Identifying the determinants of housing prices in China using spatial regression and the geographical detector technique
Yang Wang,Shaojian Wang,Guangdong Li,Hongou Zhang,Lixia Jin,Yongxian Su,Kangmin Wu
DOI: https://doi.org/10.1016/j.apgeog.2016.12.003
IF: 4.732
2017-02-01
Applied Geography
Abstract:This study analyzed the direction and strength of the association between housing prices and their potential determinants in China, from a tripartite perspective that takes into account housing demand, housing supply, and the housing market. A data set made up of county-level housing prices and selected factors was constructed for the year 2014, and spatial regression and geographical detector technique were estimated. The results of the study indicate that the housing prices of Chinese counties are heavily influenced by the administrative level of the county in question. On the basis of results obtained using Moran's I, the study revealed the presence of significant spatial autocorrelation (or spatial agglomeration) in the data. Using spatial regression techniques, the study identifies the positive effect exerted by the proportion of renters, floating population, wage level, the cost of land, the housing market and city service level on housing prices, and the negative influence exerted by living space. The geographical detector technique revealed marked differences in the relative influence, as well as the strength of association, of the seven factors in relation to housing prices. The cost of land had a greater influence on housing prices than other factors. We argue that a better understanding of the determinants of housing prices in China at the county level will help Chinese policymakers to formulate more detailed and geographically specific housing policies.
geography