Abstract

To understand differences in labor market outcomes between genders, economists must examine a complex array of potentially significant factors, such as institutional context, productivity differences, child-bearing and home production, and bargaining behavior. Many of these factors are not well captured by standard data sources. We use a new survey of academics in Japan to better understand the sources of gender pay differences. We find a 6% pay gap which persists when we control for research productivity, despite an institutional context shaped by explicit salary tables. We do not find a motherhood wage penalty, and the gender salary we document is not affected by differences in outside job offers.