The purpose of this paper is to find detenmnants and region-specific difference of preference for present residential area. An empirical analysis is applied to the survey data in which households are clustered into 21 regions in Seoul metropolitan are...
The purpose of this paper is to find detenmnants and region-specific difference of preference for present residential area. An empirical analysis is applied to the survey data in which households are clustered into 21 regions in Seoul metropolitan area. First, preference for residential area is analysed by O-D type table and the reason of preference
is analysed by ranking table and chart. Second, random intercept logit model, a basic form of multi - level model is applied to the survey data. First, Yongsan- gu, Yangcheon- gu, Gangnam-gu and Seocho- gu in Seoul and Seongnam-si and Yongin-si in Gyeonggi-do are highly preferred by residents. Second, among the independent variables of random intercept logit model, the variables of personal and household characteristics, such as woman, values having to own a house, married daughters and sons living close to the present house, living in apartment and the amount of asset are statistically significant and positive but the amount of income is negative. The variables of regional characteristics, such as the amount of local taxes and the increasing rate of housing price are statistically significant and positive. Third, the region- specific difference of preference for present residential area is
empirically verified by the variance of random intercept logit model