No abstract

We investigate the effect that competing in teams has on gender differences in choosing to enter competitions. In our experiment, subjects chose whether to compete based on the combined performance of themselves and a teammate. We find that competing in two-person teams reduces the gender competition gap by two-thirds. Independent of the sex of oneÕs partner, female subjects prefer to compete in teams whereas male subjects prefer to compete as individuals. We find that this result is driven primarily by gender differences in competitive preferences, as opposed to other potential explanations such as risk aversion, feedback aversion or confidence.Previous research indicates that there is a large gender gap in competitive fields. For example, Bertrand and Hallock (2001) show that women accounted for just 2.5% of the top five executive positions in a sample of US companies. Even in 2008, just 2.8% of firms in the S&P 500 were led by female CEOs. This type of empirical evidence is consistent with a variety of potential underlying explanations including discrimination against women, differences in performance in competitive environments and differences in preferences for entering those environments in the first place. Economists have used experiments to investigate the reasons underlying gender differences in competitive environments. These studies have generally concluded that men enjoy competing more than women do, with some studies concluding that men perform better than women in competitive environments even when no performance differences exist once the competitive aspect is removed (Gneezy et al., 2003;Gneezy and Rustichini, 2004).Notably, Niederle and Vesterlund (2007; from here, NV) find that when there are no gender differences in performance, gender differences in competitive preferences and confidence persist. They find that women choose to compete too little and men choose to compete too much, concluding that women would be better off if they competed more and men would be better off if they competed less. 1 To disentangle the roles of competitive preferences, risk aversion, feedback aversion and confidence, NV had their subjects complete a series of adding tasks under different incentives. 2 We employ a

No abstract

No abstract

No abstract