Unpacking neighborhood effects : experimental evidence from a large-scale housing program in Brazil

This paper investigates the impacts of neighborhoods on the economic outcomes of adults. We exploit one of the world's largest housing lottery programs and administrative data linking lottery registration, formal employment, and access to social programs in Brazil. Receiving a house has positive impacts on housing quality and reduces household expenditures but has negative effects on beneficiaries' neighborhood characteristics. On average, the program has a negative impact on the probability of being formally employed but no effect on the quality of jobs. Poorer individuals, however, experience better formal employment outcomes and lower welfare dependency. We find no differential impacts by distance to beneficiaries' previous homes or jobs. Leveraging a double-randomization design to allocate houses, we show that there are significant differences in effects across neighborhoods and we propose a framework to estimate the relative importance of potential underlying mechanisms. Network quality, amenities and crime play a very limited role, while labor market access explains 82-93% of the observed differences in neighborhood effects.