抑郁倾向对劳动收入的影响——基于LASSO回归的工具变量识别

The Effect of Depression on Labor Income:Instrumental Variable Identification Based on LASSO Regression

本文采用2010年中国家庭调查数据实证检验了居民抑郁倾向对劳动收入的影响,结果表明抑郁倾向能显著降低居民的劳动收入,对个人就业产生负向影响。为克服个人精神状况与收入之间的反向因果关系所导致的内生性问题,本文使用机器学习方法筛选工具变量,使得工具变量估计结果相比于传统方法更加精确,进一步验证了抑郁倾向对个人劳动收入的负向影响。异质性分析表明男性个体、位于农村地区的个体、青年个体受到抑郁倾向的影响更为严重,收入下降程度更为显著,而受教育程度的提高则能够在一定程度上减少抑郁倾向给收入带来的负面影响。机制分析表明,抑郁倾向通过抑制个人的社交能力与社交意愿、损害个人的劳动能力等方式来影响个人的劳动参与和劳动收入。

This paper uses the data from China Family Panel Studies 2010(CFPS 2010)to empirically test the impact of residents'depression tendency on labor income.The results show that depression tendency can significantly reduce residents'labor income,and has a negative impact on individual employment.In order to solve the endogenous problem caused by the reverse causal relationship between personal mental status and income,this paper uses machine learning method to screen instrumental variables,which makes the estimation results of instrumental variables more accurate than traditional methods,and further verifies the negative impact of depression tendency on personal labor income.Heterogeneity analysis shows that male individuals,individuals in rural areas and young individuals are more seriously affected by depression tendency,and the decline of income is more significant.The improvement of education level can reduce the negative impact of depression tendency on income to a certain extent.Mechanism analysis shows that depressive tendency affects personal labor participation and income by inhibiting personal social ability and willingness and damaging personal labor ability.