Zhanhan Yu

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




Alfonso Flores-Lagunes

Subject Categories

Economics | Social and Behavioral Sciences


This dissertation is comprised of three essays on labor and environmental economics. The first and third chapters study the question of whether environment affects worker productivity in the era of global warming, while the second chapter studies monopsony in the labor market and its contribution to gender pay gap. The first chapter investigates the effects of air pollution on work safety using novel data on work-related severe injuries and air pollution in the contiguous United States from 2015 through 2018. I focus on fine particulate matter (known as PM2.5), a primary air pollutant found to adversely impact human cognitive abilities and thus potentially affect the incidence of workplace accidents via biological channels. Credibly pinning down the causal effect of air pollution is challenging, because air pollutants are not randomly assigned across space or workplaces. To deal with the endogeneity of air pollution, I employ a quasi-experimental design, exploiting plausibly exogenous variation in PM2.5 driven by the changes of two commonly used instruments - rainfall and wind direction. I start by testing the validity of these instruments and provide statistical evidence that they violate the point identification assumptions. Then, I leverage partial identification strategies using the same instruments to estimate bounds on the effect of air pollution. The partial identification method replaces the exclusion restriction with weaker assumptions, exploiting the exogenous variability induced by the instrumental variables while accounting for their invalidity. In the second chapter, Alfonso Flores-Lagunes and I study another aspect of worker productivity and compensation. We investigate the existence of monopsony power in a highly-skilled labor market given by tenure-ranked faculty in public research universities in California, analyze differences in monopsony power by gender, and relate them to the observed gender pay gap. We collect and use publicly-available information on faculty salaries in the University of California system and merge it with information obtained online on faculty characteristics, career trajectories, and research productivity indicators. We infer the university-level labor supply elasticity by estimating the elasticity of separation. To deal with the endogeneity of the salary in the separation equation, we employ instrumental variables exploiting exogenous variation in salaries driven by changes in school revenues and salary scales. The final chapter examines the impact of rising temperatures on productivity in the transportation and logistics industries. In particular, Ying Shi and I measure productivity in the US airline industry by analyzing flight on-time performance on hotter days. We build a novel dataset that links the on-time performance data of flights departing from major commercial airports in the contiguous United States from 2004 through 2019 to daily meteorological data. High-frequency productivity and climate measures allow us to identify the role of heat stress by leveraging variation in temperatures over time and using a model augmented with a rich set of fixed effects. Moreover, the impact of high temperatures may be driven by several labor channels, including reductions in workers’ labor supply (extensive margin) and declines in on-the-job performance on hotter days (intensive margin). We utilize data from the American Time-Use Survey (ATUS) and find evidence supporting our hypothesis that heat affects airline productivity in part through these labor channels.


Open Access