教育真的可以影响健康吗？——来自中国1986年义务教育法实施的证据  

Does Education Really Affect Adult Health? Evidence from China’s Implementation of the 1986 Compulsory Schooling Law

作　　者：郭四维 张明昂 曹静[1] Siwei Guo;Ming-ang Zhang;Jing Cao(School of Economics and Management,Tsinghua University)

机构地区：[1]清华大学经济管理学院经济系

出　　处：《经济学报》2019年第3期148-187,共40页China Journal of Economics

摘　　要：教育能否真的影响健康一直是学术界争论的话题,传统观点认为教育能够促进健康水平的提高,但近来随着计量方法的改善,这一结论已受到挑战。本文借助我国1986年实施义务教育法这一历史事件,使用中国综合社会调查(CGSS)数据和模糊断点回归(RD)方法研究初中阶段教育对居民健康水平的影响。回归结果表明,无论是主观汇报的身体健康、心理健康、健康问题还是客观的身体形态指标"肥胖""体重过重""体重过低",都不受教育的影响,且结论在区分样本性别、城乡及改变回归方程的阶数和带宽后保持稳健。本文结论与现有文献中使用OLS、IV方法估计得出的结论不同,主要原因在于本文使用的RD、RD-DD等计量方法在很大程度上克服了遗漏变量偏误。同时,通过进一步的机制分析我们发现教育并不影响除了对待医疗卫生正确态度之外的其他健康行为指标,且高中阶段的教育对健康也没有显著影响,本文结论具有较强一般性。Whether the well-documented positive correlation between education and health outcomes is causal remains debatable in literature.In this paper,we attempt to understand to what extent this relationship is causal using CGSS data and1986 compulsory schooling law that generated sharp across-cohort difference in middle school educational attainment.Using regression discontinuity methods,we find that neither self-reported health outcomes(physical health,mental health,health problem)nor BMI-related outcomes(obesity,overweight,underweight)are affected by the reform,and the estimates are robust to different gender,rural or urban,and change of econometric equations.The main reason that our results are different from some existing literature using OLS and IV is due to the correction of the omitting bias.Also,we further find that the reform has no effect on adult health behavior,except the attitude to health care,and our conclusion can be applied to higher education.

关 键 词：居民健康 义务教育法 模糊断点回归 

分 类 号：G63[文化科学—教育学]