A Study on the Clustering of Household Characteristics for Sustainable Public Rental Housing

Abstract

Public rental housing types have been integrated recently and housing benefit support has been expanded as part of a housing stabilization support policy. Against this background, it is necessary to reinterpret the number and type of households living in public rental housing for sustainable public rental housing. This study focused on household type and used the 2016-2019 public rental housing panel data of the SH Urban Research&Insight. Based on the result of clustering according to household characteristics using the Gaussian Mixture Model (GMM) and analyzing the cluster type predictive factors using multinomial logistic regression, the characteristics of each cluster, which were categorized into four clusters, showed differences in the income type of the householder. Each cluster was named ‘earned income households’, ‘government transfer income households’, ‘social insurance income households’, and ‘general public households’. A study on the clustering according to household characteristics concluded that appropriate support should be provided for households that need customized benefit support. There is also a need for a rental type that allows elderly households with no income to live stably for a long time without the burden of monthly rent. In addition, it was derived that financial support measures should be prepared for practical housing stability specifically for those considered financially vulnerable.