Working Paper, Federal Reserve Bank of Atlanta 2001-2
This study uses data from the Survey of Doctorate Recipients to evaluate differences in employment outcomes for academic scientists by gender. A decomposition of estimated salary differences shows that over time, gender salary differences can partly be explained by differences in observable characteristics for faculty at the assistant and associate ranks. Substantial gender salary differences for full professors are not explained by observable characteristics. Probit and duration model estimates indicate gender differences in the probability of promotion, making it less likely for women to be promoted to tenure. Between 1973 and 1997, very little changed in terms of gender salary and promotion differences for academics in science. After evaluating potential explanations, the author concludes that gender discrimination similar to that observed at the Massachusetts Institute of Technology accounts for unexplained gender disparities.