Author: Wenzheng LiPublisher:
ISBN:
Category :
Languages : en
Pages : 194
Book Description
Using 16126 second-hand commercial housing samples collecting by web-crawl program in Beijing, China, this study develops traditional hedonic price model and spatial econometric models to quantitatively evaluate influential factors of housing prices in terms of housing structure, environmental and locational factors. Results show that locational factors, such as accessibility to subway, CBDs and schools, outperform other non-spatial factors as 6 locational variables explain 64% of variance. Most of housing structure variables have expected sign except floor position which manifests Beijing customers prefer middle floor. This finding does not agree with the evidence from southern Chinese cities where high floor charges higher price. We convince that this idiosyncrasy of Beijing reflects customers’ concern to safety during disasters such as earthquake and fire. Results also demonstrate customers’ desire to higher green ratio within community and closer distance to parks. We develop a feasible methodology for acquiring accurate NDVI in urban environment and test its validity in Beijing for the first time. We also demonstrate the existence of spatial error and spatial lag dependence in Beijing housing price using Lagrange Multiplier and successfully eliminate spatial dependence using spatial error (SEM) and spatial lag Model (SLM). The comparison of OLS and SEM suggests that OLS has overestimated the capitalization of housing structure and educational factors, whereas underestimated the effects of environmental, transportation and commute factors. The significant Rho (74.88%) in SLM model demonstrates the spatial spill-over effects caused by the interaction of nearby housing price and this effect is greater than other Chinese cities, such as Chengdu (Rho=19.2%) and Hangzhou (Rho=45.7%). This phenomenon can be partially explained by the potential speculation and booming of Beijing housing market. Key Words: web-crawl program; hedonic price model; spatial regression; spatial autocorrelation; spatial spill-over effect.
