사회적 혼합이 거주자의 사회적 배제와 주변 주택가격에 미치는 영향

The objective of this study is to empirically analyze the impacts of social mix development(SMD) and to provide politic implications. Based on the layout by ownership, the SMD types are classified into four: segregated types(isolated type, adjacent type) and mixed types (mixed-in-a-building type, mixed-in-a-complex type). Then this study employs the empirical analysis of 50-year public rental housings in Seoul. The empirical analysis consists of two parts. First, this study analyzes the impacts on social exclusion of public rental housing tenants. Using the survey results, analysis of variance(ANOVA) is employed to test the differences of residents' attitudes in the 'mixed type' and 'segregated type' in terms of psychological exclusion, anti-social behaviors in neighborhoods, and the discrimination or stigma from the outside. The differences of social capital characteristics are also tested by ANOVA. The second is to analyze the impacts on the housing prices around 50-year public rental housings in Seoul by one- and multi-level hedonic price models. The results can be summarized as follows. First, the degree of social exclusion is different by the SMD types: the degree of social exclusion in the 'segregated type' is higher than that in the 'mixed type.' The social capital is also different by the SMD types. Residents in the 'segregated type' show rather negative housing identity and do not show strong community solidarity compared to those in the 'mixed type.' Second, findings of both one-and multi-level models suggest that the housing prices around 50-year public rental housings tend to be negative with 'rental housing units/squared distances from public rental housing' variable. The result of multi-level models is statistically insignificant, implying that external effects of 50-year public rental housings become less significant by random errors on the 2-level model. Third, the outcome of one-level hedonic price model suggests that the negative effect of 'mixed type' variables on the housing prices is significantly lower than that of 'segregated type' variable. On the other hand, in the multi-level models the SMD type variables are not significant due to the insignificant external effects of public rental housings. Finally, findings from both models indicate that the housing prices in the 'mixed-in-a-building type' become lower than those in the 'mixed-in-a-complex type' under the same conditions. While the coefficient of SMD type variable in one-level model is significant, that in multi-level models is insignificant. It implies that the impacts of 50-year public rental housings are not critical on housing prices.