Reservation wages and unemployment duration

Based on the GSOEP (2000) for Western Germany, the paper deals with the influence of reported reservation wages on unemployment duration. The selectivity in the reservation wage observations due to only one interview per year in the GSOEP is included in the model by Heckman-correction, and the possibility of censored spells is accounted for by a Hazard-rate analysis with semiparametric unobserved heterogeneity. The results show that only by correction for the selectivity bias the theoretically expected positive effect of reservation wages on the unemployment duration can be found; it then occurs at a considerable level.