From the Magistrate’s Report and Recommendations in Tsaganea v. City Univ. of New York (S.D.N.Y. Jan. 27, 2010), adopted in relevant part by the district court a couple of weeks ago:
Finally, the plaintiff relies on purported statistical evidence to support his theory that the selection process was rigged to hire a Jewish candidate. To be sure, a plaintiff may use employment statistics to show a pattern of discrimination. However, Dr. Tsaganea’s statistics do not support his theory that CUNY conspired to hire only Jewish candidates. He cites data from 2006 from the American Jewish Committee to claim that “the probability of hiring a Jewish person under non-biased condition[s] is … 1.78%.” That figure, however, merely represents the percentage of Jewish persons in the population of the United States. Therefore, the analysis is incomplete, as it does not take into account the number of Jewish persons on one hand and Eastern Orthodox or Romanian persons on the other who hold Ph.D.s from qualified universities. Simplistic data showing the percentage population of an ethnic group does not paint a picture of disparate treatment when qualifications for hiring are not equally shared among the population at large. Also, because of the miniscule sample size, the statistics proffered here likely have no significance even if the analysis had controlled for qualifications.
Exactly right, and bears repeating: “Simplistic data showing the percentage population of an ethnic group does not paint a picture of disparate treatment when qualifications for hiring are not equally shared among the population at large.” The contrary position — that underrepresentation relative to population is itself proof of discrimination, or that groups are entitled to proportional representation regardless of proof of discrimination — would require caps on the number of Jews, Asians, and various other groups that for various reasons have ended up having higher qualifications in certain fields (or just more interest in those fields) than other groups.