Does the Order and Timing of Active Labour Market Programmes Matter?
| Date | 01 April 2013 |
| Author | Michael Lechner,Stephan Wiehler |
| Published date | 01 April 2013 |
| DOI | http://doi.org/10.1111/obes.12007 |
180
©Blackwell Publishing Ltd and the Department of Economics, University of Oxford 2013. Published by Blackwell Publishing Ltd,
9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 75, 2 (2013) 0305-9049
doi: 10.1111/obes.12007
Does the Order and Timing of Active Labour Market
Programmes Matter?
Michael Lechner and Stephan Wiehler
University of St. Gallen, Swiss Institute for Empirical Economic Research (SEW-HSG),
Varnbüelstr. 14, CH-9000 St. Gallen, Switzerland (e-mail: michael.lechner@unisg.ch;
stephan.wiehler@credit-suisse.ch)
Abstract
This study extends the traditional focus of active labour market policy evaluation from a
static comparison of participation in a programme versus non-participation (or participa-
tion in another programme) to the evaluation of the effects of programme sequences, that
is, multiple participation or timing of such programmes. Explicitly allowing for dynamic
selection into different stages of such programme sequences we analyse multiple pro-
grammes, the timing of programmes, and the order of programmes. The analysis is based
on comprehensive administrative data on the Austrian labour force. Our findings suggest
that (i) active job search programmes are more effective after a qualification programme
compared to the reverse order, (ii) multiple participation in qualification measures domi-
nates single participation and (iii) the effectiveness of several labour market programmes
deteriorates the later they start during an unemployment spell.
I. Introduction
For the last 20 years, a steadily growing number of active labour market policy evaluation
studies have been conducted for Europe.1So far, the majority of the studies concentrated on
the impact of a single programme compared with non-participation or with participation in
other programmes. In this study, we extend this focus to issues of multiple participations,
as well as the timing of active labour market programmes.
These issues are important for policy makers as European employment offices develop
more and more sophisticated strategies for particular types of unemployed. This trend
seems to be particularly pronounced in Austria. These strategies involve, for example,
sequences of similar, or quite different, types of active labour market programmes trig-
gered at particular points in time in the unemployment spell. Therefore, to understand the
effects of such strategies for the unemployed, it is necessary to evaluate the effects not of
the single programme but of the appropriate combinations of programmes, for example,
taking into account their specific timing. This comes with particular econometric problems,
JEL Classification number: J68.
1For a survey on active labour market policies in Europe see Kluve et al. (2007), as well as the summaries given
in Heckman, LaLonde and Smith (1999).
Order and timing of active labour market programmes 181
because typically the planned programme sequence and its timing are either not known to
the evaluator or not yet fully set at the beginning of the unemployment spell. Therefore,
the impact of the intended sequences has to be estimated from a sample of individuals
actually participating in complete sequences, instead of those assigned to the sequence.
The former sample is however selective, as not all who started the sequence may reach its
intended end. Instead, for reasons potentially related to the effects of the early components
of the sequence, participants may drop out. This gives rise to potential dynamic selection
bias and constitutes the econometric challenge tackled in this study. To do so, we use the
dynamic potential outcome model first suggested by Robins (1986) in the epidemiological
literature and recently discussed by Lechner and Miquel (2010) and Lechner (2008, 2009)
in the econometrics literature.
So far, the conclusions from the microeconometric literature about the effects of sin-
gle active labour market programmes in Europe are mixed. Apart from differences at the
country level, there seems to be a broad consensus about effect heterogeneity not only
with respect to the type of the programme,2but also with respect to programme partici-
pants.3The majority of studies concentrate on a comparison of participation in a particular
programme versus non-participation or participation in another programme, either at one
point in time or in the same time span.
