Can Laboratory Experiments Explain Gender Differences in Labor Market Outcomes?
The source of the gender differences in labor market outcomes has long been in dispute. The debate has been principally between free market economists, who believe that human capital and women’s family choices are the source of the differences, and institutional economists, who believe that discrimination and institutional barriers in labor markets prevent women from succeeding to the same extent as men. Recently, a new explanation focusing on personality attributes has surfaced: women are less willing to compete and perform worse under competitive conditions than men, thus possibly reducing their ability to perform in the labor market relative to men. While these behavioral tendencies have been well established in extensive laboratory research, there is virtually no research on whether these personality factors predict subsequent labor market outcomes. Anne Preston and Linda Kamas will address this question by examining how strongly competitiveness, measured in the lab, is correlated with subsequent labor market success.
Having conducted a successful pilot study in 2011 with 355 subjects, Preston and Kamas seek to test the hypothesis that competitiveness, confidence, and other personality characteristics affect labor market outcomes and contribute to the gender gap in labor market performance on a larger sample of approximately 1050 subjects. Specifically, they will compare individuals’ behavior in the lab to their subsequent career performances by following laboratory subjects for five years after graduation. The experiments will be conducted in both co-ed and all-female colleges in order to ascertain whether women who attend single-sex colleges differ in both measured personality characteristics and labor market performance. During senior year in college, subjects will participate in laboratory experiments that measure their willingness to compete, confidence, risk aversion, social preferences, and gendered personality characteristics. They will then be surveyed annually to ascertain their progress in the labor market as well as changes in personal and human capital characteristics.
The outcome of this research is expected to be a rich data set merging experimental and labor market data. Econometric analyses of these data will produce at least five research papers.