using regional labour market characteristics before the Hartz IV reform, we use an
instrumental variable approach. Furthermore, since they are worried that the composition of
long-term unemployed in Approved Local Providers differs from the composition of the
long-term unemployed in Joint Local Agencies, they focus on correcting for differences in
individual characteristics across regions.1 Given their focus, however, they are only able to
use data of 154 regions, while our dataset includes 409 out of 442 regions. The third
difference is that Broockmann et al. (2010) do not explain the rationale behind the self-
selection of Approved Local Providers and therefore do not attempt to reconcile with what
the finding that despite of positive self-selection of regions into Approved Local Providers
these did not perform better than Joint Local Agencies.
By estimating a matching function in order to determine which factors are positively
correlated with unemployment to employment transitions our paper also contributes to the
empirical literature estimating matching functions for Germany (Gross, 1997; Entorf, 1998;
Fahr and Sunde, 2005). Similar to Fahr and Sunde (2005), we use regional data to estimate
the matching function for the long-term unemployed taking spatial lags into account. Fahr
and Sunde (2005) emphasise the differences in the performance of local job centres without
being able to explain the differences in matching efficiency. Our dataset on organisational
characteristics allows us to shed some light into this black box.
The paper is organized as follows. Section 2 describes the changes in labour market
institutions and the policy experiment resulting from the Hartz IV reform. In section 3 we
provide evidence for positive self-selection of regions into being Approved Local Providers
by comparing the labour market conditions in regions with Approved Local Providers and
Joint Local Agencies before the Hartz IV reform. Section 4 describes the data used. In
section 5 we first present some basic descriptive evidence and OLS estimates for the
correlation between the job finding probability of the long-term unemployed and being cared
for by an Approved Local Provider. We then discuss the instrument we use to account for
positive self-selection of regions into being Approved Local Providers and present the IV
estimates of the overall effect of being cared for by an Approved Local Provider. In section 6
we account for the differences in the organisational structure implemented by job centres and
investigate the role of these organisational characteristics for the job finding rate of the long-
term unemployed followed by the conclusion in section 7.
1 Broockmann et al. (2010) and WZB et al. (2008) provide evidence that neither of the two institutions had an
advantage by having an easier to integrate group of long-term unemployed.