Predicting Housing Vacancies in Depopulating Areas with Spatial GAMLSS

Abstract

Housing vacancy, exacerbated by population aging and decline, generates negative externalities that extend beyond property owners, impacting real estate markets and community stability. This study examines housing vacancy patterns and determinants in Jinju-si, distinguishing urban and non-urban areas. Using Spatial GAMLSS and hotspot analysis, the research identifies high-risk zones and highlights regional disparities. Urban areas face higher risks in older neighborhoods, declining commercial districts, and poorly connected zones, while non-urban areas with weakened economic bases or limited accessibility also show elevated risks. Conversely, regions with robust infrastructure or economic activity exhibit lower vacancy probabilities. The findings emphasize the importance of tailored vacancy management policies, such as downtown revitalization, transit-oriented development, and economic base strengthening. By providing spatial insights into high-risk zones, this study supports effective resource allocation and targeted interventions for sustainable housing management.