Effectiveness of sequences of One-Euro-Jobs for welfare recipients in Germany

Abstract

Empirical evidence reveals that German welfare recipients often participate in multiple active labour market programmes (ALMPs). However, evidence on the effectiveness of ALMPs exists mostly for single ALMP participations so far. This article evaluates the sequential participation in ALMPs for welfare recipients in Germany based on comprehensive administrative data to control for dynamic selection that arises in the evaluation of sequences. Using a dynamic causal model and an inflow sample of welfare recipients, the article analyses the effects of sequences of a public employment programme called One-Euro-Jobs on labour market outcomes. For female participants in One-Euro-Jobs in the first period, especially in West Germany, the results imply that participating in two consecutive One-Euro-Jobs compared with receiving only welfare benefits for two consecutive periods better facilitates integration into regular employment. Moreover, taking part in a One-Euro-Job directly after entry into welfare receipt is also more effective for participants in One-Euro-Jobs in the first period than taking part in a One-Euro-Job in a later period, especially for East German men (although not for West German women). However, I also find evidence of so-called programme careers and stepwise integration into regular employment.

Keywords:

• sequences
• dynamic causal model
• propensity score matching
• activation
• effect evaluation

JEL Classification:

• C13
• I38
• J68

Acknowledgements

I thank Joachim Wolff, Katrin Hohmeyer, Regina T. Riphahn, Gesine Stephan, Eva Kopf and one anonymous referee for their very helpful comments. All errors are my own.

Notes

¹ Individuals aged between 15 and 64 years that could work at least 3 hours per day.

² For a detailed description of the UB-II-system, see Eichhorst et al. (2010).

³ For further information on ALMPs and the Hartz reforms, see Jacobi and Kluve (2007).

⁴ Since 2012, job centres are no longer required to place young adults in One-Euro-Jobs without delay.

⁵ For further information on the theoretical effects of One-Euro-Jobs, see Hohmeyer and Wolff (2012).

⁶ In this article, all notations and definitions of effects are based on these studies.

⁷ Sequences are denoted by parentheses. For example, (1EJ,UBII) defines the sequence with One-Euro-Job as the first state in the first period and UB-II-receipt as the second state in the second period.

⁸ I use data from the Integrated Employment Biographies (IEB) and the UB-II-Receipt History (‘Leistungshistorik Grundsicherung’ (LHG)).

⁹ Data from local authorities (‘zugelassene kommunale Träger’) are not included due to data collection problems. In these 69 districts, the Federal Employment Agency did not administer the UB II. Approximately 13% of unemployed welfare recipients were living in these districts in 2005 (Hohmeyer and Wolff, 2012).

¹⁰ Unemployment insurance benefits are labelled UB I in Germany.

¹¹ I also did some further data preparations: I do not consider individuals with programme combinations starting on the same day, individuals participating in more than six programmes and individuals with missing information on planned durations of One-Euro-Jobs.

¹² Minor employments pay up to 400 euros per month (until 2012) and are not subject to social insurance contributions.

¹³ I tested several possible durations of start windows. The trade-off was not only to obtain a short start window, but also to have enough starts within this window to provide a sufficient number of observations.

¹⁴ I use planned duration because programme participation cannot influence planned duration.

¹⁵ For more information on work opportunities as contributory employments, traditional job creation schemes and One-Euro-Jobs, see Hohmeyer and Wolff (2010). For more information on JobPerspectives, see Dengler et al. (2013).

¹⁶ Table 1, steps B.4, B.7 and C.4.

¹⁷ Not all covariates have to be included in all probit models of the three matching steps. Thus, this could lead to a poorer final matching quality, as I consider all MSBs of all covariates.

¹⁸ To calculate the respective calipers, I estimate the 75th percentile of the differences between the propensity scores of treatments and controls, using nearest neighbour matching (one-to-one) with replacement. I choose the 75th percentile as a caliper, as smaller calipers do not improve the matching quality of the dynamic matching step 3 and the final matching quality. Thus, I drop the poorest 25% of matches. Results on the MSB and effects without calipers for the matching step 3 are available upon request.

¹⁹ The results are available upon request.

Additional information

Funding

Financial support by the Graduate Programme (GradAB) of the IAB and the University of Erlangen-Nuremberg is gratefully acknowledged.