The Effect of changing the Medicare Eligibility Age on the Health of the Near-Retirement Population [Paper | Slide]
Raising the eligibility age for Medicare, the third largest program in the federal budget, could lead to a large reduction in the federal budget deficit; however, the effect of this change on the welfare and the health of the near-retirement population is unclear. Using Health and Retirement Study (HRS) dataset, I measure the effect of a change in Medicare eligibility age on the welfare of the elderly population by estimating a dynamic discrete choice model of health and retirement that endogenizes health investment decisions. The empirical model allows for tracking the health behavior, labor supply and health status among the other key variables. Using Forward Simulation and Conditional Choice Probability estimator (CCP), I incorporate a large, multidimensional state space consisting fixed unobserved heterogeneity that serves as a measure to better identification of the effects. I find that labor supply, life expectancy, and mental health will be affected positively in response to an increase in the Medicare eligibility age. The welfare effect, however, is negative and there is some evidence of cost transfers from Medicare to the Social Security Program.
Measuring Health in HRS dataset using Bayesian Approach [Paper]
This paper provides a measure of the physical health for the near retirement people based on the binary responses of individuals to the sets of questions, regarding the objective aspects of their health. Using three sets of questions: Activities of daily living (ADL), Lower Body Mobility (LBM) and Upper Body Agility (UBA), I employ the Item Response Theory and using Bayesian approach estimate the individual distribution of health for each individual in the Health and Retirement Survey (HRS) sample of 2008. The Analysis of 17,217 sets of individual responses shows that LBM and UBA affect the combined health index (Full Index) more strongly and ADL has a weak effect on the health index. While the correlation between the full measure of functional limitation and LBM and UBA is 0.92 and 0.84 respectively, the correlation between functional limitation and ADL is only 0.5.
Work in Progress
1. Consistent Incorporation of Individual Level Characteristics into Models with Health Outcomes
Measured at an Aggregate Level (with Tom Mroz and Phillip Gustafson)
2. Causal Effect of Sleep on Health
3. The Effect of Involuntary Layoff on the Health of the Older Adults