Gender and Competition: Do competitive environments favor men more than women?

University Of Chicago, Chicago IL

Investigators

Abstract

Although women have made huge strides in catching up with men in the workplace, a gender gap persists both in wages and levels of advancement. Commonly cited explanations for this gap range from charges of sex discrimination to claims that women are more sensitive than men to work-family conflicts and thus less inclined to make sacrifices for their careers. We suggest that another factor may be at work: a deeply ingrained difference in the way men and women react to competition that manifests itself even at an early age. A previous study (Gneezy, Niederle and Rustichini, QJE, 2003) involved several hundred students. Groups of six students were paid to solve simple maze problems on a computer. In some groups, subjects were paid 50 cents for each problem they solved during the experiment. In others, only the person solving the most problems got rewarded--but at the rate of $3 for each maze solved. Regardless of the sexual makeup of the groups, men, on average, performed slightly better at solving mazes when students were paid for their own performance. But when only the top student was paid, average male performance rose sharply--by about 50%--while female performance remained the same. A plausible explanation for this result is that because men are slightly better at solving mazes, they have higher incentives to increase effort in the competition. We now wish to test this explanation by investigating other tasks that may favor women, such as solving anagrams. These new tasks will be performed in a very similar environment to the maze study. The main research question is does moving to the new task change (reduce) the gender gap? The main conclusion from the findings so far is that females tend to be far less responsive to competition than males--a tendency with important implications. If the gap is affected by the nature of the task and the stereotypes, then the policy implications of the findings should be refined. For example, when considering a single sex classes, should it be only for math where the stereotype is that boys are better, or also for literature in which the stereotype is that women are better.