The other law of large numbers.

If anyone here knows which planet Justice Antonin Scalia lives on, please do tell. I’d like to move there.

[…] left 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.

This is from this week’s SCOTUS ruling on Wal-Mart v. Dukes. (Full text here.) Justice Ginsburg’s dissent summarizes the basic issue well enough (see the full text for more details):


Women fill 70 percent of the hourly jobs in the retailer’s stores but make up only “33 percent of management employees.” 222 F. R. D., at 146. “[T]he higher one looks in the organization the lower the percentage of women.” Id., at 155. The plaintiffs’ “largely uncontested descriptive statistics” also show that women working in the company’s stores “are paid less than men in every region” and “that the salary gap widens over time even for men and women hired into the same jobs at the same time.”

The plaintiffs sought class action status on behalf of about 1.5 million women currently or formerly employed at Wal-Mart. The District Court and the 9th Circuit Court granted the class certification, after some haggling over who exactly is included in the class and what’s the relative importance of the monetary claims (backpay) to the rest of the suit (injunctive and declaratory relief). The Supreme Court, in a 5-4 split decision, reversed that ruling, on the grounds that… discriminatory practices are so many and varied that they can’t all be covered by one class suit. Scalia again:


Commonality requires the plaintiff to demonstrate that the class members “have suffered the same injury,” Falcon, supra, at 157. This does not mean merely that they have all suffered a violation of the same provision of law. Title VII, for example, can be violated in many ways – by intentional discrimination, or by hiring and promotion criteria that result in disparate impact, and by the use of these practices on the part of many different superiors in a single company. Quite obviously,the mere claim by employees of the same company that they have suffered a Title VII injury, or even a disparate-impact Title VII injury, gives no cause to believe that all their claims can productively be litigated at once.

There’s more where that came from, essentially along the same lines: the Wal-Mart women don’t have much in common beyond their gender, employment and this lawsuit, to paraphrase a 9th Circuit dissenting judge quoted in the decision. They all had different managers who acted each according to his best judgement; and if maybe a few women were yelled at, and others were paid less than their male colleagues, and others were denied promotion unjustly, that doesn’t really put them in the same category, not at all. On the contrary, each such case is different, and each should be considered separately so that Wal-Mart is not deprived of the opportunity to “raise individual affirmative defenses and demonstrate that its action was lawful”. (Syllabus, page 3.)

The SCOTUS majority is saying in effect that each Wal-Mart employee who feels that she’s been discriminated against should just hire a lawyer and file her own lawsuit. On her own Wal-Mart salary.

Also? Wal-Mart has an official anti-discrimination policy, complete with penalties and everything. Beyond that, decisions on pay and promotion are left (within guidelines) to the discretion of individual managers. “In a company of Wal-Mart’s size and geographical scope, it is unlikely that all managers would exercise their discretion in a common way without some common direction” (Syllabus, page 2), but even if they did, that wouldn’t be Wal-Mart’s responsibility because policies and penalties. The lack of common direction could be disputed: Wal-Mart maintains a consistent corporate culture and regularly transfers managers between stores to maintain uniformity of operations throughout the company (Ginsburg’s dissent, page 5). Besides, if people always followed policies and were fair to each other when left to their own devices, there would be no need for courts or judges, would there? But I digress.

Wal-Mart’s “size and geographical scope” is a recurring theme in the majority opinion, a fundamental reason why the plaintiffs could not pass the commonality test. Statistical evidence isn’t relevant because it doesn’t establish enough commonality, and we know this because – wait for it – different managers will try to come up with different excuses if questioned about gender disparities. Or something. (Opinion of the Court, top of page 17.) Scalia then goes on to complain that the anecdotal evidence does not include enough anecdotes: only 120 affidavits, relating to only 235 out of 3,400 stores, do not “demonstrate that the entire company operate[s] under a general policy of discrimination”. (OotC, bottom of page 18, inner quotation marks deleted). This is followed by more kvetching about how the anecdotes break down between different states.

This is a curious reversal of a routine well familiar to women in math and science. Women who speak of their experiences of bias are told that isolated incidents don’t prove anything and that hard statistical data is called for. We now find out that if such data is actually produced, in sufficient quantity that its statistical significance should be beyond question, that’s when the judges would have preferred gazillions of anecdotes instead.

The rest of the decision doesn’t make it easier for us, either. Individual cases can always be explained by some combination of circumstances. I’ve met people who can do that every time without ever breaking a sweat. But if we try compiling statistics instead, then there’s a problem, namely that universities, faculties and departments across America might actually be even more diverse than Wal-Mart stores. Universities have anti-discrimination policies, too, and each department head has his own reasons for discretionary decisions, and it’s certainly not possible for all of them to conspire against women, is it?

Ultimately, split decisions such as this one are about different judges looking at the same set of facts and seeing vastly different pictures. The five judges who voted with the majority are all male. All three female judges dissented, as did a single male justice who broke gender ranks (Breyer, and kudos to him). I suppose that the judges in the majority could actually believe that if a company has an anti-discrimination policy – with penalties! – then there can be no company-wide discrimination. After all, it has always worked for them. Ginsburg, with her life experience, would likely have to disagree, as would Kagan and Sotomayor. Successful women start out young and naive sometimes, but can’t afford to stay that way for very long.

I could say more about the ruling, but instead I’d like to close by expressing my respect and admiration for Justice Ruth Bader Ginsburg. She’s been the voice of reason in so many recent cases that I’ve read about. Despite her bouts with cancer, her husband’s bouts with cancer, and his ultimate death from it last year, she’s hardly taken any time off and evidently has no plans yet to step down. She must know how hard she will be to replace.

(About the spelling: I’m used to “Walmart”, but the SCOTUS opinion spells it “Wal-Mart” every time, so Wal-Mart it is. Not that I’ll be shopping there anytime soon.)

1 Comment

Filed under feminism

One response to “The other law of large numbers.

  1. Pingback: Weekly Picks « Mathblogging.org — the Blog