Estimating the wage penalty for maternal leave

Abstract

The focus of this paper is the size of the wage penalty due to maternal leave incurred by working mothers in Germany. Existing estimates suggest two-digit penalties of up to 30 percent, with very little rebound over time. We apply recent panel data methods designed to address problems of sample selectivity, unobserved heterogeneity and endogeneity to German Socio-Economic Panel (GSOEP) data. The selectivity issue arises because no wage is observed for employees who are on leave. Heterogeneity takes the form of unobserved individual effects correlated with explanatory variables. Endogeneity is due to the simultaneity of the wage and participation outcomes. Heckman’s classic treatment of selectivity requires extensions to deal with both heterogeneity and simultaneity. We present an extension for the case of a censored tobit participation model and use it to exploit the actual working hours data available in GSOEP. We also investigate the sensitivity of the results to the choice of method. Our estimates imply a wage penalty due to maternal leave which although substantial remains below previous estimates. Furthermore, we find that this penalty is less persistent than other studies suggest. Five years after the career interruption mothers seem to have caught up.