Description:
This chapter focuses on neighborhood effects in housing markets. Households in effect choose neighborhood effects, or more generally social interactions, via their location decisions, which renders them endogenous. Across several classes of models that it examines, it emphasizes how we may detect empirically the presence of neighborhood effects when they may be priced by housing markets and be capitalized into housing values and rents. The chapter focuses on models that are empirically relevant and help identify neighborhood effects, and discusses actual empirical findings. The first class of models examined involves models of choice over discrete sets of individual dwelling units that allow for a multidimensional bundle of characteristics. These models extend the BerryLevisohnPakes characteristics-based approach and allow for endogenous contextual effects. The chapter develops in detail a specific application that also endogenizes contextual effects at a low level of dimensionality. The chapter examines neighborhood choice, with endogenous and contextual neighborhood effects, and housing demand (with housing being measured as a scalar) as joint decisions. This approach utilizes individual and neighborhood-level data at several levels of aggregation. The chapter next examines neighborhood effects within the canonical AlonsoMillsMuth urban model with a well-defined spatial structure. Finally, the chapter reviews hierarchical models of neighborhood location in the presence of social interactions. These models describe communities in terms of a low-dimensional vector of attributes that are aggregated into a public good whose consumption is nonrival. Such approaches are designed to utilize community-level data, along with information on the community-specific distributions of various sociodemographicsociodemographic characteristics of individuals.