Research On Urban Residential Land Price Evaluation Based On Spatial Durbin Model

In the context of the report of the 19 th National Congress of the Communist Party of China,the adjustment and introduction of a series of new policies have made the development of China's real estate industry also in a "new normal".Among them,standardizing the land transaction market is particularly prominent,and how to objectively evaluate land prices is its core issue.The problem of land price has caused extensive research in the academic community in recent years,and experts and scholars have tried to construct a more reasonable and effective evaluation model.Most domestic researchers use Hedonic Price Model(Hedonic Price Model,HPM)to analyze the characteristic variables that affect land prices and their degree of influence.However,because the data samples of each study are independent individuals as one of the assumptions of the model,and often only consider the attributes of the sample data itself,and ignore the objective interaction between the data samples.Spatial effect,so often produce some wrong evaluation results and evaluation index deviation.Compared with domestic studies,foreign studies have introduced a spatial econometric model to quantitatively analyze the spatial effects between sample data,and have achieved an ideal land price evaluation model.Therefore,on the basis of absorbing the more mature domestic characteristic price theory,this article introduces the Spatial Hedonic Price Model,analyzes the factors that affect the land price,reveals the nature of the impact and the spatial effect between the sample data,in order to obtain the ideal land price evaluation model.First of all,in the evaluation of land prices,the selection and quantification of characteristic variables is one of the difficulties in this paper.In order to select feature factors in a more reasonable and practical way,this paper summarizes a wide range of domestic and foreign research literatures,and consults with land price evaluation experts and practitioners,and finally selects 3 major categories and 12 minor feature variables.The selected feature variables are classified from three aspects: location features,neighborhood features and individual features,so as to perform more effective quantitative analysis.Secondly,using the relevant data of 200 residential land sold in the main urban area of Chongqing from 2014 to 2018 as samples,the Exploratory Spatial Data Analysis tool(ESDA)and Pan-Kriging Interpolation Method in ArcGIS were used to analyze the data structure of the sample data.Autocorrelation analysis with space.After that,based on Hedonic Price Model,Stata data analysis software was used to perform regression analysis on the characteristic variables.At the same time,the Spatial Lag Model(SLM),Spatial Error Model(SEM)and Spatial Durbin Model(SDM)are introduced,and the MATLAB mathematical software is used to analyze the spatial effect and test the results.Finally,using the ratio analysis method,the four models are compared and analyzed,and the spatial characteristic price model with the best interpretation ability for the data sample is selected to analyze the characteristic factors and spatial effects that affect the residential land price of Chongqing.The results of this study are as follows:(1)The global Moran's I index =0.24209;The local Moran's I index=0.278,both of which passed the significance test at the 10% significance level.This result indicates that there is a certain degree of spatial effect on the spatial distribution of residential land prices in Chongqing,and there are "high-high aggregation" and "low-low aggregation" phenomena.(2)The Spatial Durbin Model has the best ability to interpret sample data.It not only can better reflect the effect of feature variables on land prices and Chengdu,but also can analyze the spatial effect of sample data.The fitting degree of the model is highest.(3)The Spatial Durbin Model(SDM)analysis of the residential land price in the main urban area of Chongqing City,the evaluation results and evaluation errors are better than the other three evaluation models,of which the 2018 evaluation results are taken as an example The accuracy of SDM is 8.52%,4.82% and 2.93% higher than the accuracy of Hedonic Price Model(HPM),Spatial Lag Model(SLM)and Spatial Error Model(SEM),respectively,and the Spatial Durbin Model(SDM)evaluation results The error from the residential land price in the main urban area of Chongqing in 2018 is 3.06%,which is less than 5% of the allowable range..