影响税收因素的多元回归分析及预测

Multiple Regression Analysis and Prediction of Factors Affecting Tax Revenue

为了分析经济变量中影响税收的主要因素,对我国税收收入进行预测,利用Eviews软件对1990—2016年相关经济因素统计数据进行分析,采用普通最小二乘法进行参数估计,利用多重性、异方差和自相关性对影响税收收入的因素进行分析,最终得到税收多元线性回归模型。该模型可以得到影响税收收入的显著因素,为税收收入的预测提供了依据。

The purpose of this paper is to analyze the factors affecting tax revenue in economic variables and predict the tax revenue in China. The Eviews software was used to analyze the statistical data of related economic factors from 1990 to 2016, and the ordinary least square was applied to estimate the parameters, and the factors affecting tax revenue were analyzed from the perspectives of multiplicity, heteroscedasticity and autocorrelation, and finally, the taxation multiple linear regression model was obtained. By using this model, the factors that significantly affect the tax revenue can be obtained, which will provide the basis for the prediction of tax revenue.