Empirical Essays in Health Economics

Thumbnail Image

Publication or External Link






Chapter 1: The Economic Effects of Medical Innovations: Vioxx and Labor Supply

Despite dramatic improvements in medical technology over time, comparatively little attention has been paid to the effect of these innovations on economic outcomes. This study uses seven overlapping panels of the Medical Expenditure Panel Survey to estimate the labor supply effects of Vioxx--a selective non-steroidal anti-inflammatory drug intended for individuals with chronic pain. Fixed effect estimates analyzing the effect of individuals choosing to start and stop taking Vioxx indicates that use of the drug is associated with greater labor force participation. This paper also exploits the removal of Vioxx from the market in 2004 as an exogenous source of variation in utilization. Both methods show that Vioxx use is associated with statistically and economically significant increases in the labor supply of near elderly individuals with joint conditions. An effect is also found for usual weekly hours worked. There is no labor supply effect for other expensive medications used for chronic conditions, suggesting that a desire to work in order to obtain employer-provided health insurance is not driving the estimated effect of Vioxx. These results suggest a role for improving medical technology in explaining recent increases in the labor force participation rate of older workers.

Chapter 2: The Effect of In-Utero Conditions on Long Term Health: Evidence from the 1918 Spanish Flu Pandemic

The fetal origins hypothesis posits that in-utero stress increases the incidence of chronic conditions later in life. Utilizing 21 years of National Health Interview Survey data, this study estimates the health effect of in-utero exposure to the 1918 Spanish Flu pandemic. Exploiting the fact that people were exposed to the flu at different points during fetal development, the model tests precise predictions from the medical literature about when exposure to in utero insults should damage organs later in life. The pattern of results demonstrates the necessity of using a short duration event as a source of variation in fetal conditions and helps explain previously mixed evidence regarding the fetal origins hypothesis.