Research On The Spatial Hedonic Approach For The Mass Appraisal Of Real Estate

The real estate mass appraisal has always been the focus of academic work.The mass appraisal method is a scientific and efficient method to confirm the value of real estate compared to traditional case valuation methods such as cost method,market method and income method.At present,real estate mass appraisal generally uses the hedonic price method for multiple regression analysis.However,the hedonic price method follows the standard econometrics hypothesis that the samples are independent of each other,and can only reflect the heterogeneity of the real estate itself,ignoring its spatial dependence,so that the application of the hedonic price model for mass appraisal may have a biased result.Spatial econometrics breaks this standard hypothesis,fully considers the spatial effect between real estate,and improves the problem of biasing the results of traditional hedonic price methods,which can improve the accuracy and fairness of mass appraisal results.Therefore,the application of spatial hedonic method to the mass appraisal of real estate has important theoretical and practical significance.Based on the average price of 314 second-hand housing residential communities in Yanta District,Xi'an City in August 2018,this paper selects the architectural features,location features and neighborhood features of each residential community as control variables,and introduces the characteristics of the school district as an independent variable.District real estate prices are subject to mass appraisal.Firstly,the GIS technology of Baidu map is used to quantify the geographic coordinates,architectural features,location features and neighborhood characteristics of the residential community,and the school district features are set as dummy variables(the primary school district is 1 and the non-school district is 0).A sample database of second-hand houses in Yanta District was established,and descriptive statistical analysis was performed on the samples using Stata15.1.Secondly,the traditional hedonic price model was constructed to estimate the least squares(OLS)in housing prices.The function of the hedonic price model has various forms.In this paper,before the model estimation,the regression results of various functional forms are compared,and the linear function form of the saliency and the best fitting effect is selected for regression.Thirdly,the spatial factor is introduced on the basis of the traditional hedonic price model to carry out maximum likelihood(ML)re-estimation of housing price by referring to the idea of spatial hedonic approach.Finally,model quality of 50 Yanta residential districts that did not participate in the model regression is used to examine the scientific and accuracy of the real estate mass appraisal method.Through empirical analysis of the real estate mass appraisal of Yanta District,the main conclusions are as follows:(1)Using the spatial hedonic method to estimate the real estate volume can improve the accuracy and fairness of the evaluation results.In the empirical analysis,through the comparison test of the traditional hedonic price model and the spatial econometrics model,it is found that both the statistical test index and the ratio analysis test index indicate that the application of the spatial hedonic method for mass appraisal is closer to the real value of real estate price.(2)The Moran's I index of real estate prices in Yanta District indicates that there is spatial autocorrelation in the real estate data,and the regression results of the optimal spatial lag model also prove that the housing price in the district has strong spatial autocorrelation.Therefore,it is necessary to incorporate spatial factors into the real estate mass appraisal model,in order to obtain more scientific and reasonable real estate evaluation prices.