Research On The Residential Land Prices Based On Spatial Hedonic Model

In academic circles, the former research mostly adopt the hedonic price model(HPM) to explore the formation mechanism of land price. Assumption of the hedonic price model is each observed samples are independent, while neglecting the interaction effect to a certain extent the observation sample data of objective existence in the spatial distribution of(i.e., spatial effect). Therefore, this paper introduces spatial hedonic price model inquiring the residential land price factors and spatial effect.First of all, in this paper based on the reference of massive literature research, summarizes the factors influencing residential land price characteristics of 3 categories and 12 types of impact factors. Including the location characteristics : distance from the CBD, subway and bus route number; neighborhood characteristics : education facilities, living medical facilities and Park Plaza attractions; individual characteristics : parcel area, volume rate, green rate, building density, building height limit and the way of transfer. Secondly, the 102 residential lands public sale of main urban area in Chengdu between 2011-2014 about 5 years are selected as samples, trying to take the hedonic price theory as the foundation, from the view of spatial effect, and using spatial hedonic price model to explore the residential land price factors and spatial effect. Finally, Comparing estimation of test results of traditional hedonic price model(HPM), spatial lag model(SLM), spatial error model(SEM) and spatial durbin model(SDM); and because of strong ability to explain the observation sample, the spatial Durbin model was selected to analyze the influence factors of residential land price, spatial effect and the reasons for its formation.The conclusions of this study are as follows:①The residential land prices of Chengdu with global Moran’s I index=0.2452;local Moran’s I index =0.3550;and it satisfying the significant test at 10% level of significance. It shows that the residential land prices of Chengdu exist interaction effects to some extent in space, and also exist the "high-high aggregation" and "low-low aggregation" phenomenon.②In the spatial Durbin model, three spatial lag characteristic variables such as subway, educational facilities and public square attractions getting into the model, it illustrates a certain diffusion or agglomeration effect comes from the three variables on the adjacent residential land prices.③Spatial Durbin model has the best ability to interpret the observation sample. Spatial Durbin model can not only analyze the effect of related variables on the residential prices, also can explore the spatial effect of objective existence, and The fitting degree R2 of the model is the highest.