摘要
基于60篇教育与中国经济增长实证研究中626个回归方程的元分析发现,教育对中国经济增长的作用估计受到不同研究特征的影响。尤其是,"显著效应"的实证结果受到估计方法、教育变量和经济增长变量的设定形式、教育变量的测量指标、样本起始年份、区域和控制变量选择等研究特征的显著影响。使用变量增长率形式、运用工具变量法、使用入学率和教育经费数据估计会降低显著性效应的概率;样本区域覆盖面越小、样本起始年份越早、使用受教育年限数据估计会增加显著性效应的概率。漏斗不对称检验显示样本文献存在显著的正向偏误,即回归结果越显著的研究越有可能被发表。因此,在分析教育对我国经济增长的作用时,必须考虑研究之间的异质性和可能存在的发表偏误。未来应重点关注的是,改进测量和估计方法,通过更严谨的研究设计精确识别教育对我国经济增长的因果效应;结合中国经济发展的阶段特征,分行业和部门讨论教育的作用,同时拓展非线性模型以考察教育的收益递增的外部效应;持续关注对教育质量的监测及其对经济增长的影响,为高质量教育体系的建设提供科学的理论支持。
This study,based on a meta-analysis of 626 regression equations from 60 empirical research papers involving the effect of education on China’s economic growth,finds that effect of education on China’s economic growth is affected by different research features. In particular,the empirical results in the "significant effect" are significantly affected by such research features as the methods of estimation,the educational variables and economic growth variables set in estimation models,the indicators used to measure educational variables,the starting year of sampling,and the choice of region and control variables. The use of the growth rate of variables,instrumental variable estimation,the enrollment rate,and the data of educational expenditure can help reduce the probability of a significant effect;and a small coverage of area sampling,the early starting year of sampling,and the data on the years of education help to increase the probability of a significant effect. The test for funnel plot asymmetry shows a significantly positive publication bias,that is to say,the research outcomes with significant regression results are more likely to be published. Therefore,when analyzing the effect of education on economic growth,we need to consider the heterogeneity between the research outcomes and the potential publication biases. The future research needs to focus on the following:improving the methods of measurement and estimation to identify the causal effect of education on the economic growth more accurately;analyzing the effect of education on different industries and sectors according to the stages of China’s economic growth,and expanding the nonlinear model to explore the external effects of the increasing returns of education;and highlighting educational quality monitoring and its effect on economic growth to provide scientific theoretical support for the construction of a high-quality educational system.
作者
赵冉
杜育红
朱倩文
Zhao Ran;Du Yuhong;Zhu Qianwen(Faculty of Education,Beijing Normal University,Beijing 100875)
出处
《教育研究》
CSSCI
北大核心
2022年第1期47-61,共15页
Educational Research
基金
国家社会科学基金2021年度后期资助暨优秀博士论文项目“教育在我国经济增长中的贡献及其作用机制研究”(项目编号:21FYB073)的研究成果。
关键词
教育
经济增长
元分析
发表偏误
education
economic growth
meta-analysis
publication bias
