Parental Leave Duration and Wages: A Structural Approach

We investigate the existence of causal mechanisms from parental leave duration to subsequent wages. Our instrumental variable is a French reform giving financial incentives to take a parental leave. Two longitudinal datasets provide us with information on wages and familial background from 1976 to 2005. In our context, panel data estimations potentially suffer from unobserved heterogeneity, endogeneity and selection. We implement an innovative procedure proposed by Semykina and Wooldridge (2010) to take into account these three problems simultaneously. We find that parental leave duration has a significant and negative causal impact on later wages