-07). Sheehan finally reached the end of her rope when she was required to endure a meeting to review her job responsibilities 24a and was asked whether her main job function was typing the purchase orders. (J.A. 1911-12). Rather than blow up and cause a scene, as other Purolator employees were wont to do, Sheehan simply left the meeting. When summoned back to the meeting, she returned as directed, only to be accused of having constructed a false job description. She again left and refused to go back because she ‘‘was so upset and could not take any- more.’* {18} (J.A. 1912-20). She was immediately fired for what the District Court called ‘‘gross insubordination.’’ (J.A. 3225). Sheehan was forced to leave Purolator’s premises immedi- ately and was denied severance pay. As a result, Sheehan was not permitted to vest in her pension (which she would have done just one month later), and her son was forced to withdraw from college in order to support his mother. (J.A. 1923). The District Court’s own findings show the infirmity of its ruling that Sheehan was not retaliated against. The Court found that ‘*‘Delany’s temper was manifested indiscriminately, against men and women both subordinate and superior to him.”’ (J.A. 3233). Yet, Purolator tolerated that behavior from him, and, unlike Sheehan, allowed him to resign. Although not as dramatic, Henoch suffered from a similar pattern of retaliation—a day-to-day harassment in the form of memos, surveillance, insults and maligning her to outsiders. (See, e.g., J.A. 2271, 2304, 2339, 2340-42). Again, that Purola- tor tolerated the outrageous, unprofessional behavior from Delany, while not putting an end to the abuse suffered by Henoch, mandated a finding of unlawful retaliation. 119) ARGUMENT POINT I DENIAL OF CLASS CERTIFICATION WAS AN ABUSE OF DISCRETION The District Court’s decision refusing to certify the class was an abuse of discretion, and in direct contravention of the pre- vailing law in this Circuit. See Rossini v. Ogilvy & Mather, Inc., 798 F.2d 590, 594 (2d Cir. 1986); Society for Good Will to 25a Retarded Children v Cuomo, 737 F.2d 1239, 1243 (2d Cir. 1984). The Court’s failure to rule on plaintiffs’ motion in light of the evidence adduced at trial is also error. See Rossini, 798 F.2d at 594-96. A. The District Court Erroneously Required Affidavits from Class Members When Plaintiffs Presented Substantial Evi- dence of Defendants’ Discriminatory Policies and Statisti- cal Evidence of Discrimination The District Court incorrectly held that, in its view of Falcon, the following substantial evidence was insufficient to show an aggrieved class: 1) a lengthy affidavit describing the experiences of the three named plaintiffs, including their deposition testi- mony; 2) affidavits from two former employees, one, a woman formerly employed, describing pervasive sex dis- crimination, and one from a male officer, no longer employed by the company, describing Purolator’s person- nel policies and practices; 3) hundreds of pages of exhibits from defendants’ own records which constituted direct evidence of Purolator’s discriminatory personnel policies; and 4) unrebutted statistics showing sex segregation of jobs and an unrefuted regression analysis showing statisti- cally signficant male-female salary disparities. {20} Despite the evidence of discrimination presented by plain- tiffs, the District Court nonetheless refused to certify a class primarily because of its view that, since Falcon, plaintiffs are required to provide affidavits from class members showing that they too felt aggrieved. 103 F.R.D. at 648-49. The Court’s clear belief, couched in terms of either the ‘existence of an aggrieved class,’ or commonality or typical- ity, is that there must be a show of hands by the class on whether they support the class action and whether they want the named plaintiffs to be their representatives. 103 F.R.D. at 648; see also J.A. at 1670-71. Such an “‘opt-in’’ procedure is anti- thetical to the class action device and has been rejected. See, e.g., Cox v. American Cast Iron Pipe Co., 784 F.2d 1546, 1556- 26a $7 (llth Cir.), cert. denied, 107 S.Ct. 274, 93 L.Ed.2d 250 (1986); Rule v. Iron Workers, Local 296, 568 F.2d 558 (8th Cir. 1977): Bauman v. U.S. District Court, 557 F.2d 650 (9th Cir. 1977). Rule 23 does not require that class members opt-in. At most, class actions certified under 23(b)(3) require class members to be permitted to opt-out.'' The Manual for Complex Litigation recommends that courts not make submission of a proof of claim by class members a condition to membership in the class because this would be the equivalent of an ‘‘opt-in’’ procedure and inappropriate under Rule 23. MCL 2d (1985) § 30.232. It also provides that courts should avoid class definitions that “depend on . . . the seeking of relief (for example, persons claiming [21] injury or seeking damages from some stated prac- tice),’” because such a definition would constitute ‘‘an improper opt-in procedure.”’ Jd. ¢ 30.14. In Rule v. Iron Workers, Local 296, the district court certi- fied a class action under 23(b)(3) and directed that notice be sent to the class stating that any member who did not request exclusion or who did not return the notice would be in the class. The district court then decertified the class on the ground that the number of persons affirmatively requesting inclusion in the class was so small in relation to the entire class that it failed to meet the numerosity requirement. See Rule, 568 F.2d at 561. The Court of Appeals for the Eighth Circuit reversed, explain- ing that “‘to use as a yardstick the number of class members who affirmatively opted in (17) although not ever requested to do so, was to employ a totally unreliable, if not irrelevant, sta- tistic.”’ Id. at 564. The court noted that a required response would be inconsistent with the opt-out provisions of Kule 23. Id. at 563 n.4. See also Bauman v. United States District Court, 557 F.2d 650, 659 (9th Cir. 1977) (“‘If the district court intends, however, to use non-responses [to notices to class members] to determine that the nunerosity requirement has not been satis- fied, it is abusing its discretion because the probable practical effect of the notice would be to preclude all class actions in 11 =Plaintiffs’ motion was brought pursuant to Rule 23(b,(2), not 2b). a 27a major employment discrimination cases.’’) Similarly, in Cox v. American Cast Iron Pipe Co. , 784 F.2d at 1556-57, the Court of Appeals for the Eleventh Circuit reversed the use of discovery sanctions to dismiss class members [22] on the grounds that it had the same effect as an affirmative opt-in procedure and was contrary to Rule 23. Here, the District Court ignored the purpose of class actions, which are designed to vindicate the rights of persons in the class who need not come forward and be named, nor expend funds to litigate their claims. A requirement of affidavit submissions by class members would be tantamount to an opt-in requirement, which is unquestionably inappropriate in any Rule 23(b)(2) class action, and especially so in employment cases where many of the class members are currently employed by the defendant. The Court below, in applying its view of the need for such affidavits, referred to fifty-six sex discrimination complaints that had been filed by Purolator female employees but dis- counted that evidence. 103 F.R.D. at 649. The Court assumed that most of the complaints were automatically irrelevant on the theory that they had come from non-exempt (hourly paid) female employees. Jd. The Court failed to recognize that these complaints were relevant to the discriminatory atmosphere in which women were forced to function at Purolator and about which they were complaining. Such evidence of biased attitude in the work place is not irrelevant, as the Court apparently believed. The decisions cited by the Court below do not support its erroneous view that, since Falcon, affidavits from class mem- bers who “‘feel aggrieved’’ are essential. In Grant v. Morgan Guaranty Trust Company, 548 F.Supp. 1189, 1193 (S.D.N.Y. 1982), plaintiffs’ counsel conceded during argument on the class [23] certification motion that the class claims rested pri- marily on the allegation that defendant had discriminated against plaintiff individually. Warren v. [TT Worid Communi- cations, Inc., 95 F.R.D. 425 (S.D.N.Y. 1982) was similar. There, the individual plaintiff's EEOC charge was limited to plaintiff's individual complaint that her job description dis- criminatorily omitted some of the tasks for which she was responsible, thereby depressing her compensation. In refusing 28a to certify a class, the court concluded that her class claim rested entirely on her allegation that she had been discriminated against individually.'* 95 F.R.D. at 428-30. The Arkansas opinion, Benson v. Little Rock Hilton Inn, 30 Empl. Prac. Dec. (CCH) 4 33, 188 (E.D. Ark. 1982) is so wholly inapposite as to make puzzling the Court’s citation to it. That court initiated its own inquiry concerning the facts under- lying the class allegations: The Court’s inquiries have been directed to a number of specific areas in which the plaintiffs’ case for class certifi- cation appeared questionable. Despite the ample opportu- nity the Court has afforded the plaintiff to factually support the allegations with respect to the certification requirements enumerated above, the plaintiff has failed to adequately document her claims with affidavits, deposi- tions, interrogatories, or otherwise, and has failed even to properly proffer the supportive facts which would be adduced at an evidentiary hearing, if one were held. [24] Jd. at p. 27,701."° The District Court also referred to pre-Falcon decisions in which some courts had refused to certify Title VII class actions where the plaintiffs had failed to produce affidavits or other evidence to establish the existence of an aggrieved class. How- ever, the cases do not support the Court’s broad statement, and to the extent they support an absolute requirement of affidavits from class members, plaintiffs submit they are wrong and con- trary to controlling law of this Circuit. 12 Nor do the two cases from the Northern District of Georgia support the proposition that affidavits of class members are needed for class certifi- cation. See 103 F.R.D. at 648. Those two cases, decided by the same District Court Judge, invited renewed submissions by plaintiffs on class certification motions as a result of the Court’s view of the impact of Falcon on pending motions. Nowhere in either opinion is it said that such affidavits are required by Falcon. 13 Here, during the argument on the class certification motion, the Court inquired concerning procedures to satisfy it of compliance with Rule 23. Plaintiffs’ counsel informed the Court that the Court could hold a hear ing to receive additional evidence. (J.A. 1671) 29a Wright v. Stone Container Corp., 524 F.2d 1058 (8th Cir. 1975}, simply has no relevance to the case at bar. Nowhere in the decision did the court state that affidavits from the class members were necessary.