Abstract
Recent research documents that while men are eager to compete, women often shy away from competitive environments. A consequence is that few women enter and win competitions. Using experimental methods we examine how affirmative action affects competitive entry. We find that when women are guaranteed equal representation among winners, more women and fewer men enter competitions, and the response exceeds that predicted by changes in the probability of winning. An explanation for this response is that under affirmative action the probability of winning depends not only on one’s rank relative to other group members, but also on one’s rank within gender. Both beliefs on rank and attitudes towards competition change when moving to a more gender-specific competition. The changes in competitive entry have important implications when assessing the costs of affirmative action. Based on ex-ante tournament entry affirmative action is predicted to lower the performance requirement for women and thus result in reverse discrimination towards men. Interestingly this need not be the outcome when competitive entry is not payoff maximizing. The response in entry implies that it may not be necessary to lower the performance requirement for women to achieve a more diverse set of winners
Despite decades of striving for gender equality, large differences still remain between men and women in the labor market. Perhaps most noteworthy is the gender segregation across different types of jobs. While there is substantial horizontal segregation, with women more likely to hold clerical or nurturing jobs and men more visible in manufacturing, the vertical segregation within a sector is particularly striking (Weeden, 2004, and Grusky and England, 2004, Ander, 1998). Across fields men are disproportionately allocated to professional and managerial occupations. In a large sample of US firms Bertrand and Hallock (2001) show that women only account for 2.5 percent of the five highest paid executives.1 While it may be argued that such segregation is a result of past history, and that these differences will diminish over time, it is noteworthy that women are underrepresented among the people who have the minimum training frequently required for senior management. Only 30 percent of students at top tier business schools are women, and, relative to their male counterparts, female MBA’s are more likely to work in the non-profit sector, work part time, or entirely drop out of the work force.2
It is commonly argued that discrimination, preference differences for child rearing, and ability differences can explain the absence of women in upper level management.3 Recent research suggests that an additional explanation is that women are more reluctant to put themselves in a position where they have to compete against others (see e.g., Gneezy and Rustichini, 2005, Gupta, Poulsen and Villeval, 2005, and Niederle and Vesterlund, 2007, henceforth NV).4 For example, NV examines compensation choices in an environment where men and women are equally good at competing. They find that the majority of men select the competitive tournament whereas the majority of women select the non-competitive piece rate. While low ability men are found to compete too much, high ability women compete too little, and few women succeed in and win the tournament.
From the firm’s perspective it is particularly costly if the upper tail of the performance distribution does not enter competitions for jobs or promotions. As explained by B. Joseph White, president of University of Illinois, “Getting more women into MBA programs means better access to the total talent pool for business”.5 An additional argument for increasing the number of women in top managerial positions is that diversity in and of itself may benefit the firm.6 Indeed US corporations are concerned by their inability to attain and recruit women, and they are increasingly developing programs to improve the number of women employees.7
When instituting programs to alter the gender composition in certain jobs it is of course important that we understand how these programs influence behavior. To begin this process, we use experiments to investigate how affirmative action may affect participants’ willingness to compete. Specifically, we consider a quota system which requires that out of two winners of a tournament at least one must be a woman.8 We examine the consequences such a system may have on the individual’s decision to compete and thereby on the resulting gender composition of the applicant pool. Accounting for this response we ask how costly it is to secure that women be equally represented among those who win competitions. In particular, how much lower will the performance threshold be for women? How many better performing men will have to be passed by to hire a woman? To what extent will reverse discrimination arise? These questions are particularly interesting in light of the non-payoff maximizing tournament-entry decisions documented by NV.
No comments:
Post a Comment