April 4, 2011

Wal-Mart’s Class Action Conundrum

Why employment discrimination and class action laws don’t mix

Two of the most controversial features of modern law had their origins in the turbulent days of the 1960s. The Civil Rights Act of 1964 imposed on an employer a general obligation not to discriminate on grounds of sex in making employment decisions. Two years later, a major reform in the Federal Rules of Civil Procedure vastly expanded the potential scope of class actions under Rule 23 by introducing a procedural mechanism that allowed representative plaintiffs to bring class action lawsuits on behalf of other individuals without their cooperation or consent.

At the time of these reforms, the proponents of both legal initiatives offered extensive assurances that the expansion of liability would be carefully curbed to avoid imposing excessive burdens on defendants. It is fair to say that the drafters of those new laws and rules did try to make good on that promise.

Illustration by Barbara Kelley

Yet the argument of such reformers that any new rule is both mild and moderate should always be greeted with a large grain of salt. The long-term impact of any statute or rule does not depend exclusively, even largely, on either its text or its legislative history. It depends far more on two less obvious factors. The first is the way in which administrators and judges construe it going forward—which is always more broadly than either the words or the context suggest. The second is the synergistic interaction between different legal regimes, which magnify the impact of both legal regimes.

Which brings us to Wal-Mart, Inc. v. Dukes where that huge transformation is painfully evident. The case itself was argued this past week in the United States Supreme Court. At issue was whether decisions by two lower federal courts to certify a class action on behalf of all women who work at any of Wal-Mart’s stores should be allowed to go forward against the company. As the Wall Street Journal explains:

In Wal-Mart Inc Stores v. Dukes, the Justices will decide whether thousands of women who have worked for Wal-Mart may be certified as a class in an employment discrimination case against the company. If the case is allowed to proceed, it would lump together disparate cases as evidence of systemic discrimination by the world's largest retailer, with potential liability in the billions.

This lawsuit is no small deal, for as Wal-Mart wrote in its opening Supreme Court brief, "At the time of class certification (in 2004), Wal-Mart’s U.S. retail operations comprised 7 divisions, 41 regions, 400 districts, 3,400 stores, and more than 1,000,000 employees."

In order to bring this mammoth lawsuit, it is necessary to stretch both the law of class actions and the law of sex discrimination far beyond their original contours. Starting with the latter, the original 1964 statutory standard on discrimination contemplated defendants who were held responsible "because" they discriminated on the grounds of sex. Overt bans fall prey to this standard, as do policies that purport to be neutral on their face but which are consciously adopted in order to advance the interests of one sex over the other. In principle, moreover, the original prohibition was sex-blind in that it applied as much to favoritism of women over men as to men over women.

To make a strong case, the Wal-Mart plaintiffs will have to stretch both the law of class actions and the law of sex discrimination far beyond their original contours.

That even-handed application of the statute, however, was gone by the early 1970s when it became clear in a set of Supreme Court cases that the special rules of proof were only made available to members of "protected groups" (of which there is no mention in the 1964 Civil Rights Act). This term came to mean women and minority members only. At the same time, courts and juries began to infer discrimination based on policies that had a disparate impact on women and minority members, unless those policies were justified by some narrow conception of "business necessity," which proved impossible to satisfy in practice. The deviation from the strict sex-blind standard was intensified by key rulings in the late 1970s, the effect of which made it clear that affirmative action programs of various sorts might be justified as means to rectify past discrimination on the one hand or to promote workplace diversity on the other.

These reforms should be viewed in quite a different fashion as a matter of first principle. The willingness to allow voluntary affirmative action programs should be defended as an implicit partial repeal of the original 1964 statute, which kept all employers out of the impossible straight-jacket that would arise if disparate impact arguments were allowed in favor of men and women or whites and minority members. But the easing up of the proof standards has been the source of endless false positives, as clever lawyers and compliant courts find ever more novel ways to gin up far-fetched discrimination cases based on bad anecdotes and shaky statistical inferences, which habitually treat any unexplained variation between men and women as attributable to discrimination.

In Wal-Mart, all parties made constant reference to the 1988 Supreme Court decision in Watson v. Fort Worth Bank & Trust, which held that a mix of statistical and subjective evidence could be used to establish a discrimination claim against an African-American teller who had been passed over for promotion to a supervisory position on four separate occasions over a five year period. The Supreme Court allowed this case to go forward, thereby ushering in a massive investigation into the internal operations of a bank that had 80 employees. All the evidence that was needed in that case was fact-specific, and much of it could be subject to dual interpretations. To give but one example: passing over the same candidate four times could be read as showing racial or gender animus (or both) on the ground that the odds of this happening at random were only 1 in 16. But the better interpretation is exactly the opposite. These were not four random decisions. It is highly likely that if someone is found wanting for a supervisory promotion in one case, the same judgment is likely to be appropriate in the next.

Whatever may be the difficulties of making sense of the personnel records of one person in an individual case, they are multiplied mightily when a single class action purports to bring together hundreds of thousands of women who worked in different jobs in different divisions at different times, as in Wal-Mart. The only certain known feature is that Wal-Mart had a strong antidiscrimination policy on the books, which formed the basis for much of its top-down internal policy. The company is not known for its laid-back management style, so it seems presumptively odd to dismiss this policy as idle window-dressing when it forms the basis for many firm decisions. On the other side of the argument were scattered remarks that this enforced policy was a charade, because some managers actually believed that on average men were more aggressive in seeking promotions than women. For the plaintiffs, it would not make any difference if the statement were in fact true, such that Wal-Mart treated all aggressive candidates the same regardless of sex.

