The Power of Knowing a Woman Is in Charge: Lessons from a Randomized Experiment

As part of a term-long anonymous assessment, thousands of undergraduate university students were divided into groups, each led by a randomly selected peer. The leader's gender had no effect on the assessment's outcomes, but female students led by a female peer achieved 0.26 SDs higher course grades when the gender of the leader was revealed. They also outperformed by 0.22 SDs their counterparts in groups where the leader's gender was revealed to be male. The mechanism involved attempting more difficult practice questions. Our structural estimates suggest this operated via a stereotype threat reversal that reduced anxiety.