Doctoral Dissertation Research: Labor Supply Decisions and Fatality Risk

在做出劳动力供给决策时，工人经常要权衡有工资和无工资的工作特征。补偿差异理论为解释工人如何做出这些决策提供了一个框架，但该框架的概括性加上经验证据的不足，使研究人员和政策制定者对非工资工作特征的价格如何通过劳动力供给决策受到工人偏好的影响缺乏理解。本项目的目的是使用一种新的经验方法来推进对死亡风险补偿差异的理解。实证研究集中在阿拉斯加白令海和阿留申群岛地区的商业捕鱼和捕蟹业，在那里，从事同样工作的同一船员在一年中的不同季节面临着工资和死亡风险的变化。这种不同工作期间的独特差异，使我们能够规避以往经验估计中许多长期存在的偏见来源，这些偏见来自于未被观察到的静态工人和企业特征，对完美劳动力市场的不合理假设，以及常见的测量误差来源。此外，这种在工人-公司层面上使用变化的方法提供了关于补偿差异如何随着死亡风险水平的变化而变化的信息。估计生命统计价值的文献严重依赖于补偿差异与死亡风险线性变化的假设；然而，这一假设尚未得到验证。本项目将在观察到的死亡风险范围内正式检验这种线性假设，以评估其在经验观察到的风险水平之外外推估计的有效性，这对于估计生命的统计价值是必要的。改进对死亡风险补偿性差异和人的生命的统计价值的估计，将有助于其决定影响预期寿命的决策者，例如在工作场所安全政策、卫生政策和环境政策方面。了解劳动力市场中的个人如何做出重要决定可以指导公共政策的制定。

When making labor supply decisions, workers frequently trade off wages and non-wage job characteristics. The theory of compensating differentials offers a framework for explaining how workers make many of these decisions, but the generality of the framework combined with insufficient empirical evidence have provided researchers and policymakers with a poor understanding of how the prices of non-wage job characteristics are affected by workers' preferences through labor supply decisions. The purpose of this project is to use a new empirical approach to advance the understanding of compensating differentials for fatality risk. The empirical setting focuses on the commercial fishing and crabbing industry in the Bering Sea and Aleutian Islands region of Alaska, where the same crewmembers performing the same jobs face variation in both wages and fatality risk in different seasons of the year. This unique variation between job spells makes it possible to circumvent many of the chronic sources of bias in previous empirical estimates arising from unobserved static worker and firm characteristics, unjustified assumptions about perfect labor markets, and from common sources of measurement error.In addition, this approach of using variation within the worker-firm level provides information on how the compensating differential varies as the level of fatality risk varies. The literature estimating the statistical value of life has relied heavily on the assumption that compensating differentials vary linearly with fatality risk; however, this assumption has not been tested. This project will formally test this linearity assumption over the observed range of fatality risk in order to assess its validity for extrapolating estimates beyond empirically observable risk levels, which is necessary for estimating the statistical value of life. Improving estimates of the compensating differential for fatality risk and the statistical value of human life will assist policymakers whose decisions affect life expectancies, as in the cases of workplace safety policies, health policies, and environmental policies. Understanding how individuals in the labor market make vital decisions can guide public policy decisions.