Measuring the Pay Disparity between Typically Female Occupations and other Jobs: A Bivariate Selectivity Approach

Publisher Website
Google Scholar
Abstract
The author of this study, using a bivariate probit selectivity model with data from the 1984 Panel Survey of Income Dynamics, finds that women in female-dominated jobs earned 6–15 percent less than women with the same characteristics in other occupations. These results support the crowding hypothesis, according to which women are crowded into “female jobs” because of employer discrimination, resulting in lower wages for those jobs. Previous studies, using ordinary least squares analysis, have found a higher earnings differential; but most of those studies, unlike the present one, failed to control either for individuals' decision whether or not to work or for their choice of occupation.