The goals of federal housing policy are to provide every American family with a decent home in a suitable neighborhood. While substantial progress has been made toward stisfying the goal of a decent home, survey evidence indicates that many Americans, especially those living in central cities, are highly dissatisfied with their neighborhoods. While policymakers are fully aware that too many people have a low opinion of the overall quality of their neighborhood, there is little reliable evidence available on what neighborhood attributes matter most to people and how neighborhood preferences vary among different types of households. As a result, policymakers have little idea how best to allocate scarce public resources to achieve the greatest possible improvement in neighborhood quality.
This paper implements a new methodology with new data in order to reveal the neighborhood preferences of households categorized by race, income level, location, and type of housing occupied. The methodology involes interpreting the ranking that households assign to the overall quality of their neighborhoods on a ten-point scale as an ordinal utility index. This index enables us to observe directly the relationship between neighborhood variables and individuals' utility. To handle the ordinal nature of the dependent variable, N-chotomous multivariate probit is used as the estimating technique.
The results suggest that while many neighborhood variables affect the utility of all households similarly, there are differences in preferences among groups, especially between black and white households.