“ Richardson vy. Restaurant Marketing Associates, Inc., 83 F.R.D 268, 270 (N.D. Cal. 1978) is also inapposite. In that case, the court denied certification on the ground that plaintiffs sought to certify six restaurant facilities owned by [25] defen- dant while the allegations concerned only one restaurant that was no longer managed by the defendant and its previous man- ager, who was no longer employed by the company. Compounding its error in requiring affidavits, the District Court’s class certification decision failed to consider plaintiffs’ substantial evidence of defendants’ discriminatory policies, as well as plaintiffs’ unrebutted statistical showing of sex-based salary differences between men and women. The District Court found fault with the statistics offered by plaintiffs in support of the class certification motion, and stated that the statistics did not, by themselves, establish the existence of an aggrieved class of female employees. Since Teamsters v. United States, 431 U.S. 324, 339 (1977), it has been axiomatic that employment discrimination may be established by statisti- cal proof. See also Rossini, 798 F.2d at 604. Teamsters has been uniformly endorsed by subsequent decisions of the Supreme Court. Falcon did not overrule it or diminish its vitality. To require preliminary proof for class certification of a different 14 In the instant case, the named plaintiffs in their depositions referred to other women who, in their opinion had suffered discriminatory treatment. The Court was provided this information. (J.A. 697-704). The Court also was referred to Sheena’s deposition testimony that she had urged defendants to review their own records to confirm her statements that women were underpaid. (J.A. 698-699) When, during trial, a female employee was called as a witness by plaintiffs to testify as to her own experi- ence which showed a general attitude of discrimination against women by defendants, and specifically by a male who had supervised both Sheehan and Henoch, the Court refused to hear her testimony. (See, infra, pp. 38-39) 30a kind than is required to prove liability to the class is simply unwarranted. [26] Finally, in rejecting plaintiffs’ statistics, the District Court reiterated its view that plaintiffs needed to submit additional affidavits from class members because ‘‘statistics alone [do not] indicate that other female employees feel aggrieved,’’ citing Steffin v. First Charter Financial Corporation, 77 F.R.D. 498 (C.D. Cal. 1978). 103 F.R.D. at 649. As discussed infra, such requirement, standing alone, is error.'® B. The District Court’s Refusal To Certify The Plaintiffs As Class Representatives Was An Abuse Of Discretion The District Court’s reasons for finding that plaintiffs did not meet the commonality or typicality requirements of Rule 23(a) can be summarized as follows: (1) the plaintiffs were too ‘thigh level;’’ (2) the plaintiffs were unique; and (3) the plain- tiffs filed retaliation charges. These reasons are wrong as a mat- ter of law and fact. The Court misconstrued the class claims, placed undue emphasis on plaintiffs’ titles and elevated plain- tiffs to an unduly high status. The Court also imported a new requirement into Rule 23—plaintiffs cannot be different in any respect from other members of the class. Although the District Court purportedly rejected defendant.’ attempt to have it rule that disparate treatment Title VII class actions were barred by Falcon, the District Court’s decision, if 1S The Court, on the class certification motion, had substantial statisti- cal evidence of both sex segregation and gross male/female salary disparities presented by plaintiffs based on a study of a// exempt employees. (See J.A. 727-35). Defendants responded with a study based on an artificial sample, '/2 male and '/2 female (where females constituted less than 30% of all exempts), ‘‘randomly”’ selected employees, which in total were fewer than 8% of all exempt employees in 1982. (See J.A. 1255-60), Defendants’ expert admitted ‘‘that the sample size was not large enough to permit [him] to draw a firm conclusion about all exempt employees.’’ (J.A. 1072). The Court, nonetheless, rejected plaintiffs’ statistics on the ground that all relevant vari- ables had not been considered, a view that is directly at variance with Baze- more, 92 L. Ed. 2d at 331. 16 To the extent that Steffin turns on the absence of class member affi- davits, its holding is erroneous, and certainly not good law in this Circuit or elsewhere. 3la left undisturbed, would have nearly the same effect. In Rossini, this Court rejected a narrow reading of Falcon and reversed the District Court’s decertification of [27] promotion and training claims very similar to those alleged here. 798 F.2d at 596-600. As in Rossini, the plaintiffs here claimed they had to ‘‘ ‘per- form’ at a higher level than men in order to advance their careers.’’ Id. at 599. In Falcon, the Supreme Court did not create a new require- ment that a class representative have individual claims identical to all of the class claims. See Rossini, 798 F.2d at 596-600; Lilly v. Harris Teeter Supermarket, 720 F.2d 326, 333 (4th Cir. 1983), cert. denied, 466 U.S. 951 (1984) (‘‘the [Supreme] Court also rejected those cases. . . . requiring that the class repre- sentatives necessarily have suffered discrimination in precisely the same employment practice as did all the other members of the class’’). Class claims need only be ‘‘fairly encompassed”’ in the named plaintiff’s claims. Falcon, 457 U.S. at 156. While the Supreme Court rejected the use of unsupported inferences as satisfying the requirements of Rule 23, id. at 158, it directed the focus to defendants’ practices, and whether the different prac- tices operated ‘‘in the same general fashion such as through entirely subjective decisionmaking processes.’’ Jd. at 159 n.15. Focus on the practice at issue in the instant case reveals that, as plaintiffs both alleged and demonstrated, those practices operated in the ‘‘same general fashion’’ and Purolator’s policy of sex discrimination affected the class in the same way that it did the individual plaintiffs. Jd. at 159 n. 15. This showing is sufficient for commonality and typicality under Rule 23. See Rossini, 798 F.2d at 598-99; Cox v. American Cast Iron [28] Pipe Co., 784 F.2d at 1557-58 (holding it was an abuse of dis- cretion to decertify class raising compensation, promotion, and training claims); McKenzie v. Sawyer, 684 F.2d 62, 74 (D.C. Cir. 1982) (‘‘similar employment practices, applied to related employment decisions’’); De La Fuente v. Stokely-Van Camp, Inc. , 713 F.2d 225, 232 (7th Cir. 1983) (‘‘similarity of legal the- ory may control even in the face of differences of fact’’). Plain- tiffs, therefore, need not, for example, hold the same job title as other class members. See Rossini, 798 F.2d at 599; Carpenter v. Stephen F. Austin State University, 706 F.2d 608, 616-17 32a (Sth Cir. 1983) (custodial employees can represent class includ- ing clerical workers in case alleging channeling of female work- ers). Rather than following Falcon as it purported to do, the Dis- trict Court erroneously ventured into the merits of the class claims and required virtual identicality rather than typicality of the claims. See Eisen v. Carlisle, 417 U.S. 156, 177 (1974). See also Avagligano v. Sumitomo Shoki America, Inc., 104 F.R.D. 562, 573 (S.D.N.Y. 1984) (rejecting a similar attempt by a defendant to inject a merits of inquiry into the class certifica- tion proceeding in order to narrow the class). 1. High Level. The District Court’s refusal to certify plain- tiffs as class representatives based on their ‘‘high level’’ is espe- cially troubling. Plaintiffs sought to represent a class of exempt employees, not Purolator’s entire female work-force. The Court rejected the plaintiffs as class representatives because over the years they had received [29] ‘‘promotions’’ and salary increases. However, the class promotion claim was that females were promoted more slowly than similarly situated males, and that defendants refused to promote women into line jobs, not that they were never promoted. (See J.A. 33). The named plain- tiffs claimed, in common with the class, that if they had been male, they would have advanced more rapidly and would have reached higher levels; they would have been considered for line rather than staff positions; and they would have been paid higher salaries. (See J.A. 32-33). Plaintiffs claimed that men who started at low levels as the plaintiffs had were promoted far more often and received much higher salaries. (J.A. 33). Just because plaintiffs managed to have been promoted does not mean that their experience was unrelated to that of other women who were at different points in their careers but who, like the plaintiffs, encountered the Same sex-based barriers. The fact that a plaintiff is subject to the same discriminatory practices ‘‘and yet overcame them does not disqualify her from serving as a class agent provided there is ‘significant proof that an employer operated under a policy of discrimination.’ ’’ Hansen v. Webster, 41 FEP Cases 214, 223- 24 (D.D.C. 1986) (quoting Falcon, 457 U.S. at 159, n.15) (emphasis in original). As discussed supra, plaintiffs provided 33a such proof in the form of evidence of discriminatory personnel policies, other direct evidence of discrimination, as well as sta- tistical evidence. In any event, plaintiffs were not nearly as high level as tne District Court suggested. In rejecting defendants’ [30] argu- ment that Henoch’s officer title disqualified her from being a class representative, the District Court claimed to reject an argument that emphasized form over substance. See 103 F.R.D at 651 n.10. Nonetheless, the Court, in effect, let the plaintiffs’ titles determine its decision. The District Court simply ignored other compelling evidence, such as plaintiffs’ salaries and job functions. Sheehan was earning $28,000 per year when her employment was terminated in 1981, hardly the salary of a high level executive. (E.A. 115). Henoch’s salary at the time the class complaint was filed was $34,240. (E.A. 119). At trial, defen- dants’ strategy was to attempt to protray Henoch essentially as a clerical. (See generally J.A. 2398-2410). Purolator’s former Senior Vice President of Human Resources candidly testifed at trial that titles were often given instead of salary increases. (J.A. 2191). Although plaintiffs were among the higher-ranking women at Purolator, the Dis- trict Court in finding plaintiffs to be inadequate representatives ignored the many higher-level men and the virtual, if not total, absence of women at higher levels. Sheehan and Henoch both had titles of Staff Vice President, the ‘‘entry level into the offi- cer hierarchy’’ (J.A. 2192); there were many levels of officers above them, virtually all of which were occupied by males. (J.A. 2191-92; E.A. 105). The effect of ‘he District Court’s rule is that any employee at one end or another of a group could not be an appropriate class representative, a proposition which finds no support in the law. Rowe v. Bailar, 26 FEP Cases 1145 (D.D.C. [31] 1981), upon which the District Court so heavily relied, see 103 F.R.D. at 651, is inconsistent with the law in this Circuit, and its facts also bear no resemblance to those presented here. In Rowe, the plaintiff, unlike Sheehan and Henoch, does not appear to have alleged salary discrimination. In addition, the court there sug- gested that statistics cannot be used as the primary proof of dis- crimination in a disparate treatment case, 26 FEP Cases at 34a 1146, a position emphatically rejected by the Supreme Court in Bazemore, and by this Court in Rossini. The court also found the class allegations to be ‘‘conclusory allegations of discrimi- nation’’ and noted that the defendant had facially neutral per- sonnel policies, id. at 1147; neither factor is found in this case." Finally, the Rowe court noted that the plaintiff had received generally favorable treatment throughout his employment; that question goes to the merits of the plaintiffs’ discrimination claims and should not be resolved at the class certification stage. Ultimately, the District Court’s finding that the plaintiffs were inappropriate representatives because of their high level rests on the Court’s erroneous, unsupported, statement that: The interests of secretaries and customer-service represent- atives may not be co-extensive with the interests of a Staff {32} Vice President or Director. Nor can it be assumed that the same practices and criteria are applied to employment decisions including low level and high level employees. 