We analyse the effects of government-sponsored training for the unemployed conducted during East German transition. For the microeconometric analysis, we use a new, large and informative administrative database that allows us to use matching methods to address potential selection bias, to study...