• The editor in charge of this paper was Stefano DellaVigna

  • Acknowledgments: I am grateful to the editor, Stefano DellaVigna, and anonymous referees for their detailed feedback. I would like to thank George-Marios Angeletos, Kristin Butcher, Rachael Croson, Ernst Fehr, Robert Gazzale, Uri Gneezy, Robin McKnight, Gauri Kartini Shastry, Lise Vesterlund, Akila Weerapana, and the participants of the 2008 International ESA Conference, the 24th Congress of the EEA, and the Wellesley College working seminar series for insightful comments. I acknowledge the generous financial support provided by the Shultz Fund (MIT Economics Department) and the Wellesley College Faculty Award. I would also like to thank the Harvard Business School Computer Lab for Experimental Research for providing me with the laboratory space and the access to the subject pool for this study. Finally, I thank the FEEM Award Selection Committee for recognizing this research at the EEA annual congress in 2009. All remaining errors are my own.

E-mail: oshurchk@wellesley.edu


Gender gaps in the workplace are widespread. One explanation for gender inequality stems from the effects of the interaction between competition and two pressure sources, namely, task stereotypes and time constraints. This study uses a laboratory experiment to find that the gender gap in performance under competition and preferences for competition can be partly explained by the differential responses of men and women to the above pressures. In particular, while women underperform the men in a high-pressure math-based tournament, women greatly increase their performance levels and their willingness to compete in a low-pressure verbal environment, such that they actually surpass the men. This effect appears largely due to the fact that extra time in a verbal competition improves the quality of women’s work, reducing their mistake share. On the other hand, men use this extra time to increase only the quantity of work, which results in a greater relative number of mistakes. A labor market study suggests that the nature of the job and the stress level are correlated with the gender gap in the labor market in a manner consistent with the results of my experiment.