103 F.R.D. at 652. The District Court’s reference to secretaries and customer service representatives is inexplicable. In early 1982, Purolator reclassified the salaried position of Customer Service Representative as ‘‘salaried, non-exempt’’, although from the creation of the classification in or about 1978 through early 1982, the position was classified as exempt. (J.A. 706 n.10). At the time the job was reclassified, a few secretarial positions were also changed from exempt to non-exempt status. (/d.) The Court’s decision cites to Defendants’ Supplemental Memorandum of Law, at 13, where defendants claimed without citation that ‘‘[w]omen exempt employees at Courier run the gamut from secretaries and customer service representatives to 17 _—_— in this case, the District Court found that plaintiffs failed to specify acts of discrimination with respect to the class claims and equated that alleged failure to the general conclusory class claims in Rowe. 103 F.R.D. at 651 n.11. As discussed supra, plaintiffs provided the Court with more than should be required on a motion for class certification. Moreover, plaintiffs presented additional evidence at trial showing defendants’ discriminatory policies. The Court simply ignored that evidence and plaintiffs’ motion for _ reconsideration of the class certification motion. 35a corporate officers and professionals.’’ This memorandum was submitted in February 1983 a year after Courier had changed the status of these women to non-exempt.'® Finally, noticeably absent from the District Court’s opinion is any reference to the many decisions in which courts have cer- tified classes of exempt or professional employees similar to that sought here. See, e.g., Rossini, 798 F.2d 590; Meyer v. MacMillan Publishing Co., Inc., 95 F.R.D. 411 (S.D.N.Y. 1982). If the Court had legitimate concerns about some women, such as secretaries, being included in the class, it should have [33] certified the class expressly omitting them from the defined class. See, e.g., Lo Re v. Chase Manhattan Corp., 431 F.Supp. 189, 198 (S.D.N.Y. 1977) (certifying a class of all females who ‘fare now employed, have been employed, or have sought employment with the corporate defendants in official, manage- rial or professional positions’’ and excluding clericals, tellers and secretaries from the class.) 2. Uniqueness. The District Court also found Henoch to be an inappropriate representative ‘‘because of the particularly unique circumstances of her employment.’’ 103 F.R.D. at 652. Because Henoch might have to prove additional facts to sup- port her claim, does not mean she is not a proper class repre- sentative. ‘‘Commonality was not destroyed merely because [the plaintiff’s] individual claim also required proof of some facts that differed from those of the class claim.’’ Rossini, 798 F.2d at 599. The effect of the Court’s ruling, if left to stand, would mean that a company like Purolator, which has a myriad of job titles and functions in its exempt work force, many of which could be called ‘‘unique’’ would be immune from class actions. It simply is not required that a representative perform the same type of work as class members. See Rossini, 798 F.2d at 599 (‘‘The distinction that [the class representative] was per- forming on creativity tests while class members were perform- 18 In moving for class certification of a class of exempt employees plaintiffs should not be faulted for defendants’ erroneous inclusion of cer- tain employees in the exempt category, arguably in violation of the federal wage and hour laws. 36a ing on day-to-day work assignments is of little importance in light of the allegation that it was the same standardless, subjec- tive evaluation system that operated in both cases’’). The cases cited by the District Court to support its disqualification [34] of Henoch present extreme examples where the class claims clearly were added as an afterthought to individual litigation over ter- mination of employment. The cases do not involve employees who occupy a “‘special position.’’”” 3. Retaliation. Finally, the Court found the plaintiffs to be inadequate representatives because they made retaliation claims. The Court held that Sheehan’s retaliatory discharge claim made her an inadequate representative of the class, and, similarly, that Hellriegel’s harassment and retaliation claims and Henoch’s retaliation claims made them inadequate repre- sentatives, 103 F.R.D. at 652, 656. The disqualification of plaintiffs by virtue of their retaliation claims has no basis in fact and is contrary to the policies underlying the laws prohibiting retaliation. For a court to hoid that a plaintiff becomes an inadequate class representative by [35] virtue of claiming retali- ation invites retaliation as a means of defeating a class suit. See 19 In Williams v. Boorstin, 451 F.Supp. 1117, 1124 (D.D.C. 1978), rev'd on other grounds, 663 F.2d 109 (D.C. Cir. 1980), cert. denied, 451 U.S. 985 (1981), the plaintiff, although apparently alleging class claims of discrim- ination in recruitment, hiring, promotion, and termination, claimed only his discharge in common with the class. The termination of his employment occurred in the very unusual circumstances of his having masqueraded as a law student and lawyer during his employment although he had attended only one year of law school. See id. at 1119-21. The other two cases cited by the District Court similarly involve discrimination actions arising out of employment termination, without a showing, as here, of discriminatory poli- cies. See Martin v. Easton Pub. Co., 73 F.R.D. 678 (E.D. Pa. 1977); Odom vy. U.S. Homes Corp., 76 F.R.D. 381 (S.D. Tex. 1975). 