Memorandum, Department of Economics, University of Oslo 2003,26
We investigate treatment effects of active labour market programmes for Norwegian adults for the 1990 to 2000 period. Three types of active labour market programmes are evaluated within a competing risks hazard rate model. Non-parametric specifications on both duration dependence and unobserved heterogeneities are used. By utilising rich administrative data, we find that active labour market programmes do have intended effects on enhancing the transition probability to employment after the completion of programmes participation, but during the participation, the transition probability is low relative to that for non-participants. There is some evidence of heterogeneity of treatment effects with respect to observed individual characteristics, and effects for training programmes and wage subsidy programmes are pro-cyclical and more favourable at boom time. The positive treatment effects of labour market programmes are long lasting, at the same time diminish gradually over time when individuals remain in unemployment after completion of programmes. The net impact of active labour market programmes in terms of reduced total amount of unemployment exposure is estimated to be about 6.42%.
labour market programmes treatment effects competing risk non-parametric estimation