Regional Variation in Treatment Intensity as an Instrument for the Evaluation of Active Labour Market Policies

In this paper we evaluate the mean effects of Swiss active labour market policies (ALMP) on employmentand earnings of individual participants by nonparametric instrumental variables based on rich administra-tive data. Identification of local complier effects are obtained by using the fact that Switzerland consist ofmany fairly autonomous regions (cantons) that in several case cut through local labour markets. By somespecific policy intended to encourage the regional units to use ALMP on a larger scale, the federal gov-ernment generated an exogenous variation in the participation probabilities across cantons. Here we usethat variation to compare individuals living in the same labour market but in different cantons. Takingaccount of small sample problems occurring in IV estimation, our preliminary results suggest that ALMPappears to increase employment by about 10-20%.