An Analysis of the Change in Spatial Allocation of New Built Housing Price in Tainan City

• Main Authors: Guo-Shiuan Tzeng, 曾國軒
• Other Authors: Bor-Ming Hsieh
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/96840307984007593626

碩士 === 長榮大學 === 土地管理與開發學系碩士班 === 97 === The spatial allocation of new housing starts is the indicator of housing market conditions. The changes in spatial distribution of new housing starts present the changes in housing market development. This study primarily discusses the spatial autocorrelation and spatial clusters of new housing prices during 2005 to 2008. This study also compares the hedonic price model from the spatial autoregression models and analyzes whether the spatial autoregression model can improve the model’s goodness-of-fit and bias estimations? Regarding spatial dependence, the Moran’s Index presents significantly positive results in all four years which indicates that new housing prices are significantly correlated to each other during the four years. Higher price areas were concentrated on Hu-mei area of Middle-East district, An-ping district and Hi-an road of North district, while lower price areas were concentrated on An-nan and South districts. However, there was a significant change in the concentration of higher price areas where there was shifted to Huwei Liao area in 2008. With convenient traffic conditions, sufficient public space and facilities and good schools, Huwie Lioa area has become the hottest area attracting not only the housing developers but also housing purchasers in recent years. By comparing the hedonic price model and spatial autoregression models, the results show that the spatial autoregression models have better goodness-of-fit, which can improve bias and inefficient estimations of new housing prices, even though there is vey small in difference between these models. It is probably influenced by the sample size and allocation and also influenced by a lack of important determinants in estimating new housing price models. The contribution of this study is primarily focused on exploring the changes in spatial allocation of new housing prices with a change in real estate market condition, as well as macro-economic conditions. In addition, our results can be as an indication to local housing market development and also can be compared to local housing markets in northern Taiwan.