Studies that attempt to measure the impact of neighborhoods on children's outcomes are susceptible to bias because families choose where to live. As a result, the effect of family unobservables, such as the importance parents place on their children's welfare, and other unobservables that are common to geographically clustered households, may be mistakenly attributed to neighborhood influences. Previous studies that attempt to correct for this selection bias have used questionable instrument variables. Furthermore, results presented in this paper suggest that IV neighborhood effect parameters are sensitive to reasonable alterations in the instrument lists proposed in the literature.
This paper introduces an approach based on the observation that the latent factors associated with neighborhood choice do not vary across siblings. Therefore, family residential changes provide a source of neighborhoodbackground variation that is free of the family-specific heterogeneity biases associated with neighborhood selection. Using a sample of multiple-child families whose children are separated in age by at least three years, I estimate family fixed effect equations of children's educational outcomes. The fixed effect results suggest that the impact of neighborhoods exists even when family-specific unobservables are controlled. This finding is robust to changes in estimation techniques, outcome measures, neighborhood measures, variable definitions, and samples.