Quantitative Analyse dynamischer nichtlinearer Panelmodelle
Dynamic nonlinear panel models are estimated on the first 14 waves (waves A to N) of the German Socio-Economic Panel to test (among other things) for state dependence effects in male unemployment behaviour. Estimation of the models is based on the maximum likelihood approach. The best model turns out to be the dynamic probit model with equi-correlated disturbances where an individual"s unemployment probability at a given point in time depends on his labour force status in the previous period and which controlls for unobserved population heterogeneity. This model shows that there are strong and significant state dependence effects in individual unemployment as well as significant unobserved heterogeneity of the disturbances. This means, reducing unemployment today will have positive long-term effects on the labour market in the future