Employer-sponsored health insurance and worker productivity

Health insurance for the working population of the United States is largely provided through employers primarily because of favorable income tax treatments and employment laws that encourage employers, large or small, to provide health insurance to their employees. In fact, a recent survey shows that employer-sponsored insurance covers over 50% of the non-elderly population of the United States, 57% of firms offered health benefits to employees, and 63% of workers accepted the coverage. This dissertation addresses several interesting questions raised by this situation. First, why do some firms offer health insurance to their employees whereas others do not? Second, what determines the number of health insurance plans offered among employers who do offer health insurance? Third, how do employers’ decisions concerning how many health insurance plans to offer influence the take-up decisions by employees and, therefore, variations in the extent and quality of health insurance coverage across industries and occupations? To provide at least a partial answer to these questions, this dissertation hypothesizes that employer-sponsored health insurance may affect worker productivity and, as a result, the different types of health insurance policies offered by employers and taken up by their workers. It then empirically investigates whether and how it does so. The first part of the dissertation addresses this issue from a theoretical perspective by investigating how employers decide the types and costs of the health insurance plans they offer, and their workers decisions whether to take up those plans. Following the existing literature, I examine these issues assuming that each worker’s demand for health insurance (and health status) is unobserved by the firm, that health insurance plans are priced competitively, and that workers do not move between employers. The contribution of the theory presented here is to add into this environment the assumption that a worker’s productivity is altered by the provision of employer-sponsored health insurance. The dissertation also explores certain variants of this theoretical model to investigate workers’ take-up decisions by modifying the framework so that workers can choose to remain uninsured. The second part of the dissertation tests whether and how employer-sponsored health insurance affects worker productivity in the real world by conducting an empirical analysis using data from the Medical Expenditure Panel Survey (MEPS). It does so by using a variable measuring health-related absenteeism due to sickness as a proxy for productivity and investigating the relationship between this productivity proxy variable and a variable indicating whether a worker is insured through his or her employer. To purge this relationship of the endogeneity effects that may result from selection, the dissertation utilizes employment-related spousal variables as instruments for the potentially endogenous employer-sponsored health insurance variable. The resulting (negative) estimates suggest that employer-health insurance may enhance worker productivity by reducing health-related absenteeism. The hypothesis that health insurance improves worker productivity helps explain why firms are willing to offer health insurance to their employees and bear part of the premium costs. The dissertation makes several contributions in the field of health economics. First, the dissertation brings about the novel idea that health insurance may affect productivity. Second, it theoretically examines the take-up decisions of workers by allowing them to remain uninsured. Third, the dissertation studies the firms’ optimal decisions and equilibrium conditions when workers require reservation wages. Fourth, it finds a statistically significant empirical relationship between a proxy for worker productivity (days missed for health reasons) and employer-sponsored health insurance