This dissertation is a socio-economic approach to the residential location model. Heterogeneity of households’ residential location decisions and actual trade-offs made in the housing market are explored. Specifically, this dissertation explores both heterogeneous residential location decisions among different categories of households and trade-offs between location and housing quality. It is assumed that social values are reflected in locations, and locations with highly valued characteristics are preferred by homebuyers, and that living in preferred locations means that the household has higher social status. In order to live in preferred locations, households may have to make a trade-off by giving up some housing quality. This research uses a repeat homebuyers’ dataset in Franklin County, which is the central county in the greater Columbus, Ohio metropolitan area. I utilize the two-stage nested logit model with interaction terms for different household characteristics.
The estimation results provide a range of information about households’ residential location decisions. First, households with school-aged children are attracted to locations with high school quality and high-income neighbors, while choosing older houses (with more bedrooms). In other words, they make a trade-off by choosing older houses—less expensive homes in the housing market—in order to live in these preferred locations. Second, middle-income households with school-aged children are more likely to choose these same preferred locations, but they tend to choose older and smaller houses with more rooms. As expected, their relatively limited incomes seem to stimulate the trade-off between location and housing quality. They even accept higher crime rates in their community choices. Third, households with householders who are 50 years old and over prefer safer communities, indicating that safety is an important social value to this group. In their dwelling choices, this group is more likely to choose multi-family attached houses (e.g., condos, duplexes, and townhouses) with fewer bedrooms. This group’s age-specific circumstances seem to affect their housing choices significantly.
The operationalization of the concept of social status with the status of locations as preferred locations is one contribution of this work. The second contribution is that this study broadens the perspective of the residential location modeling by providing a variety of determinants associated with community and dwelling characteristics. The utilization of the random utility theory—particularly the nested logit modeling—makes the third contribution. I believe that this study will help extend the understanding of actual determinants of residential location decisions in the U.S. housing market, and improve the urban residential location theory.