The Effects of Fmla on Women's Promotion in the Federal Government

Abstract

We examine the effect of the Family and Medical Leave Act (FMLA) on the relationship between fertility rates and promotion of women . We combine OPM data on the careers of U.S. federal civil servants with Centers for Disease Control and Prevention natality data on the average number of children born per female by year, race, and age bracket. The idea is that each female worker can be assigned an expected fertility for a woman of her age, race, and year, and we can relate this expected fertility to her chance of promotion, both before and after the passage of the FMLA.

We estimate a linear probability model explaining promotion into managerial (central) positions. The explanatory variable of interest is an interactive term between expected fertility and a dummy variable for 1993-2011 (years in which the FMLA was active). After controlling for age, job tenure (experience), level of education, yearly dummies, age dummies, race dummies, and occupational dummies, we want to know if the relationship between average fertility rates and promotion probabilities changes in the years after the FMLA was implemented and, if so, how much and in which direction.

We find that after the FMLA was passed in 1993, there was a significant change in the relationship between expected fertility and promotion, with expected fertility becoming more negatively associated with promotion. Compared to the relationship prior to 1993, a 10 percentage point increase in expected fertility is associated with an additional 0.86 percentage point decline in the probability of being promoted. This relationship is robust to several definitions of promotions, and suggests that something changed in the years since the FMLA was passed that may have inhibited the relative progress of women in high-fertility demographic groups up through the U.S. federal civil service system. When we run the same model on men, matching them with the corresponding female fertility, the coefficient on interactive term between fertility and a dummy variable for 1993-2011 is insignificant.