Give it another try: What are the effects of a public employment scheme especially designed for hard-to-place workers?

Job creation schemes (JCSs) have for a long time been considered to function as stepping stone into the labour market for unemployed workers. However, previous research has shown that public job creation schemes have mostly negative effects on the employment outcomes of participants, probably due to strong lock-in effects, which are particularly strong for unemployed workers with higher labour market attachment. This raises the question of whether JCSs could be an effective policy tool for very-hard-to-place workers who have very low job chances anyway. We contribute to this discussion by analysing a JCS that employs a special selection mechanism to identify these workers. Relying on a combination of administrative data and survey data, we employ radius-matching with regression adjustment to estimate treatment effects. Our results indicate that the effects are still negative but weaker than the ones reported in previous studies. Furthermore, we point to effect heterogeneity with regard to contract duration which suggests that the results are more optimistic for shorter contracts. Finally, we show that the results are sensitive to the inclusion of survey variables, even if high quality administrative data are available.