The Demographic Impact of Leadership: Mayoral Education and COVID-19 Excess Mortality in Italy

Does the quality of elected leaders matter when institutions face an unprecedented shock? Political economists have documented that political selection shapes outcomes in ordinary times, but acute crises — with no script to follow and no time to learn — offer the sharpest test of whether who governs matters. We study this question in Italy during COVID-19. Combining weekly death registers with a Bayesian hierarchical spatio-temporal model, we estimate municipal-level excess mortality and exploit close mayoral elections in a regression discontinuity design to isolate the causal effect of electing a mayor with predetermined characteristics. College-educated mayors significantly reduced excess mortality during the first wave, lowering both its incidence and its intensity. The effect disappeared in the second wave, once national protocols and information had caught up. Two newly collected large-scale datasets — municipal policy acts and mayors’ social media communication — reveal two channels: educated mayors acted earlier and with more targeted measures, and communicated in a less alarmist, more guidance-oriented way. Political selection thus shapes a community’s resilience precisely when institutional routines break down.