The distribution of unemployment duration in our equilibrium matching model with spell-dependent unemployment benefits displays a time-varying exit rate. Building on Semi-Markov processes, we translate these exit rates into an expression for the aggregate unemployment rate. Structural estimation using a German micro-data set (SOEP) allows us to discuss the effects of a recent unemployment benefit reform (Hartz IV). The reform reduced unemployment by only 0.3%. Contrary to general beliefs, we find that both employed and unemployed workers gain (the latter from an intertemporal perspective). The reason is the rise in the net wage caused by more vacancies per unemployed worker.