1 论文标题
The Role of Health Status on Income in China
2 作者信息
Xie, Ruizhi
Awokuse, Titus O.
3 出处和链接(比如,NBER working paper No.11000)
Selected Paper prepared for presentation at the Agricultural & Applied Economics
Association’s 2013 AAEA & CAES Joint Annual Meeting, Washington, DC, August 4-6, 2013.
4 摘要
The economic benefits from improving health status are obvious, yet there remains a lack of agreement on how to quantify and compare the benefits and the accompanied costs. In our study, we extend Liu et al. (2008)’s study on the role of health status on income in China and examine whether their conclusions still hold under new specifications and in a broader time horizon. Our results show a larger impact of health status on income after replacing household income with individual income. We find this effect becomes even more pronounced in the 2000s. Moreover, our results show an inverted-U relationship between age and income, which is an improvement over Liu et al. (2008)’s work and is in line with other empirical studies. By admitting the endogeneity issue, we find the impact of health status becomes even larger after instrumenting health status. The results of GMM estimation, which allows for efficient estimation under heteroskedasticity of unknown form, are consistent the IV estimations.
Health economics这个方向的理论文章比较少,Grossman(1972)是奠基性的文章。该领域的文章主要都是致力于在因果意义上研究SES和Health Status之间的相互影响。最近几年James Heckman花了很多精力讨论幼年期和儿童期健康同成年后健康和收入之间的逻辑关系,这些研究既有理论研究也有实证研究,相辅相成,感兴趣的话可以找来看看。我觉得这个领域理论研究少的原因也在于问题本身更偏向于政策导向,理论内涵较少,研究者关心的通常是即给予某个健康干预政策或者SES干预政策,SES或健康会有何反应。健康资本和人力资本的关系文献中也有讨论,比如儿童期健康好,才能更好的接受教育,教育水平高,就更容易选择健康的生活方式等等。另外,你说的“过劳死”这种教育同健康负相关的关系,这里面有许多别的因素, 比如工作的强度,环境等等,而文献中讲教育水平高导致健康好的原因包括收入高,健康知识更多,对健康的自我控制更有效,我觉得两者并不矛盾,因为有不同的作用渠道。个人观点,仅供参考。
寻找健康的工具变量确实很困难。现有文献利用了大饥荒,大流感的爆发来做height,weight这种健康度量的工具变量;还有一些则直接做的政策实验,直接给营养不良的个体以发放药丸的形式补铁,或者补碘,主要是在非洲和东南亚;还有少数研究利用一些ZF的健康干预政策,比如产假的长短;我印象中还有一篇很有趣的文章,利用穆斯林的斋月来做工具变量。无论如何,在这个领域,一个好的IV,基本上就意味着一项好的实证研究。除了采用工具变量这种识别策略外,James P. Smith(1999,2004,2006)觉得可以利用health onset来作为比较外生的健康度量。背后的逻辑是,一旦控制了基期的SES和各种健康指标,未来健康的发生时比较外生的,因为没有人可以准确预期自己何时生病;经济学季刊最近有一篇文章用食盐的摄入来做高血压的工具变量。