SSE/EFI Working Paper Series in Economics and Finance 465
We estimate a labor supply model on a random sample of Swedish male and female blue collar workers to study the effect of economic incentives on work absence behavior. We observe work absence for each day during 1990 and 1991 for each worker in the sample. We use non-parametric (Kaplan-Meier) techniques; semi-parametric stratified models, where individual effects are removed; and fully prametric Cox regression models, where observed characteristics are used to control for heterogeneity. An exogenous change in the cost of being absent due to a reform of the sickness insurance, which took place during the time period covered by the data, is used as identifying information. The empirical analysis is focused on explaining gender differences in work absence behavior. We find that about one third of this difference in our sample can be attributed to differences in costs of being absent.