Fixed-term contracts as sorting mechanisms : evidence from job durations in West Germany

We estimate the effect of initial episodes under fixed-term contracts (FTCs) on job duration in the further course of the employment spell, using data from the German Socio-Economic Panel (SOEP) from 1985 to 2002. Using a statistical matching approach, we find that job exit rates are initially much higher if the employment spell began with an FTC. However, exit rates fall below those of comparable spells spent entirely in permanent employment after a few years time. This suggests that FTCs accelerate a sorting process and that they may at least to some part be understood as prolonged probationary periods. Strikingly, the probability of long-term employment of more than five years duration is not lower in spells that are initially concluded as FTCs. Hence, the sorting processes taking place in both forms of contracts seem to be of similar intensity.