生命价值、职业伤害成本低估与安全事故

Value of Statistical Life,Undervalued Cost of Occupational Injuries and Production Accidents

本文立足于生命价值理论,结合享乐主义工资模型,采用中国家庭动态追踪调查(2010)与《安全生产统计年鉴(2009年)》的匹配数据,测算了2010年中国劳动者统计学生命价值(VSL)区间,并与近年我国企业实际负担的职业伤害成本进行比较。研究发现:我国劳动者的统计学生命价值与计量模型选取、变量内生性控制以及劳动者个体收入特征等有关;无论采用参数方法(线性或非线性)、半参数方法或者非参边界分析方法,均表明我国劳动者的统计学生命价值要远远高于因工死亡的实际补偿。职业伤害成本的低估引发了现阶段的不可承受之重,"外在于全社会共同承担"的职业伤害成本要远高于企业承担的实际成本。据此,本文建议:动态调整我国劳动者因工死亡的补偿标准,将现有"维持劳动者及其家属的一定生活水平"的补偿目标转化为"补偿性提升劳动者及其家属的生活水平",通过提升劳动者因工死亡的标准倒逼我国当前的安全生产改革。

Worker safety and occupational injuries are critical issues in government regulation policies in China.If China s workplace safety policy is to be clear and precise,evaluating and measuring the cost of occupational injuries,that is,assigning the value of occupational injuries,are key to workplace safety policy design in China.Based on comparing several methods of calculating the cost of occupational injury-life value theory,the compensation practice of Chinese laborers occupational injuries,and the mainstream approach to measuring value of statistical life(VSL)in current research-this paper chooses a hedonic model to evaluate the cost of Chinese laborers occupational injuries.Using data from China Family Panel Studies(CFPS)(2010)and The Safe Production Statistics Yearbook(2009),this paper proposes a VSL regression model and obtains the VSL of Chinese laborers in 2010.If the relation between lethal risk and wage is a quadratic curve,the estimate result shows the laborers VSL is RMB51,298,200.If the relation between lethal risk and wage is a linear curve,the estimate result shows the laborers VSL is RMB3,573,500.Furthermore,we propose a quantile regression to simultaneously explore the result of VSL,and find that the VSL of laborers in the 50%quantile is RMB51,323.To eliminate heterogeneity and endogeneity issues,we introduce the non-parametric boundary analysis method and apply the MTR-MTS joint hypothesis,and find the workers VSL to be between RMB12,750 and RMB41,731.No matter which measurement method is used,the value of statistical life is much higher than the current compensation standards in China.This discrepancy has two serious consequences.One is that the cost of occupational injuries is transferred from enterprises to laborers in China.Therefore,production accidents occur frequently because there is no significant impact on enterprises.The other consequence is that the laborer s value of statistical life is far higher than the cost that an enterprise can afford in China.Thus,this paper suggests that compensation standards for workers death or injuries should be adjusted over time.In particular,the compensation standards should be transformed from maintaining the living standards of workers and their relatives into promoting their living standards.Doing so will give high priority to safer production and push the government to develop safer production standards in China.This paper suggests the following marginal improvements.First,it attempts to combine risk theory of production and life value theory with Chinese practice.Based on theory of the value of statistical life,this paper measures the social cost and affordable capability of Chinese enterprises.It enriches the research related to social ethics,corporate ethics,and social welfare distribution.Second,this paper further considers endogeneity problems under limited information on occupational injury data.Through discussing two types of individual ability(cognitive and non-cognitive capabilities),the measurement of occupational socioeconomic status,and the degree of occupational risk,this paper considers the endogeneity problems that arise from ignorance of laborers abilities and the heterogeneity of risk preferences and production conditions(Hwang et al.,1992).In addition,this paper uses the control function method proposed by Wooldridge(2009)to examine the value of statistical life under different conditions of income level distribution,which effectively controls endogeneity problems and expands on the work of Qin et al.(2010)and Qian(2011).Third,using the non-parametric boundary analysis method proposed by Manski(1990,1997,2000),this paper measures the range of the value of statistical life.It reduces the bias caused by the lack of instrumental variables and also narrows the differences in the estimated value of statistical life obtained by various methods.