Households amid Urban Riots

Abstract

This article analyzes the determinants of household riot victimization, based on a unique survey collected in Maharashtra, India. We adopt a multilevel framework that allows neighborhood and district effects to randomly influence household victimization. We find that economically vulnerable households, households living close to unsafe areas, and shop owners are more prone to suffer from riots. Households report lower levels of victimization if they live further from police stations, exhibit higher levels of trust, are able to rely on outside help in times of need and accumulate savings. The results show, however, a double-edge effect of income: wealthier households are better able to cope with the adverse effects of riots, but also have more to lose from riots and are more exposed to opportunistic violence and looting. We find further that affluent neighborhoods and neighborhoods where caste fragmentation is high report higher levels of victimization. Neighborhoods with stronger social interactions experience lower levels of victimization.