20 In rejecting Hellriegel as a class representative, the Court cited the lower court decision in Rossini, 80 F.R.D. 131, 135-36, reversed by this Court, to hold that because Hellriegel supervised some members of the class, she could not be a representative for those members. 103 F.R.D. at 655 n.12. That position was rejected in this Court's decision in Rossini, 798 F.2d at 596. 37a Holsey v. Armour & Co., 743 F.2d 199, 217 n.8 (4th Cir. 1984).”" The test of adequacy of representation is whether the pro- posed representative has interests antagonistic to those of the class members. Eisen v. Carlisle & Jacquelin, 391 F.2d 555, 562 (2d Cir. 1968). Any antagonism between the representative and the class must be real rather than academic, Herbst v. Able, 46 F.R.D. 11 (S.D.N.Y. 1969), and must relate to the subject mat- ter in controversy. Sperry Rand Corp. v. Larson, 554 F.2d 868, 874 (8th Cir. 1977); Mersey v. First Republic Corporation, 43 F.R.D. 465, 468 (S.D.N.Y. 1968). Here, there was no showing of a conflict with the proposed class members over the subject matter of the litigation. Defendants made no such showing, nor did they submit any affidavits from class members in opposi- tion to certification of the plaintiffs as class representatives. Accordingly, the Court abused its discretion in refusing to certify a class of exempt employees. [36] POINT Il THE DISTRICT COURT’S FINDINGS ON SHEEHAN’S AND HENOCH’S DISCRIMINATION CLAIMS ARE CLEARLY ERRONEOUS A. The District Court’s Failure To Consider The Discrimina- tory Atmosphere And The Sex-Biased Attitude Of Purola- tor’s Decisionmakers Constitutes Reversible Error A district court’s findings are clearly erroneous ‘‘to the extent they fail to recognize [important] incidents and reject or fail to draw inferences which [are] found [to be] inescapable from the record.”’ Griffin v. City of Omaha, 785 F.2d 620, 628 (8th Cir. 1985) (quoting Alexander v. National Farmers Organization, 21 During oral argument, the Court asked plaintiffs’ counsel whether the plaintiffs would forgo their ‘‘personalized’’ retaliation claims. (J.A. 1672-73). Although believing this to be a totally improper request, counsel stated that if necessary plaintiffs would. (/d.) Nonetheless, the Court per- sisted in holding that the retaliation claims made the plaintiffs unique and, therefore, atypical. 38a 687 F.2d 1173, 1203 (8th Cir. 1982), cert. denied, 461 U.S. 637 (1983)) (bracketed [important] in original). It is also error for a court to ignore, without comment, highly probative evidence of discrimination. ‘‘When direct evidence of discrimination has been introduced, the lower court must, as an initial matter, spe- cifically state whether or not it believes plaintiff's proffered direct evidence."’ Thompkins v. Morris Brown College, 752 F.2d 558, 564 (llth Cir. 1985). If the court rejects such evi- dence, it should state ‘‘why this apparently highly probative evi- dence was discredited. . . [since] [i}n these circumstances some indication of the Court’s reasons for rejecting this evidence must be given in order for us to exercise properly our function of appellate review.’’ Jd. (quoting Lee v. Russell County Board of Education, 684 F.2d 769, 775 (11th Cir. 1982) (bracketed [since] [in] in original.) See also Bell v. Birmingham Linen Ser- vice, 715 F.2d 1552, 1557 (11th Cir. 1983). The District Court committed such error here, where it totally [37] ignored plain- tiffs’ direct evidence of discrimination, offered both at the class certification stage and during trial, and provided no explana- tion in either opinion for the failure to give weight even to uncontraverted, direct evidence. Plaintiffs offered written policies in which Purolator specifi- cally referred to sex as an important element to be considered in personnel decisionmaking (E.A. 1), expressed a preference for promoting males (J.A. 726), and advised supervisors to con- sider the number of children that female employees have in order to determine level of responsibility to give them. (E.A. 1). Significantly, these policies were signed by James Rapp and Robert Ulrich, former President of Purolator, who made deci- sions with respect to plaintiffs Sheehan and Henoch, and whose Signatures on those policies indicated their attitudes with respect to gender as a consideration in Purolator’s decision- making. The testimony at trial was that these policies were dis- tributed company-wide to all managerial personnel and not superceded until December 1983 or January 1984—three years after this litigation was commenced. (J.A. 2101-02). None of this evidence was even commented upon by the District Court. In addition to those written policies the Court also ignored testimony from a former manager of Purolator that he had Wa been told by Rapp, among others, to hire women at lower sala- ries than comparably situated men because women have ‘‘hus- bands.”’ (E.A. 399). The witness also observed that, based on his managerial experience within Purolator, it was more diffi- cult to get women promoted than men and that if you were female ‘‘you weren't going to move ahead with Purolator.”’ (E.A. 381). At trial, the Court precluded a witness, Providence Balzano (‘‘Balzano’’), from testifying that Delany (one of the defen- dants’ major policy makers) told her she could not have a job because men would not respect a woman in that particular posi- tion. (J.A. 2525). The Court’s rationale for precluding this tes- timony was that this was not a class action case.” The Court abused its discretion in precluding this testimony, where it was highly relevant to show Delany’s bias. The Court’s utter failure to acknowledge such highly relevant evidence of discrimination is puzzling, and its decision should be reversed on that ground alone. See Hopkins v. Price Waterhouse, _____ F.2d ___, 44 FEP Cases 825, 833 (D.C. Cir. 1987) (violation of Title VII established by plaintiff's showing that ‘‘Price Waterhouse was impermissibly infected by stereotypical attitudes towards female candidates’’); Wheeler v. City of Columbus, 686 F.2d 1144, 1154 (Sth Cir. 1982) (decision vacated for lower court’s failure to consider and analyze dis- crimination claim in the light of sexist comments and pattern of discriminatory hiring practices, which ‘“‘would constitute proof probative of the existence of intentional discrimination’’); McKenna v. Weinberger, 729 F.2d 789, 790 (D.C. Cir. 1984) (‘‘Where a woman is frustrated in her attempts to work [39] cooperatively by the sexist attitudes and actions of her male co- workers, she is a victim of discrimination.’’) See also Equal Employment Opportunity Commission v. Sage Realty Corp., 87 F.R.D. 365, 368 (S.D.N.Y. 1980) (‘‘[Njo matter what form the sex-biased or sex-stereotyped decision may take, when it operates as an impediment to employment opportunity for 22 It is ironic that the Court prevented Balzano from testifying in view ’ the Court's concern in its class certification decision that plaintiffs had r supplied affidavits from women complaining about Purolator’s discrimina 40a women, Title VII requires its elimination.”’); Jordan v. Wilson, 649 F.Supp. 1038, 1058 n.13 (M.D. Ala. 1986) (general state- ments about women probative evidence of discrimination because they “‘help| ] illuminate the department's general atti- tude towards female officers.’’) B. The District Court Improperly Evaluated The Statistical Evidence At trial and in support of class certification, plaintiffs offered an array of statistical evidence in support of their claims, including a regression analysis which showed that on average Purolator’s exempt male employees were earning between three and five thousand dollars per year more than Purolator’s exempt female employees, and that difference in pay was Statistically significant at 2 standard deviations or greater.’ (See J.A. 727-36; E.A. 106). Plaintiffs also demonstrated that, based upon a comparison between the 1982 census-derived, nationally qualified female work force and the Purolator work force, there was an [40] underrepresentation of females in Purolator’s exempt job cate- gories, and the underrepresentation was statistically significant at two standard deviations or greater. (J.A. 732; E.A. 110). Furthermore, unlike what appears generally to be true in the United States, the salary differential between Purolator’s male and female exempt employees, based on seniority, is increasing: 1982 PERCENT MALE SALARY, ADVANTAGE BY YEAR OF EXEMPT EMPLOYEE’S HIRE AT PUROLATOR Pre- 1966 1966-75 1976-7 1978-79 1980-81 1982 (N/A—only 48.87% 49.1% 62.3% 49.5% 66.1% one female current employee hired pre-1966) 23 Plaintiffs’ regression analysis included variables for sex, age, years of service with Purolator, and job itle. (E.A. 106). A study including these same variables and geographic jocation showed similar results. (J.A 1255-60) ———eooO 4la (E.A. 111). Thus, for example, in 1982, men who had from seven to sixteen years of service with Purolator were earning about 49% more than women with the same seniority; and men who had from three to four years of service with Purolator were earning 62% more than women with equal seniority. This in a company which has an express policy of promoting from within, and in which a high school education ‘‘with experience”’ are sufficient qualifications for most jobs. (J.A. 2860). As to traiuing, plaintiffs showed that during 1981-82, ten men and no women were allowed to participate in Purolator’s Management Training Program and during the period [41] 1975-1980 of thirty-four participants in that same Program, only five were women. As of 1982, a female employed by Puro- lator who had gone through the Management Training Pro- gram was earning about $108.00 less per week than the remaining male who had been appointed to that same program. (E.A. 113). Finally, with respect to fringe benefits, plaintiffs showed that the proportionate difference between the number of women employed in job categories eligible to receive the automobile fringe benefit and men employed in job categories eligible to receive that same benefit was statistically significant at two standard deviations or greater. (E.A. 114). Defendants did not challenge the validity of any of plaintiffs’ studies. Their expert conceded the accuracy of plaintiffs’ regression analysis and testified that the methodology employed for that study was sound. (J.A. 3026-27). Nor did he comment on or challenge plaintiffs’ work force analysis or plaintiffs’ automobile fringe benefit study. Rather, defendants’ expert criticized plaintiffs’ regression analysis on the ground that it excluded variables {primarily edu cation and prior experience) that might make a difference in the results. He testified that, although he had performed an analy- sis of only one hundred of Purolator’s employees, he could draw no conclusion about what effect, if any, the omission of these variables would have on plaintiffs’ regression analysis, which included a// exempt employees: 42a Court: You indicated several minutes ago that two very Significant variables are education and total job experi- ence? [42] Witness: Yes. * * * The Court: What effect would the exclusion of those vari- ables have, upon the validity of this? [Plaintiffs’ regres- sion analysis (Ex. 227) (E.A. 106)}. * + * The Witness: That’s a terrific question. I don’t know. The answer is, it could have a big effect. And unless you have the data you don’t know. (J.A. 3037). With respect to his own regression, defendants’ expert stated “‘the sample size of his study was not large enough to permit me to draw a firm conclusion about all exempt employees.’’ (J.A. 3142). Accordingly, defendants’ expert con- cluded: [I]t [defendants’ study] doesn’t prove there is no signifi- cant difference for the whole group of exempts. [T}hat could only be shown by doing the calculations and doing the collecting the data for all the exempts. (J.A. 3143). Plaintiffs’ studies, however, included ‘‘the whole group of exempts’’ and remained unrebutted. !n Bazemore, the Supreme Court reaffirmed its approval of the use of regression analysis to prove a plaintiffs’ case of dis- crimination, and in so doing, stated: Importantly, it is clear that a regression analysis that includes less than all measurable variables may serve to prove a plaintiff’s case. A plaintiff in a Title VII suit need not prove discrimination with scientific certainty; rather, his or her burden is to prove discrimination by a prepon- derance of the evidence. [43] 92 L.Ed.2d at 331 (citations omitted). The Supreme Court in Bazemore also disapproved the rejection based on specula- tion of uncontradicted statistical evidence. Thus, in finding that 43a the lower court had improperly rejected plaintiffs’ regression studies by crediting the questionable evidence offered by the defendant, the Court explained: There was very little evidence to show that there was in fact no disparity in salaries between blacks and whites, or to demonstrate that any disparities that existed were the product of chance. . . . Respondents’ strategy at trial was to declare simply that many factors go into making up an individual employee’s salary; they made no attempt that we are aware of—statistical or otherwise—to demon- strate that when these factors were properly organized and accounted for there was no significant disparity between the salaries of blacks and whites. Bazemore, 92 L.Ed.2d at 333 n.14. This Court, too, found the rejection of a plaintiff's regres- sion study to be an abuse of discretion when that rejection was based ‘‘on the speculative basis that the table’s results might ‘possibly’ have been different if [the] military experience [vari- able] had been included.’’ Rossini, 798 F.2d at 604, citing Baze- more, for the proposition that regression analysis excluding some variables may still be valid evidence of employment dis- crimination.” See also Eastland v. TVA, 704 F.2d 613, 623 (11th Cir. 1983) (variables in statistical models should be those [44] likely to have actual effect in decision-making process under study), cert. denied, 465 U.S. 1066 (1984)).*° It was error for the District Court to reject plaintiffs’ uncon- tradicted statistical proof on the ground that the regression analysis excluded certain variables, when defendants’ own expert testified that he did know what effect the exclusion of the variable had on the results. As in Bazemore, the defendants here offered no evidence to show that there was ‘‘in fact no dis- 24 This Court found an abuse of discretion, notwithstanding the pres- ence of other statistical studies in which inclusion of a military experience variable had reduced the male-female salary differential. 798 F.2d at 603-04. 25 The inclusion of a variable for education in a study of Purolator’s workforce would have been highly questionable given defendants’ admission that high school was deemed a sufficient educational background for most jobs. (J.A. 2860). 44a parity in salaries between [males] and [females], or to demon- Strate that any disparities that existed were the products of chance.’’ Bazemore, 92 L.Ed.2d at 333 n.14. It offered only a limited study, from which its expert conceded, no conclusions could be drawn. And with respect to plaintiffs’ work-force and automobile benefit analyses, defendants offered no evidence to undermine them and simply never questioned them. The Court’s rejection of these studies without comment is reversible error. POINT Ill THE DISTRICT COURT’S FAILURE TO EXAMINE FULLY PLAINTIFFS’ RETALIATION CLAIMS REQUIRES REVERSAL To establish a prima facie claim of retaliation under Title VII, a plaintiff may show: (a) known statutorily protected opposition; (b) adverse employment action against the plaintiff; and (c) a causal connection between the two, i.e., a retaliatory [45] motive playing a part in the adverse employment action. Davis v. State University of New York, 802 F.2d 638, 642 (2d Cir. 1986). See also DeCintio v. Westchester County Medical Center, F.2d , 44 FEP Cases 33, 36-37 & n.8 (2d Cir. 1987); Dominic v. Consolidated Edison Co.., F.2d , 44 FEP Cases 268, 272 (2d Cir. 1987). Once the plaintiff has established those three elements, a prima facie case of retalia- tion has been made, and the employer is obliged to articulate a legitimate, non-retaliatory reason for the adverse action in order to attempt to refute any inference of retaliation which may be drawn from the plaintiff’s prima facie showing. See Grant v. Bethlehem Steel Corp., 622 F.2d 43, 46 (2d Cir. 1980). If such a justification is offered, plaintiff then is entitled to offer evidence calculated to show that the employer’s articu- lated justification for the employment action was pretextual, i.e., *‘a cover up for retaliation.’’ See Womack v. Munson, 619 F.2d 1292, 1296 (8th Cir. 1980), cert. denied, 450 U.S. 979 4Sa (1981); see, generally, McDonnell Douglas v. Green, 411 U.S. 792, 802-04 (1973). Evidence that an employee who has complained of discrimi- nation is disciplined for an infraction, while the employer does not take similar action against others for ‘‘similar derelic- tions,’’ provides the causal connection between protected activ- ity and retaliation. DeCintio, 44 FEP at 36. See also at id. at 37 n.8. The District Court simply made no effort either to analyze defendants’ [Text truncated at 120,000 characters. The full text is on the page linked above.] --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/brief%3Amicro_IA40385019_3008%3A2. Public record. Not legal advice.