This paper uses data from the Medical Expenditure Panel Survey to estimate the effect of COBRA on health care utilization among a sample of individuals who experience employment separation. The empirical specification employs a structural simultaneous equations model of insurance choice and utilization that is estimated by Maximum Simulated Likelihood. Results indicate that employment separators who elect COBRA appear to consume more health care compared to individuals who become temporarily uninsured. In addition, results do not indicate adverse selection into COBRA. Although COBRA enrollees consume more health care than temporary insurance losers, election appears to exhibit favorable selection with respect to physician utilization.
©2011 Walter de Gruyter GmbH & Co. KG, Berlin/Boston