Houses for the People, Power for the Gangs

Housing policies are designed to reduce housing deficits and provide secure property rights to low-income families. We argue that, in contexts where organized crime is present, large-scale social housing can unintentionally generate territories of opportunity for criminal organizations. We exploit population thresholds that determine the maximum size of a large-scale housing program in Brazil (Minha Casa, Minha Vida—MCMV), together with longitudinal homicide data for all 5,570 municipalities to estimate the causal effect of housing on lethal violence. We show that municipalities assigned more housing units experience higher homicide rates, with effects that are concentrated among populations that are more vulnerable to organized crime victimization, that grow over time, and are stronger in larger cities. Drawing on original fieldwork, we trace how criminal groups invade developments, expel residents, and use housing complexes as bases for extortion and territorial control. Our findings highlight how redistributive policies can inadvertently strengthen criminal governance.

Under Review. REPAL Conference Best Paper Award, 2025.