Unraveling the dynamics of vacant housing: Multi-scale analysis of socio-demographic-economic, physical environmental, and spatial clusters in South Korean cities (2015–2020)

초록

Vacant housing has emerged as an increasingly pressing urban challenge, impacting economic stability, social cohesion, and land use efficiency. This study investigates the multi-scale determinants of vacant housing across South Korean cities during 2015 and 2020, exploring the interplay of city-level socio-demographic-economic (SDE) factors, neighborhood-level physical characteristics, and spatial clustering effects. Using multi-level regression (MLR) models, we assess how city-wide trends such as population dynamics, migration patterns, and housing supply shape vacancy rates at the neighborhood level. Our findings reveal that population growth, higher birth rates, and net migration at the city scale are negatively associated with vacant housing rates, while housing oversupply increases vacancies. At the neighborhood scale, aging housing stock, a high proportion of detached homes, and elevation significantly affect vacancy rates, whereas improved road infrastructure correlates with lower vacancy rates. Furthermore, spatial clustering analysis identifies localized hot and cold spots of vacancy, highlighting spatial dependencies beyond city- and neighborhood-level factors. These findings emphasize the need for scale-sensitive policy interventions, including city-level economic revitalization strategies, targeted neighborhood renewal, and spatially adaptive housing policies to mitigate vacancy concentration. By integrating spatial clustering techniques with multi-level modeling, this study enhances predictive accuracy and provides a more comprehensive understanding of urban vacancy patterns, offering data-driven solutions for sustainable urban development.

키워드