Is it what you do or where you work that matters most? Gender composition and the gender wage gap revisited

The purpose of this study is to examine the impact of gender segregation on wages using matched employer-employee private-sector data from Sweden. The questions that we are interested in examining are two-fold. Has the effect of gender segregation on the gender wage gap been overestimated and what matters more for gender wage differentials, occupation or establishment segregation? Our results show that a too detailed aggregation of occupations and/or establishments leads to an overestimation of the segregation effect on gender wage differences. We also show that occupational segregation contributes more to explaining the wage gap than establishment segregation.