Bent on Saving Systemic Disparate Treatment from Wal-Mart

In our second installment on mitigating the effect of Wal-Mart, we have Jason Bent's (Stetson) new article, Saving Systemic Disparate Treatment by Exposing Hidden Priors, which he has just posted on SSRN.  The abstract:

What remains of the systemic disparate treatment theory of discrimination after Wal-Mart Stores, Inc. v. Dukes? Is it now defunct, or can it be saved? This Article contends that the systemic disparate treatment theory can be saved only by identifying the flaws in its statistical foundation, and rebuilding that foundation anew with a recognition of the inescapable role of Bayesian priors. Recent scholarly efforts to understand systemic disparate treatment law can roughly be sorted into two strands – methodological and contextualist. In the methodological strand, scholars call attention to the inability of current statistical methodologies to support an inference of discrimination. In the contextualist strand, scholars argue that systemic disparate treatment theory should be conceptually expanded to impose liability on employers for wrongdoing located at the organizational level, rather than as simply an aggregation of individual-level claims. This Article aims to reconcile these divergent scholarly strands and, in the process, to rebuild systemic disparate treatment law. Taking a Bayesian view, the statistical shortcomings identified in the methodological strand are not fatal. Yet, the Bayesian view also provides conceptual space for the organizational approach advanced in the contextualist strand. After Wal-Mart, scholars have an opportunity to reshape this sorely misunderstood area of antidiscrimination law. To do so, legal scholars, social scientists, and statisticians will need to convince courts to acknowledge, embrace, and ultimately, to manage the vital role of Bayesian priors.

I'm always a sucker for arguments that seek to take the advantage of statistics, while still recognizing that it has limits and can benefit from other types on analysis.  So check it out!.

-JH