文摘
The purpose of this dissertation is to explore the relationship between low-income subsidized households/housing units and neighborhood quality while using a neighborhood change theoretical framework. Specific research questions that are addressed are: do low-income subsidized households/housing units impact neighborhood change or are low-income subsidized households/housing developers responding to changing neighborhood conditions? The research design is a case study of 16 US metropolitan areas. The two primary data sources are census data from the Neighborhood Change Database (NCDB) and Department of Housing and Urban Development (HUD) administrative data from the Picture of Subsidized Households (PSH) data file. Census tract median household income and median home value are the two neighborhood quality indicators used with the categorical outcome variable of large decline, stable/small change, and large gain. Predictor variables include neighborhood change variables: population, housing stock, and neighborhood characteristics. Multinomial logit regression, using base subsidized housing, neighborhood change, and PSH models, provides hypotheses testing of the impact of neighborhood indicators and PSH on how neighborhoods changed from 1990 to 2000 and impact of neighborhood ranking in 1990. The results indicate that subsidized households are more likely to be located in lower ranked neighborhoods based on income and home value, compared to general metropolitan trends. Second, PSH housing units significantly, though at an extremely small level, increased the likelihood that a neighborhood experienced a large decline and decreased the likelihood of a large gain from 1990 to 2000, compared to stable neighborhoods. Third, housing stock age and neighborhood characteristics are not primary indicators of neighborhood change. Fourth, neighborhoods with high levels of assisted households and total poor households may or may not experience neighborhood change compared to neighborhoods with no or low levels of assisted/poor households.