Ninth Circuit Approves Largest Title VII Sex Discrimination Class Action in History

April 30, 2010

In 2007 we reported on the Ninth Circuit’s 2-1 affirmance of the certification of the largest Title VII gender discrimination class action in history. Following that ruling, Wal-Mart asked an en banc (11-judge) panel of the Ninth Circuit to reconsider the matter. On April 26, 2010, the larger panel ruled 6-5 to affirm the certification ruling, with some notable changes.


The Dukes action was brought in 2001 by six current or former female employees of Wal-Mart, then (if not still) the largest private employer in the world, for alleged sex discrimination under Title VII of the 1964 Civil Rights Act. Plaintiffs claimed that women were paid less than men in comparable positions despite having higher performance ratings and greater seniority, and that they received fewer, and had to wait longer for, promotions to in-store management positions than men. They sought injunctive and declaratory relief, lost pay, and punitive damages. The six plaintiffs sought to represent a class estimated in 2001 to be over 1.5 million female employees and former employees who worked in Wal-Mart’s 3,400 stores in 41 regions in 50 states.

In 2004, the district court certified the class with respect to each of plaintiffs’ claims, including their claims for back pay and punitive damages. In 2007, the original three-judge panel of the Ninth Circuit, over a vigorous dissent by Judge Andrew Kleinfeld, affirmed the district court’s certification ruling.

The En Banc Majority’s Holding

In its 236-page opinion released April 26, 2010, the en banc panel affirmed the certification ruling, but with some limitations. First, it limited the class to persons who were employed by Wal-Mart as of the date the original complaint was filed (June 8, 2001). This had the effect of reducing the class size from 1.5 million to approximately 500,000. Second, it ordered the district court to reconsider whether the punitive damage claim should be included in that certification, suggesting that that claim might be more appropriately certified under a different provision of Rule 23. Third, it suggested that the district court consider whether a separate class should be certified covering Wal-Mart’s former employees (those not still employed on June 8, 2001). If the district court were to do that, the combined class size would presumably again approximate 1.5 million people, if not more.

The Lessons of Dukes v. Wal-Mart

I - Many of the lessons of the case are technical, which only a lawyer could love, but one fundamental difference between the approach taken by the 6-judge majority versus the 5-judge group of dissenters — and a difference that matters a lot to employers defending these cases — is, how much evidence or proof should be required of plaintiffs to show that a nationwide class really exists?

 Plaintiffs’ theory for class certification was that Wal-Mart had a corporate policy of discrimination which it implemented through the discretionary decisions of individual store managers, resulting in classwide injury in violation of Title VII. Both majority and dissenters agreed that plaintiffs’ burden was to show that the class was sufficiently numerous to warrant class treatment (there was no dispute about that); that there were issues of law or fact “common” to all class members; that the six named plaintiffs’ claims were “typical” of other class members; and that plaintiffs and their counsel were “adequate” representatives of the purported class.

Where the majority and dissenters diverged is the amount of “evidence” or “proof” that should be required of plaintiffs to meet those requirements. To support their claim of company-wide discrimination, plaintiffs presented (a) anecdotal evidence in the form of declarations from 120 female employees, each attesting to alleged discrimination suffered at the hands of individual store managers, (b) statistical evidence purporting to show, on a regional basis, disparities in pay and rates of promotion between female and male employees, and (c) the expert testimony of a sociologist and labor economist who testified, respectively, as to Wal-Mart’s alleged “strong corporate culture” that was allegedly likely to have led to discriminatory personnel decisions, and a study that purported to show that Wal-Mart promoted women at a lower rate than other retailers. The majority agreed with the district court that this evidence satisfied plaintiffs’ burden of showing a nationwide, corporate policy and practice of gender discrimination.

The dissenters, on the other hand, expressed skepticism that six plaintiffs, who worked in only 13 of Wal-Mart’s 3,400 stores, and who claimed to have suffered discrimination at the hands of only a “few” individual store managers, could appropriately represent a class of 1.5 million women working in 41 regions in 50 states in 3,400 stores. The dissenters, citing the Supreme Court’s decision in Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 159 (1982), concluded that absent “significant proof” of a general, company-wide policy of discrimination, which they found to be missing in plaintiffs’ evidence, “there is nothing to bind these purported 1.5 million claims together in a single action” and therefore class treatment was not appropriate. It is implicit in the dissenters’ view, that absent such “significant proof,” it does not comply with Rule 23 to put the defendant to the time and expense of having to defend a class action, especially a very large class action like this one.

The majority disagreed that Falcon’s “significant proof” standard applied, and instead opted for a standard somewhere between “mere allegations” and “significant proof.” Since, in a class certification motion, the burden of showing compliance with Rule 23’s requirements is on the plaintiff, this fundamental difference in approach between the majority and the dissenters may encourage Supreme Court review.

II - Another fundamental difference between the approach of the majority versus the dissenters — also of extreme importance to the defense of Title VII and other class actions — is, do defendants have the due process or statutory right to present individual evidence to rebut individual class members’ claims?

 The majority holds that in a certified class action, both back pay liability, and, most notably, punitive damages, can be determined on a classwide basis, without individual proof. What this means is that if, in an employment case, classwide discrimination were proven (i.e., plaintiff proves that “discrimination was the company’s standard operating procedure, the regular rather than the unusual practice,” see Int’l Bhd. of Teamsters v. United States, 431 U.S. 324, 336 (1977)), back pay, and possibly punitive damages, may be awarded to each class member without giving the employer the opportunity to prove, with respect to any given class member, that the class member was not in fact discriminated against, or that the same employment decision would have been made regarding the class member even absent the discriminatory practice, or, for purposes of imposing punitive damages, that plaintiff has not proven that the employer acted with “malice or with reckless indifference to the federally protected rights” of the class member, see 42 U.S.C. §1981a(b)(1). The majority’s ruling implicitly acknowledges that some class members, who were not the victims of discrimination, can nevertheless receive compensation, possibly even including punitive damages. The majority’s view is that the risk of “over-compensating” these non-victims of discrimination should be borne by the employer against whom the discriminatory “pattern or practice” was proven.

