Previous studies of female suffrage have interpreted the change in voting patterns as reflecting a change in voter composition, in part because only aggregate voting data was available. We exploit the existence of separate counts for women and men votes in Chile before and after female suffrage. We show that inference based on aggregates is inaccurate because men also change their voting behavior. Two potential explanations are provided: men responded to female suffrage through strategic voting and men previously represented in part women's vote due to negotiation within the household. We show evidence consistent with both hypotheses.