Gender and Risk Taking in the Classroom

Abstract

We examine whether differences in risk preferences explain gender differentials in test scores amongst a large class of undergraduate microeconomics students, where students were evaluated using multiple choice questions. In each of five class tests, the negative penalty associated with an incorrect answer was randomly varied across questions. We show that female students exhibit lower risk propensities on average, and that they are more responsive than males to an increase in the penalty for an incorrect answer. Controlling for differences in risk preferences, we show that the gender differential in relation to answering any given question correctly reduces by a third, and that the gender differential in overall test scores becomes statistically insignificant. This result is robust to a variety of distributional assumptions.