Child Penalty Estimation and Mothers’ Age at First Birth

Abstract

Motherhood continues to pose significant challenges to women’s careers, and a correct assessment of its effects is crucial for understanding the persistent gender inequality in the labor market. We show that the prevalent approach to estimate post-birth earnings losses – so called “child penalties” – is prone to yield substantially biased results. We demonstrate that the biases stem from conventional event studies pooling together first-time mothers of all ages, without considering their distinct characteristics and the varying impact of motherhood. To address the biases, we propose a novel approach that accounts for the heterogeneity by building upon recent advancements in the econometric literature on difference-in-differences models. Applying it to administrative data from Germany, we find substantially larger estimates of earnings losses after childbirth (by 25-30%), indicating that the costs of motherhood and related gender gaps in Germany are even larger than previously thought. Moreover, we document that effects and their interpretation differ significantly depending on maternal age at birth. Our results demonstrate that heterogeneity of effects by age is more than just a source of bias – it is an opportunity for better understanding of motherhood impact. Joint paper with Lukas Riedel.