Untangling the Causal Effects of Sex on Judging
Published in the American Journal of Political Science 54: 389-411
Paper presented at the annual meeting of the Midwest Political Science Association, Chicago IL and at faculty workshops at Stony Brook University, the University of Chicago, Harvard University, and William & Mary University.
Winner of the 2008 Pi Sigma Alpha Award for the Best Paper Presented at the 65th Annual Meeting of the MPSA
Christina L. Boyd
Lee Epstein
Andrew D. Martin
We enter the debate over the role of sex in judging by addressing the two predominant empirical questions it raises: whether male and female judges decide cases distinctly ("individual effects") and whether the presence of a female judge on a panel causes her male colleagues to behave differently ("panel effects"). We do not, however, rely exclusively on the predominant statistical models---variants of standard regression analysis---to address them. Because these tools alone are ill-suited to the task at hand, we deploy a more appropriate methodology---non-parametric matching---which follows from a formal framework for causal inference.
Applying matching methods to sex discrimination suits resolved in the federal circuits between 1995 and 2002 yields two clear results. First, we observe substantial individual effects: The probability of a judge deciding in favor of the party alleging discrimination decreases by about 10 percentage points when the judge is a male. Likewise, we find that when a woman serves on a panel with men, the men are significantly more likely to rule in favor of the rights litigant. Both effects are is so persistent and consistent that they may come as a surprise even to those scholars who have long posited the existence of gendered judging.
Click here for the article (.pdf) (posted on April 27, 2010)
Click here for the web appendix (posted on October 22, 2009)
Click here for the data (Stata .dta file) (posted on March 17, 2008)
Click here for a related paper on gendered judging in state courts of last resort