Public employment services under decentralization: Evidence from a natural experiment

Abstract

This paper studies whether the decentralization of public employment services (PES) increases job placements among the unemployed. Decentralizing PES has been a widely applied reform used by governments aiming to enhance their efficacy. However, economic theory is ambiguous about its effects, and empirical evidence has been scarce. Using a difference-in-differences design, we exploit unique within-country variation in decentralization provided by the partial devolution of German job centers in 2012. We find that decentralization reduces job placements by approximately 10%. Decentralized providers expand the use of active labor market programs and monitoring strategies which diminish job seekers' reemployment prospects but shift costs to higher levels of government.

Integrating Immigrants into the Nordic Labour Markets

“…The studies find no evidence of geographical lockin effects 13 of decentralisation in Sweden and Germany, but the two German studies show negative effects of decentralisation on job-finding rates. Mergele and Weber (2017) find that decentralisation reduced the number of new job matches by approximately 10% during the first four post-reform years. This is equivalent to an increase in the average unemployment duration by three months.…”

“…Both Lundin and Skedinger (2006) and Mergele and Weber (2017) find that decentralisation increases local ALMP initiatives. For instance, Mergele and Weber find an immediate and permanent shift after decentralisation towards public job creation programmes despite the fact that these programmes are less effective than other measures in increasing reemployment rates.…”

“…There are even fewer studies focusing on decentralisation/centralisation issues than on contracting out employment services. The lack of empirical evidence is probably related to empirical constraints: the level of decentralisation typically does not vary within a country, and therefore it is difficult to construct a treatment and a control group (Mergele and Weber 2017). Nonetheless, we have identified three studies on decentralisation of employment services which exploit policy experiments that brought about variation within a country.…”

“…One of the studies focuses on decentralisation in Sweden (Lundin and Skedinger 2006), and two focus on decentralisation in Germany (Boockmann et al 2013, Mergele andWeber 2017). The Swedish study examines the effects of a Swedish pilot programme in 1996 that strengthened the role of local authorities in labour market policy in certain regions.…”

“…The two German studies exploit large-scale German policy experiments. The experiment described in Mergele and Weber (2017) exploits the devolution of responsibility for public employment offices (job centres) to the district level within 41 of the 402 districts.…”