住宅負擔能力之分量迴歸分析──以新竹市公寓大廈住戶為例

Quantile Regression Analysis for the Local Level Housing Affordability-An Example of Households Living in Condominium in HsinChu City

住宅是人一生中最重要而且不可或缺的商品，但是價格非常昂貴，當住戶購買住宅時，若房價與家戶所得間差距過大時，不僅會增加家庭負擔，甚至會生活失序，影響都市發展。公寓大廈是新竹市最常出現的社區型式，本文乃以居住在新竹市公寓大廈的住戶為研究對象，來分析家庭的住宅負擔。本文利用房價所得比建立家戶的住宅負擔能力指標，再以分量迴歸模型分析住宅屬性與社區特徵，對不同分量住宅負擔能力的影響性質與程度。利用最小平方法估計解釋變數的係數，模型操作結果發現解釋變數間存在多重共線性的問題，而且無法分析解釋變數的異質性。在剔除共線性的解釋變數並且操作四個跨分量迴歸分析之後，本文發現住宅負擔二極端（第0.1分量與第0.9分量）的住戶中，住宅負擔最為沉重的住戶，選擇住宅應該以「北區」或是「香山區」，「住宅居住面積」愈大、「住宅樓層」愈高、「住宅衛浴套數」愈多的住宅居住，這種選擇會相對優於住宅負擔最不沉重的住戶選擇相同屬性的住宅居住。相反的，住宅負擔最不沉重的住戶應該避免選擇「住宅居住面積」大、「住宅樓層」高、「住宅衛浴套數」多、「北區」與「香山區」的住宅。 Housing is the most important and expensive economic goods for the individual to consume. When the gap between house price and family income has been widened would not only increase family affordability but also influence urban development negatively. For the reason that condominium shares the largest proportion among the patterns of community, this research uses the households living in condominiums as the research population. House price to income ratio of family surveyed in this research is used to construct family housing affordability index. The property and level on quantile housing affordability affected from housing characteristics and neighborhood features are investigated with quantile regression model. In this research, the ordinary least square method is applied to estimate the independent variables coefficients and the demonstrative result shows that the multicollinearity existed among independent variables and it fails to explain the variables’ heterogeneities. Afterwards dropping variables with high value of Variance Inflation Factor and operating four inter- quantile regression equations, this research finds that the households with the highest level of housing affordability among both ends of housing affordability had better to select a house in either North District or Shang Shan District with large “dwelling area”, high “floor level”, and more “sanitary and health equipment” to live than the households with the lowest level of housing affordability selected the house with the same characteristics as the ones of the households with the highest level of housing affordability.

Housing is the most important and expensive economic goods for the individual to consume. When the gap between house price and family income has been widened would not only increase family affordability but also influence urban development negatively. For the reason that condominium shares the largest proportion among the patterns of community, this research uses the households living in condominiums as the research population. House price to income ratio of family surveyed in this research is used to construct family housing affordability index. The property and level on quantile housing affordability affected from housing characteristics and neighborhood features are investigated with quantile regression model. In this research, the ordinary least square method is applied to estimate the independent variables coefficients and the demonstrative result shows that the multicollinearity existed among independent variables and it fails to explain the variables’ heterogeneities. Afterwards dropping variables with high value of Variance Inflation Factor and operating four inter- quantile regression equations, this research finds that the households with the highest level of housing affordability among both ends of housing affordability had better to select a house in either North District or Shang Shan District with large “dwelling area”, high “floor level”, and more “sanitary and health equipment” to live than the households with the lowest level of housing affordability selected the house with the same characteristics as the ones of the households with the highest level of housing affordability.