Motherhood and the onset of intimate-partner violence

This paper estimates the causal impact of first-time motherhood on the onset of physical or sexual intimate partner violence (IPV). We use retrospective data from Peru's Demographic and Health Surveys (2004-2019), and conduct a stacked difference-in-differences estimation comparing first-time mothers to observationally similar women who conceive three years later, matching on pre-conception characteristics. Unlike previous studies based on administrative data, we address the key empirical challenge of constructing a credible counterfactual by exploiting detailed survey data on pre-conception relationship characteristics and incorporating these characteristics into the matching procedure. We find that motherhood increases the likelihood of IPV onset by 4.3 percentage points (a 29% increase) within two years after conception. Importantly, we find that failing to account for pre-conception relationship characteristics when constructing the comparison group leads to substantial upward bias, highlighting their importance for credibly identifying the causal effect of motherhood on IPV.