地区生产总值影响因素实证研究——基于多元回归分析

Empirical Study of Factors Affecting Regional Gross Domestic Product—A Multivariate Regression Analysis Approach

摘要: 本文运用多元回归方法，对2010年至2020年26个省GDP总量进行研究，选取了选取RD科技创新实验发展支出、CM社会消费品零售总额、ET能源消耗作为衡量地区生产总值GDP的指标体系，建立了多元回归方程，并对结果进行检验。结果显示：RD科技创新实验发展支出、CM社会消费品零售总额、ET能源消耗对GDP地区生产总值呈正相关关系。且影响地区生产总值GDP最大的是CM社会消费品零售总额，其次是ET能源消耗，应影响最小的是RD科技创新实验发展支出。本文还对模型进行了异方差检验和内生性处理，其中UE普通高等学校本科招生数作为Ln RD的工具变量，对Ln RD的贡献率最大。基于研究结论得出的政策建议有：促进消费品零售总额增长、优化能源消耗以及加强科技创新和研发，提升人才培养和科技创新。

Abstract: This article employs a multiple regression approach to investigate the total GDP of 26 provinces from 2010 to 2020. The selected indicators for measuring regional Gross Domestic Product (GDP) include RD technology innovation experimental development expenditure, CM total retail sales of consumer goods, and ET energy consumption. A multiple regression equation is established, and the results are tested. The findings indicate a positive correlation between RD technology innova-tion experimental development expenditure, CM total retail sales of consumer goods, ET energy consumption, and regional GDP. Furthermore, it is observed that CM total retail sales of consumer goods has the greatest impact on regional GDP, followed by ET energy consumption, while RD technology innovation experimental development expenditure has the least influence. This study also conducts tests for heteroscedasticity and endogeneity. Among these tests, undergraduate en-rollment in regular higher education institutions (UE) is employed as an instrumental variable for Ln RD, showing the highest contribution to Ln RD. Based on the research conclusions, policy rec-ommendations are as follows: promote the growth of total retail sales of consumer goods, optimize energy consumption, strengthen technology innovation and research and development, and enhance talent development and technological innovation.