The dissenters, on the other hand, relying on the Supreme Court’s decision in Teamsters, posit that an employer’s Title VII statutory rights — including the right to establish by affirmative defense either (a) that the adverse employment action taken against the particular class member was taken “for a[ ] reason other than discrimination,” 42 U.S.C. §2000e-5(g)(2)(A), or (b) that, even if motivated in part by discriminatory motive, the employer “would have taken the same action in the absence of the impermissible motivating factor,” id., §2000e-5(g)(2)(B)(ii) — cannot be abridged by determining relief, including compensatory damages, on a classwide basis without individual proof applicable to each class member. The dissenters would similarly insist that before punitive damages could be imposed with respect to any given class member, a showing would have to be made that the employer acted with malice or reckless indifference with respect to that class member. In this case, what that would mean is that 500,000 or more “mini-trials” would be necessary, which, according to the dissenters, would render class treatment inappropriate.

III - Finally, a word about the difference between the majority and the dissenters regarding the consideration of plaintiffs’ proffered expert testimony. Because they adopted a “significant proof” approach to measuring plaintiffs’ burden, the dissenters, but not the majority, would require the district court to delve into the details of the experts’ opinions, including those of defendant’s experts, to see whether those opinions, when taken with plaintiffs’ other evidence, provide “significant proof” of a company-wide pattern or practice of discrimination.

For example, plaintiffs’ statistician analyzed Wal-Mart’s pay and promotion data on a regional basis. Wal-Mart and its statistical expert asserted that the data should be analyzed at the store rather than the regional level, since plaintiffs’ theory was that it was at the store level that the overly “subjective” pay and promotion decisions alleged to be discriminatory were being made. The district court rejected this argument, rejected Wal-Mart’s statistical evidence showing that on a store-level basis pay and promotion disparities between male and female employees were isolated to a small number of stores, and accepted plaintiffs’ statistical evidence as having established an overall pattern or practice of discrimination at all of Wal-Mart’s 3,400 stores. The en banc majority found no abuse of discretion in this approach.

The dissenters, on the other hand, found the district court’s approach “superficial” and constituting “legal error.” The dissenters concluded that in failing to consider Wal-Mart’s countervailing expert’s store-level analysis, which reached a statistical conclusion opposite of plaintiff’s expert’s analysis (and which, according to the dissenters, was arguably more accurate in describing whether or not discrimination had occurred at the store level), the district court had abdicated its responsibility to “vigorously analyze” whether an overall, nationwide pattern or practice of discriminatory treatment — that “would bind the[] purported 1.5 million claims together in a single action — had been proven. Absent plaintiffs’ presentation of such proof, according to the dissenters, certification was not appropriate.

Another major difference between majority and dissent was with regard to the treatment of plaintiffs’ expert sociologist, who plaintiffs offered to employ a “social framework analysis to examine the distinctive features of Wal-Mart’s policies and practices and evaluate[] them ‘against what social science shows to be factors that create and sustain bias and those that minimize bias.’” The expert’s conclusion, on which the district court relied in certifying the class, was that: “Wal-Mart’s centralized coordination, reinforced by a strong organizational culture, sustains uniformity in personnel policy and practice; there are significant deficiencies in Wal-Mart’s equal employment policies and practices; and Wal-Mart’s personnel policies and practices make pay and promotion decisions vulnerable to gender bias.” The majority of the en banc panel found the district court’s reliance on this opinion permissible.

The dissenters strongly disagreed. Relying on the Supreme Court’s decision in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), the dissenters said the district court’s reliance on such an opinion without first testing its scientific validity and reliability under the Daubert standards was legal error. Whereas the majority was unwilling to hold Daubert applicable in the context of a district court’s decision to grant or deny class certification, the dissent posited that it should apply; that its “gatekeeping” function, to establish that expert testimony is “relevant and reliable” before relying on it in making any substantive ruling, should be equally applicable in the class certification context as in any other. The dissenters took note of the “dispute among social psychologists over whether it is appropriate for experts like [plaintiffs’] to use social framework analysis to draw specific factual conclusions about a company’s operations without the use of case-specific research,” and faulted the district court for failing to determine the “scientific reliability” of the testimony before relying on it in certifying a huge, nationwide class.

Based on media reports, and because of the substantial differences between the reasoning of the majority and the dissent with respect to the interpretation and application of at least three Supreme Court cases —TeamstersFalcon and Daubert, not to mention the sheer size of the case (and its potential monetary exposure, which could run into the hundreds of millions if not billions of dollars), it can be expected that Wal-Mart will challenge this ruling by a petition for review by the United States Supreme Court.

For more information on this alert or any other labor and employment issues, please contact any of the lawyers listed below:

John Adkins,, 617.951.8551
Jenny Cooper,, 617.951.8473
Louis Rodriques, Co-chair, Labor and Employment Group,, 617.951.8340

Los Angeles/Orange County
Jacqueline Aguilera,, 213.229.8439
Debra Fischer,, 213.680.6418

New York
Douglas Schwarz,, 212.705.7437

Mie Fujimoto,, 81.3.6721.3138

This article was originally published by Bingham McCutchen LLP.