Tuesday, June 21, 2011

Always Low Principles, Always

A big decision yesterday by the Supreme Court dealing with gender bias and workplace discrimination. The case is Wal-Mart v. Dukes (2011). From the NYT:

The Supreme Court on Monday threw out an enormous employment discrimination class-action suit against Wal-Mart that had sought billions of dollars on behalf of as many as 1.5 million female workers.

The suit claimed that Wal-Mart’s policies and practices had led to countless discriminatory decisions over pay and promotions.

The court divided 5 to 4 along ideological lines on the basic question in the case — whether the suit satisfied a requirement of the class-action rules that “there are questions of law or fact common to the class” of female employees. The court’s five more conservative justices said no, shutting down the suit and limiting the ability of other plaintiffs to band together in large class actions.

The court was unanimous, however, in saying that the plaintiffs’ lawyers had improperly sued under a part of the class-action rules that was not primarily concerned with monetary claims.

It was basically a split decision of 5-4 and 9-0, and while the issues underlying the case involved gender bias and workplace discrimination, the case before the court was essentially regarding class-action status and whether thousands of female employees (perhaps 1.5 million) could band together to sue.

Sociology was dragged into drama via the work of Dr. William Bielby, who was featured in this NYT portrait back in March.

When the Supreme Court considers on Tuesday whether hundreds of thousands of women can band together in an employment discrimination suit against Wal-Mart, the argument may hinge on the validity of the hotly disputed conclusions of a Chicago sociologist.

Plaintiffs in the class-action suit, who claim that Wal-Mart owes billions of dollars to as many as 1.5 million women who they say were unfairly treated on pay and promotions, enlisted the support of William T. Bielby, an academic specializing in “social framework analysis.”

A central question in the case is whether he should have been allowed, in preliminary proceedings, to go beyond describing general research about gender stereotypes in the workplace to draw specific conclusions about what he called flaws in Wal-Mart’s personnel policies.

Here's a rundown of Bielby's assertions.

Social framework analysis gives courts general information — a framework — drawn from social science. Testimony about the reliability of eyewitness identification can, for instance, serve a valuable role in cases in which prosecutors seek to rely on such evidence.

Professor Bielby, who declined a request for an interview, told the trial court that he had collected general “scientific evidence about gender bias, stereotypes and the structure and dynamics of gender inequality in organizations.” He said he also reviewed extensive litigation materials gathered by the lawyers in the case.

He concluded that two aspects of Wal-Mart’s corporate culture might be to blame for pay and other disparities. One was a centralized personnel policy. The other was allowing subjective decisions by managers in the field. Together, he said, those factors allowed stereotypes to infect personnel choices, making “decisions about compensation and promotion vulnerable to gender bias.”

The methodology he used, Professor Bielby explained, was social framework analysis. He cited the seminal work of the two law professors at the University of Virginia, Professors Walker and John Monahan, in the first of 123 footnotes in his 41-page sworn declaration in the case.

As the article notes, however, the very professors upon whose work he based social framework analysis were skeptical of Bielby's contentions.

“Bielby made a conclusion that he had no basis to make,” said Laurens Walker, one of two University of Virginia professors who coined the term for the analysis almost 25 years ago. “He hasn’t done the research.”

Professors Walker and Monahan contend in their academic writing that Professor Bielby has misused social framework analysis. It is fine, they say, to give courts general information about social science research. But it is improper, they continue, to draw conclusions about the matter in dispute without conducting first-hand research.
Yesterday, the Supreme Court agreed, harshly discounting Bielby's work (from Scalia's majority opinion):
The only evidence of a “general policy of discrimination” respondents produced was the testimony of Dr. William Bielby, their sociological expert. Relying on “social framework” analysis, Bielby testified that Wal-Mart has a “strong corporate culture,” that makes it “‘vulnerable’” to “gender bias.” He could not, however, “determine with any specificity how regularly stereotypes play a meaningful role in employment decisions at Wal-Mart. At his deposition . . . Dr. Bielby conceded that he could not calculate whether 0.5 percent or 95 percent of the employment decisions at Wal-Mart might be determined by stereotyped thinking.”

