Local Spillovers, Externalities and Sorting of Households by Race and Income in Locational Equilibrium

National Bureau Of Economic Research Inc, Cambridge MA

Investigators

Abstract

In this project, we investigate the determinants and consequences of household residential location decisions. In particular, we study factors leading to stratification of households across communities by income, demographic characteristics, and preferences. For example, households may choose communities based on housing prices and the quality of local public goods and amenities such as education, public safety, and environmental quality. In addition, preferences for income, race, or other demographic characteristics of neighbors may play a role in household location choices. We also investigate how residential choices of individuals affect neighborhood and community outcomes. Such effects may operate through collective (e.g., voting) decisions that affect expenditures on education, police, and other public services. Effects may arise more directly via neighborhood effects-interactions among individuals that affect the character and quality of neighborhoods and the services they provide. For example, parental involvement and peer effects both within and outside schools may affect children's educational achievement and social development. Interactions within neighborhoods may also affect public safety and the quality of the neighborhood environment. Choices of individuals may also affect satisfaction of others with a neighborhood if households care about income, race, and other demographic characteristics of neighbors. A particular focus of our work, then, is to understand the interaction of preferences for neighborhood demographic composition and local public goods in determining the sorting of population by race and other characteristics across communities. We also investigate the importance of spillover effects and externalities within communities and across neighboring communities. Our contributions are both to development of new methods and to application of those methods. In previous research, we and others have developed strategies for studying stratification in models that impose considerable a priori structure on the patterns of household sorting across communities and associated variation in housing prices and public service quality. In particular, such models imply a common ordering across communities of household incomes, housing prices, and public service quality. While these models provide many valuable insights and have considerable predictive power, they are nonetheless restrictive. In particular, they allow for relatively limited variation across individuals in demographic characteristics and preferences, and, where multiple local public goods are present, they imply a common ordering of qualities across communities for the various goods. Our research advances the state of the art by developing models and computational and econometric methods that permit consideration of multiple observed and unobserved characteristics of individuals and multiple local public goods and amenities. The framework that we are developing also accommodates neighborhood effects and spillovers across neighborhoods. Efforts to understand the adjustments of heterogeneous households to spatial differences in local public goods, environmental amenities, and neighborhood quality have important policy implications. Our comprehensive analysis of the sorting and mixing of households provides improved understanding of racial segregation and income stratification observed in metropolitan areas. Our framework also permits investigation of the effects of large-scale changes such as tax limitations, school finance equalization programs, changes in environmental quality, and falling crime rates. In particular, our approach permits investigation of how such changes affect location decisions of households, how gains and losses arising from the changes are distributed across households, and how the changes affect the distribution of housing prices across communities.