At this juncture, the ever-resourceful plaintiff’s lawyer used some fast-stepping class action law to magnify the effect of a hyper-aggressive employment discrimination policy. No one of course denies that class actions have their use when all the plaintiffs are similarly situated against a given wrong. The question that arises is just how similar do these positions have to be in order for the aggregation of class actions to make any sense. For our purposes, Rule 23 gives two possible alternatives. Rule 23(b)(2), under which this case was brought, provides that class action certification is proper when:

(2) [T]he party opposing the class has acted or refused to act on grounds that apply generally to the class, so that final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole[.]

There is no provision that allows any one individual to opt out of the class because there is no need to let this happen. The typical case is an injunction against a private nuisance that blows soot on the land of 100 neighbors. If the injunction is granted to stop the manufacturing activities, opting out makes no sense because once the nuisance is stopped, the party who opted out enjoys the benefit of clean air with everyone else.

The Supreme Court should redeem its reputation by bouncing this class action out on its ear.

The Rule 23(b)(3) action is a different kettle of fish in that it explicitly allows for opting out. This form of class action allows for individual damage actions arising out of a common nucleus of facts even if there are some differences in the relative position of the parties.

The plaintiffs in Wal-Mart knew that appealing to Rule 23(b)(3) was hopeless because of the huge variations inherent in the nature of the claim. So the plaintiffs sought to stuff this action into Rule 23(b)(2) by seeking two sorts of relief, neither of which qualifies for an ordinary injunction. The first form of relief asks for constant remedial oversight on how Wal-Mart hires, fires, promotes, and demotes, which could utterly undercut the efficiency of the company’s operations. The second was for "automatic" back pay without full compensatory damages (e.g., anguish or lost job opportunities) for all these individuals, come what may.

The ground for that action was this: the common element in all cases was that the central administration of Wal-Mart delegated all these decisions to individual managers, and thus allowed discrimination to fester given that managers used a mix of objective and subjective factors. Justice Ruth Bader Ginsburg, while duly aware of the massive administrative complications, showed some affection for this theory on the ground that unconscious preferences could take over these decisions. "Most people prefer themselves," she observed. The next sentence referred to the decision-maker as "he," which implies that a male supervisor might have such biases but hardly a female supervisor. Oh.

The dangers of Justice Ginsburg’s one-sided (and somewhat stereotypical) thinking should be apparent to anyone who is prepared to ask the larger questions related to firm’s operations that the justices studiously avoided. How does the aggressive version of the sex discrimination laws impact efficient employer practices?

Here are the two key ways in which they collide. First, an efficient firm rarely seeks to centralize individual personnel decisions in the home office, or even at the regional level, because remote actors do not have the information to make these hard choices. Second, an efficient firm never relies exclusively on objective or subjective evidence in employment decisions, because each supplies information that the other does not. By any sound standard, Wal-Mart was handling its personnel management just about right. Yet for its pains, the company gets hauled into court on charges that these practices could, in the overconfident words of Justice Elena Kagan, result in "excessive subjective" judgment, as if she or anyone else knows exactly how to mix the objective and subjective together in one seamless web.

We have come a long way from the darker days of 1964 when sex discrimination was a common practice across the land. But what should be apparent is the one point that the justices ignored: powerful market forces have shown just how costly it is for any firm to discriminate against a group of potential employees that could easily constitute over fifty-percent of its work force. It is also equally bizarre to assume that female managers, of which there is more than a token handful, are keen on perpetuating various stereotypes in the places in which they work.

Reading the transcript of the oral argument creates the profound impression of a Supreme Court that is wholly detached from how labor markets work on the ground. We know that every time Wal-Mart is allowed to open up a new store, prospective employees of both sexes line up around the block. That evidence, not the cock-eyed theories of the plaintiff’s lawyers and expert witnesses, should control this case. The Supreme Court should redeem its reputation by bouncing this class action out on its ear.

Indeed, I do think that the unfortunate episode of this case will have a happy ending. In light of the testimony of Justices Ginsburg, Sotomayor, and Kagan, it is not likely that that Wal-Mart will get the unanimous decision that it deserves. But on this issue, the conservative coalition should hold firm, and could easily pick up Justice Breyer and perhaps one of the three female justices. Whatever one thinks of the abstract merits of class actions, Wal-Mart v. Dukes should be an easy win for the defendant.

Richard A. Epstein, Peter and Kirsten Bedford Senior Fellow at the Hoover Institution, Laurence A. Tisch Professor of Law at New York University, and senior lecturer at the University of Chicago, researches and writes on a broad range of constitutional, economic, historical, and philosophical subjects. He has taught administrative law, antitrust law, communications law, constitutional law, corporate law, criminal law, employment discrimination law, environmental law, food and drug law, health law, labor law, Roman law, real estate development and finance, and individual and corporate taxation. His publications cover an equally broad range of topics. His most recent book, published in 2013, is The Classical Liberal Constitution: The Uncertain Quest for Limited Government (2013). He is a past editor of the Journal of Legal Studies (1981–91) and the Journal of Law and Economics (1991–2001).

Letters to the editor may be sent to definingideas@stanford.edu. Editors reserve the right to reject or publish (and edit) letters.