The Effects of a Job‐Creation Scheme: Evidence from Regional Variation in Program Capacities

• Author: Rainer Eppel
• Published date: 01 January 2017
• DOI: http://doi.org/10.1111/irel.12167
• Date: 01 January 2017

The Effects of a Job-Creation Scheme: Evidence

from Regional Variation in Program Capacities

*

RAINER EPPEL

In direct job-creation schemes, unemployed individuals at risk of permanent labor

market exclusion are offered temporary subsidized employment in public- or non-

profit-sector firms in combination with skills training and sociopedagogical support.

The main aim is to stabilize and qualify them for later re-integration into the regular

labor market. Exploiting exogenous regional variation in population-group–specific

program capacities, I find that, on average, such a program eases the transition out

of unemployment but has no long-term effect on regular employment.

Introduction

Even in the countries with the lowest unemployment rates, there are seg-

ments of individuals with severe difficulty gaining a foothold in the regular

labor market. Often these people face multiple employment obstacles, such as

long periods of joblessness, a lack of qualifications, old age, disability, and

social problems. As a consequence, they seem to have few prospects of finding

a regular job, and conventional modes of public employment assistance reach

their limits. Direct job-creation schemes (JCS) have emerged as a policy tool

to help these individuals at risk of permanent labor-market exclusion. In such

a measure, unemployed workers are provided with temporary subsidized jobs

in autonomous public or nonprofit sector “social enterprises”that trade in the

market but serve social needs. The main aim is to stabilize and qualify them

for later re-integration into the regular labor market. In this article, I evaluate

the impact of participating in an Austrian JCS that combines subsidized

employment in a relatively sheltered environment with training and

*The author’saffiliations are Austrian Institute of Economic Research (WIFO), Vienna, Austria, and Kiel

Institute for the World Economy (IfW), Kiel, Germany. E-mail: Rainer.Eppel@wifo.ac.at.

The author is grateful to the Austrian Federal Ministry of Labour, Social Affairs and Consumer Protection

(BMASK) for financial support of the research project that built the starting point for this paper (see Eppel

et al. 2014). Further thanks go to the project co-authors from WIFO and Prospect Unternehmensberatung,

especially Helmut Mahringer and Tom Horvath, as well as to Christine Zulehner and three anonymous refer-

ees for very valuable comments and suggestions.

JEL: C26, J08, J68.

INDUSTRIAL RELATIONS, Vol. 56, No. 1 (January 2017). ©2016 Regents of the University of California

Published by Wiley Periodicals, Inc., 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington

Road, Oxford, OX4 2DQ, UK.

161

sociopedagogical support. I assess whether the program is effective in provid-

ing a bridge to stable employment in the regular labor market.

Previous evidence. Previous studies of direct job creation in the public or

nonprofit sector tend to find that this measure has an insignificant or even

adverse impact on postprogram employment outcomes. This is documented by

several meta-analyses of program evaluations (see Heckman, LaLonde, and

Smith 1999; Kluve and Schmidt 2002; Card, Kluve, and Weber 2010; and

Kluve 2010). Much of the available evidence comes from microevaluation

studies of the traditional German JCS, which was very similar to the Austrian

one (cf. Wolff and Stephan 2013). These studies suggest that, despite some

heterogeneity in the effects with respect to subgroups and timing of program

starts, the German JCS generally performed poorly in improving the re-integra-

tion of participants into regular employment. Even after the typical lock-in

effects, the estimated effects were found to be zero or negative for most partic-

ipant groups (Caliendo, Hujer, and Thomsen 2004, 2006, 2008a, 2008b; Hujer

and Thomsen 2010; Hujer and Zeiss 2007; Lechner and Wunsch 2009; Ste-

phan and Pahnke 2011; Wunsch and Lechner 2008).

1

The results for Germany

are generally in line with recent evidence for other countries that, except for

one study by Van Ours (2001) for the Slovac Republic, suggests zero or nega-

tive effects (see Gerfin and Lechner [2002] for Switzerland, Fredriksson and

Johansson [2003] for Sweden, as well as Hardoy [2005] and Lorentzen and

Dahl [2005] for Norway).

Hence, previous results are consistent. However, with only very few excep-

tions that apply to duration models (see Van Ours [2001] and Hujer and Zeiss

[2007]), all of the recent studies use propensity score matching to identify pro-

gram effects—an approach that critically relies on the selection-on-observables

1

The results of only one study are more optimistic: Hohmeyer and Wolff (2010) found that—after a

period with moderate lock-in effects—the traditional German JCS contributes to a significantly better

employment performance of the participants. A possible explanation for this difference in results is the dif-

ferent time period analyzed and, thus, the different participant groups that were studied. The traditional Ger-

many JCS, in force from 1969 to 2012, was very similar to the Austrian one, as persons in need of

assistance were offered temporary jobs mainly in public or noncommercial institutions and received a regular

wage during participation (cf. Wolff and Stephan 2013). Before the Hartz welfare reforms in 2005, primarily

unemployment insurance (UI) recipients participated in the scheme. After the introduction of the reforms,

the scheme was available only for welfare recipients (unemployment Benefit II recipients). Whereas all other

evaluations analyzed the period before the Hartz welfare reforms and mostly found negative average pro-

gram participation effects of the German JCS on regular employment, Hohmeyer and Wolff (2010) provide

positive evidence for welfare recipients who participated after the introduction of the reform. A possible

explanation for the more optimistic result is that welfare recipients have far lower re-employment prospects.

Because they receive fewer job offers, program participation does not as much keep them from accepting

regular jobs, and there is a larger scope for improving their employment prospects (cf. Hohmeyer and Wolff

2010).

162 / RAINER EPPEL