Female leadership remains strikingly low in most democracies, and voter preferences are
often suggested as a likely explanation. In this paper, we present experimental evidence
from India which suggests that, on average, villagers, especially men, are prejudiced
against female leaders. For example, men rate a hypothetical leadership speech more
negatively when the speaker's voice is experimentally manipulated to be female, rather
than male. However, randomly assigned exposure to a female leader (due to mandated
political representation for women) reduces such prejudice by 50-100% depending on
the measure. We also provide suggestive evidence that prejudice influences perceptions
of actual performance. Despite outperforming their male counterparts on many dimensions of performance, first time women leaders receive worse evaluations. Consistent
with our experimental evidence that exposure reduces prejudice, second time female
leaders are rated at par with male leaders.