Memorandum, Department of Economics, University of Oslo 1999,14
We investigate how unemployment exit probabilities are affected by economic incentives, spell duration and macroeconomic conditions. Building on a database containing all registered unemployment spells in Norway in 1989-1998, we apply an econometric model in which exit probabilities vary freely over spell durations as well as calendar time. We find that i) the replacement ratio affects the exit rate negatively and that this effect is counter-cyclical and stable over spell durations; ii) the exit rate rises just prior to benefit exhaustion; and iii) the predicted exit rate declines as the spell lengthens, but this may be explained by unobserved heterogeneity.
Unemployment spells business cycles unemployment income semiparametric duration analysis