Working Paper: CEPR ID: DP6891
Authors: Evren R. S. Frédéric Palomino; Eloc Peyrache
Abstract: Using data from a natural experiment with high payoffs in education, we examine whether the competitive nature of tournament structure explains the performance gender-gap. We find that performance is statistically lower for women, the variance of performance is higher for men, and the tails of the performance distribution are significantly fatter for men. For the same participants in non-competitive settings with similar academic content, the performance of women first-order-stochastically dominates that of men. We reject differences in risk aversion and ability as reasons for performance gender-gap.
Keywords: gender gap; relative performance; evaluation; tournament
JEL Codes: I29; J16
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Competitive environment (L13) | Women's performance (J16) |
| Competitive environment (L13) | Men's performance (D29) |
| Performance distribution (D39) | Gender performance gap (J16) |
| Variance of performance (D29) | Men's performance (D29) |
| Variance of performance (D29) | Women's performance (J16) |
| Competitive structure (L11) | Proportion of men admitted (I24) |