Mind the (Gender) Gap A Follow-up Experimental Analysis of Gender Differences in Performance in Competitive Environments

Gender differences in the composition of competitive high-profile jobs have persistently contributed to the gender gap in earnings among men and women. Social scientists have suggested two important explanations for workforce gender inequity: discrimination and differences in preferences. In their landmark paper, Uri Gneezy, Muriel Niederle, & Aldo Rustichini (2003) presented experimental evidence for an additional factor: controlling for ability, women may be less effective than men in competitive environments. They observed no gap in mean performance in the non-competitive piece rate and a significant gap in the competitive tournament; the effect was stronger in mixed groups, as women competed against women in single-sex groups. In this paper, I test for the robustness of their study by adopting and rerunning two important details of their experimental design: (1) a comparison of tournament and piece rate performance as a measure of “competitiveness” and (2) the task of mazes. I also introduce a new task for which women excel, matching pictures, to see if their results are generalizable across tasks. Like the previous study, we also find a gap in the competitive tournament; however, we find that women significantly learned or competed against men, although the improvement was not substantial. We find for matching pictures a gap in favor of women in the piece rate but no gap in tournament performance for mixed groups; however, men significantly learned or competed against men. In general, we find that men improve significantly, compete, or learn more than women for both tasks.