Recent literature has seldom addressed the effects of differential timing of labour mar-
ket programmes or the effects of programme sequences, that is, participating in either
the same programme more than once or in different programmes: Carling and Richard-
son (2004) estimate the impact of labour market programmes on subsequent unemploy-
ment in Sweden. In a duration modelling framework they find that the duration until
programme entry has no impact on the estimated hazard rates. Another example is Sia-
nesi (2004) who groups Swedish labour market programmes according to the length of
the respective unemployment spell before the programme. She finds that earlier allocation
increases subsequent employment prospects. Fitzenberger and Speckesser (2007), B. Fit-
zenberger, A. Osikominu and R. V¨olter (unpublished data) and Fitzenberger and V¨olter
(2007) adopt this approach for Germany and find similar results. However, a particular
problem of that approach is that at any point in time ‘starting’participants are compared with
‘not yet’ participants. Therefore, the implicit counterfactual changes consequently, which
makes the interpretation of the estimated quantities difficult. Flores-Lagunes, Gonzalez and
Neumann (2007) estimate average causal effects of different lengths of exposure to Job
Corps (JC) training in the US. They suspect that the length of a programme is potentially
2For job creation schemes in Switzerland see Gerfin and Lechner (2002). Similar results appear in Lechner and
Wunsch (2006) and in Caliendo, Hujer and Thomson (2004, 2006, 2008) for Germany.For wage or integration sub-
sidies in Sweden see Sianesi (2008) and Forslund, Johannson and Lindqvist (2004), and for Switzerland see Lalive,
van Ours and Zweim¨uller(2002) and Gerfin, Lechner and Steiger (2005). For business start-up programmes, we refer
to Carling and Gustafson (1999). For training measures, comprising formal qualification, further training of any kind
and retraining see Richardson and van den Berg (2001) for Sweden, Gerfin and Lechner (2002) for Switzerland, and
Hujer, Thomsen and Zeiss (2005) for Germany. Lechner, Miquel and Wunsch (2011) investigate long-run effects
and Lechner and Wunsch (2006) analyse business cycle effects. Winter-Ebmer and Zweim ¨uller (1996), Hofer and
Weber (2004a, b) and Lutz, Mahringer and P¨oschl (2005) investigate employment effects for different instruments
of the Austrian ALMP.
3See Puhani (1999) and Kluve, Lehmann and Schmidt (1999) for Poland, Friedlander, Greenberg and Robins
(1997) and Heckman et al. (1999) for the US and other Western economies. Lechner et al. (2011, and Lechner
and Wansch 2007) for Germany, as well as a survey by Bergemann and van den Berg (2006) for gender specific
differences.
©Blackwell Publishing Ltd and the Department of Economics, University of Oxford 2013
182 Bulletin
influenced by personal characteristics or by the participant’s history in the labour market
at any point in time. Thus, they use the idea of the ‘generalized propensity score’ (intro-
duced by Imbens, 2000), that is, the assumption that the length of the individual’s JC spell
is randomly assigned, conditional on a rich set of covariates. They find a hump-shaped
positive relation between the programme length and weekly earnings for their White and
Hispanic subsamples.
As mentioned above, we use the dynamic potential outcome approach for this eval-
uation study. This approach requires having informative data to allow correcting for the
potential dynamic selection biases. Furthermore, as the evaluation heavily relies on indi-
viduals having completed the sequences of interest, the data have to be large enough to
contain a sufficiently large number of such individuals. The comprehensive administrative
labour market data from Austria used in this study meets these criteria. We use it to eval-
uate a variety of programme sequences to answer questions that are not restricted to the
timing issue per se, but also relate to the issue of multiple participations in programmes
of a different or the same kind. Therefore, this study is probably the first labour market
evaluation study that deals with the issue of a dynamic programme allocation for the case
that participation in a later stage of a sequence may depend on earlier stages of the sequence
and on intermediate outcomes of the latter.
Among other results, we find that earlier programme allocation dominates allocation
in a later stage of the unemployment spell. Furthermore, active job search programmes are
more effective after participating in a qualification measure, and less effective before such
a programme.
The article is organized as follows. Section II briefly summarizes key features of the
Austrian labour market and sketches the institutional setting. In section III, we introduce
the data, discuss the underlying identification strategy and provide a descriptive analysis
of the participants of the various programme sequences under consideration. Estimation
is discussed in section IV. The results of the estimation and a brief sensitivity analysis
are provided in section V. Section VI concludes. An online appendix with supplementary
material is available (http://www.sew.unisg.ch/lechner/at).
II. The Austrian labour market – development and institutions
Compared with other European countries, Austria has experienced a rather low and sta-
ble unemployment rate over the last decade after Austria’s EU entry in 1995. In Figure 1
we observe a slight decrease of the unemployment rate between 1998 and 2000 with a
subsequent convergence to the former level of about 5.9%. Only Switzerland features an
unemployment rate that is lower over the entire observation period. In recent years also
Denmark and Sweden undercut the Austrian unemployment rate.
The objectives for the Austrian government in fighting unemployment are legally de-
fined in the Public Employment Service Act. In particular, they comprise (i) matching job
seekers and vacancies efficiently, (ii) removing any barrier that prevents this matching,
(iii) increasing the flow of information about potential matches, (iv) mitigating quantitative
and qualitative differences between labour demand and supply, (v) securing existing jobs
and (vi) providing funds for the unemployed in the case of a job loss. These goals are
pursued by means of a wide range of passive and active labour market policies.
©Blackwell Publishing Ltd and the Department of Economics, University of Oxford 2013
To continue reading
Request your trial