Do Lower Caseloads Improve the Performance of Public Employment Services? New Evidence from German Employment Offices

• Author: Katja Wolf,Gerhard Krug†,Jens Hainmueller,Barbara Hofmann
• Date: 01 October 2016
• Published date: 01 October 2016
• DOI: http://doi.org/10.1111/sjoe.12166

Scand. J. of Economics 118(4), 941–974, 2016

DOI: 10.1111/sjoe.12166

Do Lower Caseloads Improve the

Performance of Public Employment

Services? New Evidence from German

Employment Offices∗

Jens Hainmueller

Stanford University, Stanford, CA 94305-6044, USA

jhain@stanford.edu

Barbara Hofmann

Institute for Employment Research, DE-90478 Nuremberg, Germany

barbara.hofmann@iab.de

Gerhard Krug†

Institute for Employment Research, DE-90478 Nuremberg, Germany

gerhard.krug@iab.de

Katja Wolf

Institute for Employment Research, DE-90478 Nuremberg, Germany

katja.wolf@iab.de

Abstract

The caseworker-to-clients ratio is an important, but understudied, policy parameter that af-

fects both the quality and cost of public employment services that help job seekers find

employment. We exploit a large-scale pilot by Germany’s employment agency, which hired

490 additional caseworkers in 14 of its 779 offices. We find that lowering caseloads caused

a decrease in the rate and duration of local unemployment as well as a higher re-employment

rate. Disentangling the mechanisms that contributed to this improvement, we find that offices

with lowered caseloads increased monitoring and imposed more sanctions but also intensified

search efforts and registered additional vacancies.

Keywords: Caseworker; labor market policy; statistical matching; unemployment

JEL classification:C14; H43; H83; J68

I. Introduction

Active labor market policies (ALMP) are widely used in many countries to

help unemployed workers find jobs, but there is mixed empirical evidence

about the effectiveness of these policies. Some studies show that public

∗We thank Michael Rosholm, Julia Schneider, Jeff Smith, and participants at an ASB/IAB

workshop in Nuremberg, as well as JSM and EALE conferences, for helpful comments. We

are grateful to Frank Sowa and Stefan Theuer for sharing information about the implemen-

tation of the pilot project. The usual disclaimer applies.

†G. Krug is also affiliated with University of Erlangen-Nuremberg (gerhard.krug@fau.de).

CThe editors of The Scandinavian Journal of Economics 2016.

942 Do lower caseloads improve public employment services?

intervention in the labor market to match the unemployed with employers

can help job seekers to find work more quickly (Yavas, 1994; Foug`

ere

et al., 2009). In contrast, other studies suggest that ALMP can also result in

inefficiencies, if, for example, the formal monitoring of job search efforts

by caseworkers crowds out the informal private search by job seekers

(van den Berg and van der Klaauw, 2006).

One of the most important unresolved questions about ALMP is how

the behavior of caseworkers in public employment offices affects the re-

employment chances of job seekers. Caseworkers play a crucial role in

most labor market policies because they work directly with job seekers and

try to help them find new employment. Differences in the quality, work

conditions, and training of caseworkers can therefore affect the success of

such policies (e.g., Behncke et al., 2008; Rinne et al., 2013). Consistent

with this idea, Lagerstr¨

om (2011) finds that caseworkers in Swedish em-

ployment offices vary dramatically in their effectiveness at bringing their

clients back into regular employment.

Why is it that some caseworkers are more effective than others? Very

few studies shed light on this important issue. For example, using data

from Switzerland, Behncke et al. (2010) find that caseworkers who follow

a less cooperative and less harmonious approach towards the unemployed

increase the employment chances of their clients. In this study, we examine

the effect of another potentially important factor: caseload (i.e., the ratio

of caseworkers to unemployed clients). Caseload is an important policy

parameter for at least two reasons. On the one hand, caseload influences the

effectiveness of the assistance because it dictates how much time and effort

a caseworker can devote to each client. On the other hand, the caseload

is also a key driver of the administrative costs of the policy, as lower

caseloads require that public employment offices hire more caseworkers.

Despite the importance of this question for public policy, we have almost

no empirical evidence about the effects of caseload on key indicators such

as the unemployment rate, the duration of unemployment, and the outflow

rate from unemployment to regular employment.

Our study contributes new evidence by drawing upon a large-scale pi-

lot project of Germany’s federal public employment agency (Bundesagen-

tur f¨

ur Arbeit, BA), which lowered the caseload in 14 of its 779 local

employment offices in May 2007. The ratio of caseworkers to recipients

of unemployment insurance (UI) benefits was set to 1:40 in the pilot

offices, while it was about 1:100 in the non-participating offices (mea-

sured as full-time equivalents). Although not randomized, the BA chose the

participating offices based on well-documented criteria that were mostly

designed to achieve a representative sample. As a result, participating

offices were fairly similar to non-participating offices prior to the pilot

project. Our empirical strategy relies on a combination of matching and a

CThe editors of The Scandinavian Journal of Economics 2016.

J. Hainmueller et al. 943

Fig. 1. Monthly re-employment rate before and after start of pilot project in pilot and

control employment offices (May 2005–April 2008)

Notes: Monthly re-employment rate: number of re-employed individuals per month divided by stock of unem-

ployed at the end of the preceding month.

Source: Calculations based on micro data of the German Federal Employment Agency (see text for details).

difference-in-differences (DiD) estimator to isolate the causal effects of the

caseload decrease on several outcomes, including the unemployment rate,

the cumulative unemployment duration, and the re-employment rate. We

also conduct various robustness checks to corroborate the main findings,

including tests for differential trends in the pre-program period. More-

over, we consider two intermediate outcomes (sanction rates and number

of new vacancies registered) to shed light on the potential causal mecha-

nisms through which caseload affects outcomes. We also consider potential

negative side effects such as spillover into neighboring regions. Finally,

we analyze the cost effectiveness of the pilot program to obtain a broader

measure of the policy returns.

Overall, we find that the pilot project led to an improvement in the per-

formance of participating local employment offices. Lowering of caseloads

resulted in a decrease in the duration and rate of local unemployment and

an increase in the re-employment rate. To preview the main result and iden-

tification strategy, the upper panel in Figure 1 shows the average monthly

re-employment rate in the group of pilot offices and the comparison group

of non-participating offices over a three-year period, including the two

years prior to the start of the pilot in May 2007 and the year following the

CThe editors of The Scandinavian Journal of Economics 2016.