A German policy experiment from 2012 allows me to identify the short-run and longruncausal effects of decentralization on the placement efficiency of public employmentservices (PES). I exploit variation over time and across districts with different types ofPES in a difference-in-differences framework. Decentralization had significant negativeeffects on placements in the short run, but not in the long run. This points to a transitionprocess, during which caseworkers acquire relevant placement skills. Decentralizationalso had a persistent negative effect on sanctions on the welfare recipients, which probablyindicates a reluctance of local authorities to reduce the well-being of their potential voters.
Public employment services job placement matching efficiency differencein- differences.