Wednesday, May 4, 2011
Selmi on Theorizing Systemic Disparate Treatment Law
Michael Selmi (George Washington) has posted on SSRN his forthcoming piece in the Berkeley Journal of Employment and Labor Law: Theorizing Systemic Disparate Treatment Law.
Here is the abstract:
The pattern or practice cause of action is the most potent, but least understood, of the causes of action recognized by Title VII. The massive sex discrimination case filed against Wal-Mart has renewed scrutiny on the nature of the pattern or practice claim, and this essay seeks to explain under what circumstances statistics can prove intentional discrimination. This essay first explores the history of the pattern or practice claim, noting that the primary case law is now three decades old and was developed around issues of overt race discrimination. Claims of gender discrimination are more complicated because the regression analyses that are at the core of the case do not create as strong an inference of discrimination as occurs in the context of race discrimination. The essay, however, rejects the notion that the pattern or practice claim is merely an aggregation of individual claims, and instead suggests that the statistics prove a more subtle form of discrimination that would not be evident if one focused solely on individual claims. The paper also critiques the presentation of the sex discrimination claim in the Wal-Mart litigation for its generic quality, and suggests that plaintiffs must provide a narrative that explains the statistical story with a focus on the particular employer.
Mike points out that this piece is part of Tristan Green's (San Francisco) project on systemic discrimination. It looks like an interesting idea and I look forward to reviewing it. Check it out!
PS
https://lawprofessors.typepad.com/laborprof_blog/2011/05/selmi-on-theorizing-systemic-disparate-treatment-law.html
Comments
Yes, I think it may support drawing the inference of intent to discriminate. What bothers me about the case is that the proposed Phase II does not give Wal-Mart the opportunity to rebut that inference as to individual class members before awarding individual relief. That is a departure from Teamsters.
Posted by: Jason Bent | May 6, 2011 8:37:00 PM
I am less certain that would be a basis to draw an inerence of discrimination, at least under existing law. Employment decisions seem quite different from the objective measures of a freezer (I suspect nobody is actaully watching the freezer temps at each store anyway). But in terms of personnel decisions, there is no requirement that an employer satisify some numerical standard, which is what the home office would be working from. Maybe if the numbers were way out of whack, and it was clear that the central office knew about the numbers, one might be able to draw some inference, but in general, I think existing law requires significantly more than knowledge of disparities from afar.
Posted by: Mike Selmi | May 9, 2011 7:55:43 AM
Mike, I think you might be conflating "intent" with something closer to "animus." The juxtaposition of hyper control from the central administration over the whole enterprise with the grant of unchecked discretion to local management where the consequences of the exercise of that discretion is known certainly supports the inference of intent.
Posted by: Mike Zimmer | May 11, 2011 10:48:48 AM
Dukes v. Wal-Mart is a mammoth case from a number of points of view. Certainly the "stereotyping" explanation for the shortfall of women in terms of pay and promotions is part of it. But, as I understand the case, the plaintiffs have shown that, in general, Wal-Mart is known for its extensive knowledge and control of most every detail in its far flung enterprise. Apparently, when a freezer turns on in a facility in Shanghai, the Bentonville headquarters computers "know" it and analyze power use, etc. If that freezer's use of electricity is too far out of the norm for freezers, orders go out to Shanghai to do something about it. Like the operation of its freezers, the Bentonville headquarters knows every hiring and promotion decision. But, here is the difference, Wal-Mart doesn't analyze that data the way it does electric power usage and it, therefore, doesn't tell local management to do anything, even investigate, about the fact that the hiring and promotion decisions are out of whack from what should be expected based on probability statistics. If Wal-Mart can't explain this quite different reaction to different data points -- power usage for freezers, shortfall of women in terms of pay and promotions, doesn't that support drawing the inference of intent to discrimination?
Posted by: Mike Zimmer | May 5, 2011 8:37:40 PM