考虑多重共线性影响的中国财政收入实证研究  

An Empirical Study of China’s Fiscal Revenue with the Effect of Multicollinearity

作　　者：李盈 LI Ying(Jinan University Guangzhou,Guangdong 510632)

机构地区：[1]暨南大学,广东广州510632

出　　处：《中国商论》2023年第21期95-99,共5页China Journal of Commerce

摘　　要：作为财政收支平衡的一部分,全国财政收入保障基层政府正常运转、保障各项民生基本支出,因此密切关注全国财政收入的变化及其影响因素对财政部门制定政策、推动社会经济发展具有重要意义。本文以1996—2022年全国财政收入数据为研究对象,主要研究全国财政收入及其影响因素,从解决多重共线性问题的角度出发,对比多种回归方法以得到更准确、更有代表性的模型及更为稳定的参数估计、预测全国财政收入。结果显示,偏最小二乘估计法可以更准确地估计参数,全国财政收入主要受全国财政支出、社会消费品零售总额、居民消费价格指数和全社会固定资产投资额显著正向影响。As a part of fiscal balance,the national fiscal revenue guarantees the normal operation of the local government and the basic expenditures of people’s livelihood,so it is important to pay close attention to the changes in the national ffscal revenue and its infl uencing factors for the ffnancial department to formulate policies and promote social and economic development.This paper takes the national ffscal revenue data from 1996 to 2022 as the research object,mainly studies the national ffscal revenue and its infl uencing factors,and from the perspective of solving the problem of multicollinearity,compares various regression methods to get a more accurate and representative model and more stable parameter estimation and prediction of the national ffscal revenue.The results show that the partial least squares estimation method can estimate the parameters more accurately,and the national ffscal revenue is mainly affected by the national fiscal expenditure,the total retail sales of consumer goods,the consumer price index and the total investment in fixed assets in the whole society in a signiffcant positive way.

关 键 词：全国财政收入 多重共线性 参数估计 偏最小二乘估计法 政府财政健康 

分 类 号：F812.4[经济管理—财政学]