The parties dispute whether Bielby’s testimony even met the standards for the admission of expert testimony under Federal Rule of Civil Procedure 702 and our Daubert case … The District Court concluded that Daubert did not apply to expert testimony at the certification stage of class-action proceedings. We doubt that is so, but even if properly considered, Bielby’s testimony does nothing to advance respondents’ case. “[W]hether 0.5 percent or 95 percent of the employment decisions at Wal-Mart might be determined by stereotyped thinking” is the essential question on which respondents’ theory of commonality depends. If Bielby admittedly has no answer to that question, we can safely disregard what he has to say.
Ouch. The ASA offered a defense of its support of Bielby and the plaintiffs back in March as well:

Sociological research and data were used to document widespread sexual discrimination across Wal-Mart stores, and the Ninth Circuit Court affirmed this as a valid source of evidence to certify the plaintiffs as a “class” and that the class experienced discrimination. Wal-Mart critiqued the research data as lacking validity citing legal scholars writing in law review journals that are not peer-reviewed.

If the Supreme Court rules in favor of Wal-Mart, the validity of social science research in legal opinions could be significantly diminished.

I'm not sure that's true. From my cursory reading of social framework analysis, the methodology relies on plugging in anecdotal evidence of discrimination into an elaborate regression framework and then coming up with probabilities and possibilities regarding gender bias and discrimination. Of course, probabilities and prediction are the gold standard in social scientific research, and if Bielby's application (which I can't seem to find anywhere) is accurate, then it should carry weight.

But in a court of law, where facts carry more weight than probabilities or anecdotal evidence, I'm not sure social framework analysis was the proper thing to use. Not only was it dismissed outright by the majority, but the dissenters (Ginsburg, Breyer, Sotomayor & Kagan) didn't even mention it.

They did mention Dr. Richard Drogin's work, which was far more damning in its indictment of Wal-Mart's treatment of female workers (and used actual evidence and more traditional methodologies). The majority opinion acknowledged the validity of Drogins' work as well, but felt it was "insufficient to establish that respondents’ theory can be proved on a classwide (sic) basis."

Regardless, Ginsburg's dissent is probably more a sociological tour de force than social framework analysis.

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.”

Wal-Mart’s delegation of discretion over pay and promotions is a policy uniform throughout all stores. The very nature of discretion is that people will exercise it in various ways. A system of delegated discretion, Watson held, is a practice actionable under Title VII when it produces discriminatory outcomes.
I'll leave it to constitutional and legal scholars to explain the finer points of the ruling, but suffice it to say, I don't think this was a rejection of sociology, gender bias and workplace discrimination at Wal-Mart, or the validity of social scientific research.

In fact, the 5-4 split (all men v. all women on the court) pretty much sums it up. Gender filters how we interpret and perceive social facts (and social scientific research): the female justices (plus Breyer) understand gender bias and the insidious nature of workplace discrimination.

The men remain clueless and dismissive, as usual.

UPDATE: Sociological Images has more on the data of the case.


Rachel said...

Bahahaha!! I had to laugh out loud when I read your last sentence!!

Bill Bielby said...

"Social framework analysis" is a term made up and used by legal scholars and attorneys. It has no meaning in sociology. I didn't do a "social framework analysis" (though I referenced the term once in my report). I did an organizational case study, an approach that has a long history in sociology. My report is at: http://walmartclass.com/all_reports.html (I also did a lengthy rebuttal report that is not on the web). Also see the ASA amicus brief at:

Todd Krohn said...

Bill, I appreciate the response and the links. I'm glad you clarified social framework analysis, which sounds sociological, but for the life of me I could not remember ever reading about until the first NYT piece came out in March.

The links also reiterate the conclusion you, Drogin and others have made: Wal-Mart may have dodged the class action bullet with this decision, but they are a long way from being exonerated of workplace discrimination and gender bias.