Drawing on a large georeferenced data set on workers and unemployed persons in Germany, we evaluate the effectiveness of a active labor market policies (ALMP) in Germany by means of a spatial regression discontinuity design. We exploit the fact that unemployed individuals who have exhausted their benefits from unemployment insurance are assigned strictly geographically to local job centers, which provide job search support in the form of ALMP. The fact that the average treatment intensity of these policies changes discontinuously at county borders while regional labor market conditions vary smoothly in space allows for a precise estimation of the effects that different types of policies (public employment schemes, training programs, wage subsidies, coaching) have on subsequent employment prospect of job seekers. The structure of the data allows us to control for potential lock-in effects and to shed light on dynamic employment effects over longer periods. Our preliminary results indicate that, at least over a horizon of 24 months, participants in ALMP does not raise the number of days in employment but, if anything, leads to a slight decrease. |
