공동주택 관리비 결정요인 분석: 다수준 종단분석

Abstract

The rising concerns about housing management fee appear to focus on fraud detection, ignoring value neutral analysis, which motivates this study to identify factors affecting housing management fee within objective analysis framework. The data used in the study are monthly maintenance fees which are nested in the APT complexes as well as in the measurement period of month. We employed a multilevel modeling approach taking account of the spatio-temporal cross-nested structure of data, and chose to apply the Hamiltonian Monte Carlo algorithm(HMC), one of the Markov Chain Monte Carlo(MCMC) numerical techniques, to estimate parameter values in the multilevel model. Analysis results indicate that seasonality is one of the maintenance fee determinants. Private utility fee, one component of the maintenance fee, shows a particularly apparent seasonality. The traits of APT complexes that go unmeasured are found to be extremely strong determinants of the maintenance fee; geographical location, building layout and other various factors that are not available or cannot be measured are all examples of the traits of APT complexes. We identified that the households per building is an important management fee-increasing factor. Finally, scale of housing estate and property age are also revealed to be determinants of the maintenance fee. The interaction effect of the two determinants showed that property age impact, representing building deterioration, is more apparent in large housing estate than in small one.