The aim of the project is to exploit very large and comprehensive new Swedish datasets to evaluate the effects of the Swedish Active Labour Market Programmes (ALMPs) for the unemployed on a variety of labour market outcomes of participants. In assessing the effectiveness of the Swedish labour market policy in the 1990s, two institutional features are of crucial importance. First, there is a wide variety of different programmes from which the unemployed can choose from; second, participation in a programme renews eligibility to comparatively generous unemployment benefits. In addition, the standard selection problem needs to be addressed in order to recover causal programme impacts, in particular the observed differences in the performance of participants and non-participants need to be purged of the potentially different composition of the unemployed workers who select into the different programmes.

From a methodological point of view, both non-parametric and parametric evaluation methods will be employed, which will allow us to assess the robustness of the results to the different assumptions underlying each methodology.