A spatial analysis of housing growth

Abstract

This study accounts for spatial spillover and spatial heterogeneity to estimate housing growth between 1990 and 2000 using Tennessee census-block group data. A deterministic estimation method, inverse distance weighted averaging is used to create neighborhood variables that can capture more accurate measurement of spatial spillover while reducing multicollinearity. The block-group specific local effects on housing growth are computed using locally weighted regression, and they are mapped using a geographical information system.