A Tale of Two Medicare Transitions : Can Moral Hazard Influence Health Behaviors after a Health Shock?

We use longitudinal data from the Health and Retirement Study to investigate the effects of exogenous changes to financial risk exposure to medical spending. We focus on the transition to Medicare coverage at the age 65 threshold as well as the introduction of Medicare Part-D in 2006. Our identification strategy rests on the timing of a health shock as unknown to the individual. Utilizing individual fixed effects, we find evidence that those experiencing a health shock are less likely to smoke if the shock occurs during the pre-Medicaid years. Conversely, shocks occurring at age 65 or later, when insurance generosity (i.e., financial risk protection) is greater, tend to be much less influential on smoking status. Unsurprisingly, insurance status preceding age 65 also matters, as different pre-existing insurance and financial situations shape some of the impact of transitioning to the Medicare program. We also account for labor supply considerations in certain models, which may also occur around age 65 and perhaps effect tobacco consumption. Relatedly, we show having a shock in the post-Part-D period leads to worse health behaviors for Medicare-eligible individuals relative to a shock in the pre-2006 period