School of Public Policy Capstones

Publication Date



This paper utilizes Census Bureau and USPS-HUD data sets to examine the relationship between address vacancy and percent of residents who are housing cost burden at the Census Tract level. This research determines there is a statistically significant positive linear relationship for four of the six New England States. Additionally, I examine spatial autocorrelation patterns among the residuals to determine if error term is clustered, dispersed, or random. Finally, I compare my results to how building codes disincentive rehabilitating buildings and incentivize greenfield development. I offer a number of policy suggestions to assist at the local, state, and Federal level in adopting rehabilitation building codes.