Date of Award




Document Type


Degree Name

Doctor of Philosophy (PhD)


Department of Information Science

Content Description

1 online resource (vii, 102 pages) : illustrations (some color), color maps.

Dissertation/Thesis Chair

Shiguo Jiang

Committee Members

Shiguo Jiang, Alexander Buyantuev, Catherine T. Lawson, Kai Zhang


Housing Vacancy, Spatial Heterogeneity, Spillover Effects, Housing

Subject Categories

Geographic Information Sciences


The housing vacancy issue has long been a concern for community development and urban planning, especially in old industrial cities since the second half of the 20th century. Following the subprime lending crisis in 2007, the problem of housing vacancy again attracts attention as it causes problems like housing inequity, neighborhood redevelopment, and spatial justice in most U.S. cities. This dissertation systematically builds up an analytical framework to explore the mechanism of housing vacancy, using Buffalo urban area as a case study. Firstly, the research pays attention to the factors contributing to housing vacancy. The dissertation also investigates if there exists a time lag effect between housing vacancy and the associated factors and if the use type of vacant properties plays a role in shaping housing vacancy. Secondly, the dissertation explores the spatial spillover effect of vacant houses. Spatial regression models will be adopted to investigate the underlying spatial pattern of spillover effects for housing vacancy. In addition to the simplified spatially weighting method (Queen weight) which is popular in previous literature, this research develops two refined spatial weights to capture the spillover effects: inverse-distance weight (IDW) and shared boundaries weight (SBW). Finally, the dissertation explores spatial variations in the relationships between housing vacancy and the significant factors, using Geographically Weighted Regressions (GWR). The research finds that there exist remarkable spatial disparities in housing vacancy, especially at the finer block group level. Two datasets are utilized in the dissertation. The dataset on housing vacancy and related factors in the research is the U.S. Census Bureau 5-year American Community Survey (ACS), which is a nationwide dataset with a uniform standard for data collecting and recording. Another dataset is the US Department of Housing and Urban Development from the US Postal Service (HUD-USPS) which quarterly collects counts of vacant residential properties and aggregates them to the census tract level. Through the systematic analysis of housing vacancy, this dissertation not only identifies the factors of housing vacancy at the intra-city level but also advances knowledge of housing vacancy by understanding how spillover effects of vacant houses spread among neighborhoods and how the relationships between housing vacancy and associated factors vary over the urban area. With the findings of this research, cities can strategically allocate limited public resources to impacted areas in order to maximize the effectiveness of interventions.