Research On Real Estate Tax Base Evaluation Method Based On Spatial Characteristics

Real estate tax base assessment is the basic link of the entire real estate tax levy process,affecting the efficiency and fairness of the comprehensive tax levy,so academic research on real estate tax base assessment whose importance can't be underestimated has been ongoing.Compared with the traditional market method,cost method and income method,the real estate tax base mass appraisal is distinguished by virtue of its scientific model,the high efficiency of large sample regression and the accuracy and fairness of the results.The tax base mass appraisal generally uses the Hedonic price model for regression analysis,but the Hedonic price model follows the assumption that the samples are independent of each other,ignoring the spatial dependence,so there is a possibility of biased results.The spatial autoregressive model and the spatial error model can fully take the spatial dependence into account and correct the evaluation results of the traditional Hedonic price model to improve its accuracy.This article analyzes the tax base assessment value types,assessment objects,assessment subjects,assessment period,and tax base preferences.This article selects 823 ordinary second-hand housing data from Lixia District,Jinan City,Shandong Province to apply the Hedonic price model.20 characteristic variables were selected and divided into three levels of architectural characteristics,neighborhood characteristics and location characteristics for large sample regression.First,use the second-hand house list data in the "Fang Tianxia" website and the observed values of Baidu map to quantify the feature variables,and set the "road grade","whether it is a key school district","whether there is a large shopping mall" and "whether there is a park" as the dummy variable,then construct the Hedonic price model and perform regression analysis using linear function form and semi-logarithmic function form respectively.After comparing the goodness of fit and the significance of characteristic variables,finally this article selects the semi-logarithmic function form and obtains 17 characteristic variables that have a significant impact on housing prices,including the building area,building age,total number of floors,decoration status,number of rooms,accessibility of Quancheng Square,bus traffic status,road grade,property management,greening rate,floor area ratio,whether there are large shopping malls,number of schools,whether it is key school districts,whether there are parks and the number of hospitals and vegetable markets.The results of calculating the Moran's I,the Geary's C and the Getis-Ord's G all show that the sample has significant spatial dependence,so the spatial autoregressive model and the spatial error model are applied to modify the traditional Hedonic price model.This article finds that the spatial autoregressive model is superior to the spatial error model and traditional Hedonic price model.In order to further analyze the accuracy and scientificity of the assessment results,this article uses the three indicators of the ratio analysis test developed by the International Association of Assessing Officers:Median Ratio,Coefficient of Discreteness,and Price Related difference.All three indicators indicate that considering spatial factors can improve the accuracy of the model.