Gender Bias in Evaluation Processes

Women's representation at the top is still an issue on the agenda for equality. In this paper, we contribute to the literature on the effect of female application reviewers on the performance of women in selection processes. We define gender bias as any differential treatment of female and male applicants. We are particularly interested in whether the gender of the evaluator plays a role in this bias. Using a unique dataset from Mexico, we exploit the fact that applicants cannot select their evaluators and vice versa, that the gender composition of the reviewers chosen is unknown to individual evaluators, and that information is available on the scoring of different aspects of the application. Even though the pairing of applicants to evaluators is not entirely random, we show that the results are robust to specifications and sample restrictions where random assignment is much more likely. The key results are as follows. First, female evaluators assign male applicants a score close to 0.10 standard deviations higher than male evaluators. Second, male reviewers score male and female applicants similarly. Third, female reviewers grade female applicants more harshly than male applicants (by 0.04-0.05 standard deviations), although the estimate is imprecise. We explore three possible mechanisms behind these results and find evidence consistent with gender stereotyping.