Temporary Agency Work and Firm Competitiveness: Evidence from German Manufacturing Firms

• Date: 01 July 2014
• Published date: 01 July 2014
• DOI: http://doi.org/10.1111/irel.12062

Temporary Agency Work and Firm

Competitiveness: Evidence from German

Manufacturing Firms

*

SEBASTIAN NIELEN and ALEXANDER SCHIERSCH

This paper addresses the relationship between the utilization of temporary agency

workers by firms and their competitiveness measured by unit labor costs, using a

rich, newly built, dataset of German manufacturing enterprises. We conduct the

analysis by applying different panel data models while taking the inherent selec-

tion problem into account. Making use of dynamic panel data models allows us

to control for firm-specificfixed effects as well as for potential endogeneity of

explanatory variables. The results indicate an inverse U-shaped relationship

between the extent that temporary agency workers are used and the competitive-

ness of firms.

Introduction

TEMPORARY AGENCY WORK IS A TOOL THAT ALLOWS FIRMS TO ADJUST LABOR INPUT

on short notice. In Germany, it has become increasingly important since 1994,

when regulations concerning temporary employment were relaxed. Between

1994 and 2007, the number of temporary agency workers in Germany quadru-

pled from roughly 175,000 to more than 700,000. However, it must be noted

that despite this growth, only 2.4 percent of the working population in 2007

were hired by temporary agencies (Schmidt and W€

ullerich 2011). The growth

in the use of temporary agency workers is by no means only a German phe-

nomenon; it can be observed throughout the industrialized world. The share of

temporary agency workers in Japan, for example, grew in the active population

by more than 1.3 percentage points to 2.1 percent between 2000 and 2007. In

European countries such as Switzerland, Austria, Finland, and Italy, the growth

*The authors’affiliations are, respectively, Institute of Applied Economic Research T€

ubingen, T€

ubingen,

Germany. Email: sebastian.nielen@iaw.edu; German Institute for Economic Research, Berlin, Germany

Email: aschiersch@diw.de.

JEL codes: D24, L23, L60.

*The authors would like to thank two anonymous referees, Werner B€

onte, and Alexander Kritikos for their

helpful and valuable comments.

INDUSTRIAL RELATIONS, Vol. 53, No. 3 (July 2014). ©2014 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.

365

in temporary agency work was 0.7 percentage points over the same period

(Eichhorst, Marx, and Thode 2010).

The increasing importance of temporary work is of growing interest to econ-

omists. Although there is an extensive discussion on this form of employment,

the discussion is driven by a labor market perspective. Only a few studies have

taken into account or explicitly analyzed the effect of temporary agency work

on firm performance. Arvanitis (2005), Kleinknecht et al. (2006) and Bryson

(2007) investigate the effects of the utilization of temporary agency work for

Switzerland, the Netherlands, and Great Britain. All three studies, while mak-

ing use of pooled ordinary least squares (OLS), find no significant effects on

sales, value added, or sales per capita. In addition, Beckmann and Kuhn

(2009) as well as Hirsch and Mueller (2012) study the question for the Ger-

man case. Both studies apply panel data models to account for time variant

and invariant heterogeneity on data from the Institute for Employment

Research (IAB) establishment panel to analyze the effect of temporary agency

work on value added and sales. Both papers find a hump-shaped relationship

between temporary agency work and productivity. Finally, Beckmann and

Kuhn (2012), also using the IAB establishment panel, analyze which strategy

pays off: The use of temporary agency work to increase input flexibility or to

screen for new productive workers. Also controlling for the self-selection prob-

lem, they find negative effects for the flexibility strategy on firm performance,

i.e., sales per capita.

This study extends the existing literature on the effects of temporary

employment on firm performance in two ways: We apply panel data models

on a rich, newly combined, panel dataset of more than 26,000 German manu-

facturing enterprises. The advantage of this dataset is that it contains annual

data for the use of temporary agency work instead of date data. Moreover, the

study focuses on competitiveness, which is measured by unit labor costs

(ULC). This has the advantage that not only the effect of temporary employ-

ment on productivity is taken into account, but also potential cost effects. Fur-

thermore, the analysis explicitly controls for the inherent selection problem

using temporary agency work, because some firms systematically do not use

this instrument. To offset the strict exogeneity assumption of explanatory vari-

ables and to account for dynamic effects, we use system GMM in addition to

OLS and the fixed-effect approach.

Almost all econometric results confirm our hypothesis of an inverse U-

shaped relationship between the share of temporary agency work and firm

competitiveness. Only when temporary agency work is modeled as being

endogenous is this pattern not found for the entire dataset. However, when

testing the hypothesis separately for industries and industry groups, the inverse

U-shaped relationship is found even for the model taking into account the

366 / SEBASTIAN NIELEN,AND ALEXANDER SCHIERSCH