Thursday, August 17, 2017

White v. Office of the Cook County Public Defender: Can you establish disparate impact without an expert?

In White v. Office of the Cook County Public Defender, No. 14-cv-7215 (N.D. Ill. Aug. 14, 2017), Federal District Court Judge John Blakely rejected the disparate impact claim of Patrick White on the grounds that he failed to establish a prima facie case of disparate impact. Given the disparities between the rates at which men and women were promoted and the number of individuals involved, I think White should have been able to proceed with his claim.

White alleged that the process for promotion to Grade IV Assistant Public Defender, which was based on interview scores assigned by a board composed of three women, had a disparate impact on men. There were 15 positions available for immediate promotion, and the board interviewed 18 minimally qualified male applicants and 18 minimally qualified female applicants. The 15 positions were awarded to 11 women and 4 men. The board identified 4 additional applicants -- two men and two women -- for future promotions when more slots became available.

No matter how you slice it, these figures seem sufficient to establish a prima facie disparate impact claim. Judge Blakely rejected White's claim based on the "exceedingly small sample size." The sample size, however, was not that small, and the disparity in the promotions rates between men and women was large. Here, the board selected 13 out of 18 women for immediate or future promotion, but it selected only 6 out of 18 men. The promotion rate for men (.33) was less than half that for women (.72). It is true that such a disparity in promotion rates might not have been sufficiently probative if there had been significantly fewer applicants or selectees. Similarly, if the two-to-one disparity had been less extreme, then, absent a larger sample size, the plaintiff's evidence might not have established a prima facie case of disparate impact.

Although Judge Blakely faults White for failing to submit expert evidence, he also notes that there is no formal requirement that he do so. A simple Google search reveals multiple sources that provide tests for evaluating whether a disparity is statistically significant, meaning whether it is unlikely to arise by chance. A particularly easy test to apply is the chi-squared test (see, for example, here and here), which is endorsed by the EEOC's Compliance Manual Section on Compensation Discrimination. If the test reveals the likelihood of a disparity arising by chance to be less than 5%, the EEOC and most courts consider that sufficient to establish a prima facie case of disparate impact.

In this case, you could compare only the men and women selected for immediate promotion or you could compare the men and women selected for either immediate or future promotion. The first comparison reveals that the probability of the disparity is only .018, and the second comparison reveals a probability of only .019, both of which are well below the threshold of .05. Given the absence of any formal requirement that a plaintiff present expert evidence in order to establish a prima facie case of disparate impact and the simplicity of evaluating the statistical evidence in this case, I would think this one of those cases where expert evidence is not needed.

On the other hand, White may have been required to do more than merely present evidence of statistical disparities. Even if he was not required to submit expert evidence, he may have been required to have at least explained the application of the chi-squared test or some other statistical tool for establishing a prima facie case based on his statistical evidence. It is not apparent that White did so. What we're left with then is the rejection of a disparate impact claim that probably should not have been rejected -- at least on the grounds that there was no prima facie case -- but it's not clear who's to blame.

This blog reflects the views solely of its author. It is not intended, and should not be regarded, as legal advice on how to analyze any particular set of facts.