Wal-Mart v. Dukes and the future of meritocracy

First, a big thank you to Jack for inviting me to contribute to the blog.

The fulcrum of the 5-4 divide in Wal-Mart v. Dukes, which came down this morning, turned out to be the following question: was the plaintiffs’ case bolstered, or was it eviscerated, by the fact that Wal-Mart gave mid-level managers nearly unfettered discretion over whom to tap for promotions and what wages (within a range) to pay each employee?  The majority today answered: it was eviscerated.  This is a big change.  In this post I’ll offer a couple of observations about where this change came from and what it might mean going forward.

First it is worth taking a step back.  In 47 years, what has Title VII accomplished?  In my view, it has worked not one, but two, major changes to the landscape of American employment.  One change is the obvious one: Title VII brought some measure of racial and gender integration to the workforce.  But a second, subtler change was in some ways equally profound.  Title VII helped spur both public and private employers to replace subjective, non-meritocratic systems of hiring and promotion with systems that were fairer, more uniform, and more merit-based.  Today it is hard to imagine, for example, a big-city police department deciding simply to hire only the sons and nephews of incumbent officers.  But at one time such practices were common.  What changed was that corporate leaders, HR professionals, public sector managers, and the many, many lawyers who advise all of these people learned that if one does not want to be accused of discrimination—and for that matter, if one does not want to discriminate—it helps, a lot, to adopt relatively rational, formalized, uniform, merit-based policies for hiring, performance evaluation, promotion, compensation, and termination.  An employer that instead adopts policies of total subjectivity and discretion is needlessly painting a litigation target on its back, so why do it?  Or so any competent employment lawyer would have advised, at least until this morning.

Is it still good advice?  Probably.  But Dukes raises the disquieting possibility that rather than inviting litigation, subjective and standardless policies might—for a large enough employer—have the perverse effect of insulating the employer from large-scale litigation by helping to defeat class certification.  This possibility could have far-reaching effects on the landscape of American employment practices.  We would only expect those far-reaching effects under certain conditions.  Large employers would only have reason to switch (back) to more subjective, Wal-Mart-style employment practices if they valued a potential shield against class certification more highly than they valued either (a) the shield against individual discrimination lawsuits that more objective, meritocratic practices provide and (b) all the other salutary effects of more objective, meritocratic practices, which include having a workforce that is more qualified and more rationally compensated and promoted.  (More after the jump.)

To see how Dukes marked a change in the Court’s approach to subjective discretion, one need only compare Dukes to (almost any of) the canonical cases of systemic disparate treatment discrimination.  The plaintiffs’ case in Dukes consisted primarily of statistical evidence.  At Wal-Mart, the plaintiffs alleged, women constituted the vast majority of all employees (about two thirds), but the further you went up the chain of command, the fewer women you saw.  The plaintiffs alleged that 90% of the Store Managers were men; 96% of the District Managers were men; and of the top corporate executives, 19 out of 20 were men.  These were fairly extreme numbers, especially in comparison to Wal-Mart’s competitors.   The plaintiffs’ expert also crunched some numbers on pay and found substantial disparities between male and female Wal-Mart employees, after controlling for various variables, in every region of the country.

Okay.  But what is a court to do with that?  Even if a court credits such evidence, how does one tell whether it represents a company-wide pattern or practice of discrimination?  In the canonical cases of systemic disparate treatment discrimination, such as Teamsters v. United States, courts reasoned that large company-wide statistical disparities had to come from somewhere.  Plaintiffs and defendants offered competing explanations.  (Plaintiffs offered anecdotes of specific instances of discrimination; defendants offered arguments about qualifications, lack of interest, etc.)  If a court believed the plaintiffs, it concluded that the company—despite its protestations to the contrary, and despite the many independent decision-makers in charge of hiring decisions at its many sites—showed a company-wide pattern of race discrimination.

To put it simply, the Court in the 1970s and 1980s was willing to believe that pervasive bias throughout a company and its culture could explain extreme disparities based on race.  In 2011, a five-justice majority of the Court is not willing to believe that anything similar could explain observed disparities based on sex.  In a remarkable passage, Justice Scalia first acknowledges that the Court has, in the past, found employers liable for “giving discretion to lower-level supervisors,” then writes: “[L]eft to their own devices most managers in any corporation—and surely most managers in a corporation that forbids sex discrimination—would select sex-neutral, performance-based criteria for hiring and promotion that produce no actionable disparity at all.”  Slip op. 15.  Scalia found it so obviously true as to require no support that “left to their own devices most managers in any corporation” would not discriminate.  They would not engage in disparate treatment; they would not choose practices with a disparate impact; they would not exhibit cognitive bias.  Discrimination is something bad apples do.

In light of this factual assertion, it is unsurprising that Scalia was so utterly unmoved by the plaintiffs’ two strategies for persuading the trial court that the disparities represented some common, overall pattern or practice of discrimination.  The plaintiffs had anecdotes; but those were presumably the bad apples.  The plaintiffs offered a sociological expert, whose social framework evidence was an attempt to show how and why delegating all hiring and promotion decisions to a nearly all-male group of managers, whose leaders commonly referred to female employees as “little Janie Qs,” might not be the most unbiased way to make decisions.  But how could any of that be convincing, if it is correct—so correct that one can take judicial notice of the fact—that most managers do not discriminate?

There are many things to be said about Dukes, including its effects on class certification, systemic discrimination litigation, backpay remedies, and the disparate treatment/disparate impact distinction. But it seems to me that the two most important questions the case raises are not so doctrinal. They are, first, what does this opinion mean for courts’ power to address forms of discrimination that are widespread and diffuse throughout an enterprise, rather than turning on the bias of a particular decision-maker?  And second, as discussed above, what effect will this opinion have on the conversations among executives, HR departments, and the lawyers who advise them, about the use of subjective employment practices?