Analysis of Factors Influencing Public Budget Revenue in Nanjing Based on LASSO Regression

Abstract

Based on the data of Nanjing Statistical Yearbook from 2008 to 2020, the influencing factors of public budget income in Nanjing were investigated by using LASSO method for screening and analysis. The evaluation index system of the study covers the annual per capita disposable income of urban households in Nanjing, the consumer price index of urban residents, etc., with a total of 12 explanatory variables. The study adopts a cross-validation method to adjust the values of the parameters, and uses the LASSO regression model to finally screen out three main influencing factors. According to the estimated values of regression coefficients, the annual per capita disposable income of urban households, total retail sales of consumer goods and fiscal expenditure are positively correlated with public budget revenue; based on these findings, suggestions for the future development of Nanjing are proposed.