Over the last twenty years women became much more prominent in the economics profession. Analyzing faculty and graduate students data for the top-10 economics departments between 1983 and 2007, we find persistent differences in gender composition across institutions. We test whether such differences are driven by female faculty influence or by departments’ gender preferences. Using a matched sample of PhD students and faculty, we find positive correlation between the share of female faculty and the share of women in the entering PhD class. We show that this correlation is explained by departments’ gender preferences that vary both across departments and over time, and not by influence of female faculty on the gender composition of the entering PhD class. We also find that these top-10 departments are targeting a certain trend in the number or share of women on their faculty, lowering on average the probability of hiring a woman by about 10 percentage points for each additional woman already on the faculty. This evidence suggests that gender differences across institutions are explained by gender preferences and policies rather than hysteresis related to women’s influence.