The relationship between built environment and housing vacancy rates: Can compact development reduce vacancy?

Urban housing vacancies represent a mismatch between the population and available housing. Studies investigated the direct influence of specific built environment factors on housing vacancy rates; however, discussions from the perspective of compact built environments remain scarce. Structured around the 5D framework, this study empirically investigates 154 subdistricts in Guangzhou and employs a random forest approach to assess how built environment factors can affect housing vacancy rates. This study obtains several key findings. (1) A compact built environment can effectively reduce housing vacancy rates. Within certain thresholds, housing vacancy rates decline as the POI density and the building density increase. Low housing vacancy rates are observed in areas with small plot areas, high road density, and high education accessibility. (2) Among the five dimensions of the built environment, the density dimension plays the most significant role in reducing housing vacancy rates, followed by the design dimension. Moreover, the effects of built environment factors on housing vacancy are nonlinear and exhibit clear threshold effects, thereby suggesting that urban planning and design should account for reasonable threshold levels. The conclusions highlight the complex nonlinear impact of urban built environments on housing vacancy and provide empirical evidence supporting the role of compact built environments in addressing housing vacancy. The findings offer insights for sustainable urban development.