Suchergebnisse

Abstract
We estimate a dynamic model of health that is rooted in "stress models" from Epidemiology. Health is determined by time-invariant endowments, permanent shocks, and transitory shocks. We estimate that the variance in health at age 60 ranges between 2.5 and five times its variance at age 25 depending on which demographic group we consider. We show that the stress model performs better than a simple alternative, the random effects Probit, particularly for less educated people.

We investigate the impact of exogenous income fluctuations on health using twenty years of data from the Panel Study of Income Dynamics using techniques from the literature on the estimation of dynamic panel data models. Contrary to much of the previous literature on health and socio-economic status, we find that, on average, adverse income shocks lead to a deterioration of health. These effects are most pronounced for working-aged men and are dominated by transitions into the very bottom of the earnings distribution. -- Gradient ; health ; dynamic panel data models ; recessions