Does electing female leaders empower women’s political engagement? While past research focused on the effect of descriptive representation on women’s political attitudes and stated preferences, we have little evidence for its behavioral impact on the citizens’ level. Measuring this effect is challenging because data, such as voter turnout, typically occur before these significant events and are too sparse in time to allow causal comparisons. In this study, we utilize a unique dataset from Slovakia to measure civic engagement in a continuous fashion. We collect data from an online platform (\Odkaz pre starostu) that allows citizens to inform their municipal authority of street problems needing their attention. We then use an established machine classification method to determine the gender of the users, exploiting that many of them use their full names when submitting problem reports. We employ this rich dataset (N = 92,178) in a natural experiment: we set out to identify the effect of Zuzana Čaputová's election as Slovakia’s first female president in 2019 on the gender gap in problem reports using a regression discontinuity in time (RDiT) technique. We find that the submission rate of female versus male users increased by 11 percentage points in the immediate aftermath of the election. Our study offers evidence for a plausible input-side effect of descriptive representation on mass democratic inclusiveness and political efficacy.