Moral hazard and sickness insurance

We study if the replacement level in the Swedish national sickness insurance, which replaces foregone earnings due to temporary illnesses, affects work absence behavior. We use micro data and estimate the effects of a major reform, whereby the replacement level during the first 90 days in each absence spell was reduced, on work attendance. To separate out the effect of the reform from any trend in work absence, we distinguish between the implications on the incidence of work absence (i.e., the frequency of spells) and the duration of the spells. We also use a regression-discontinuity approach to estimate the effects on the prevalence of work absence. Finally, we estimate elasticities with respect to the replacement level in the sickness insurance. (c) 2005 Elsevier B.V. All rights reserved.