Productivity in German manufacturing firms: Does fixed‐term employment matter?

• Date: 01 December 2016
• Author: Alexander SCHIERSCH,Sebastian NIELEN
• Published date: 01 December 2016
• DOI: http://doi.org/10.1111/ilr.12019

International Labour Review, Vol. 155 (2016), No. 4

Copyright © The authors 2016

Journal compilation © International Labour Organization 2016

* IfM Bonn, email: nielen@ifm-bonn.org. ** German Institute for Economic Research,

Berlin, email: aschiersch@diw.de.

Responsibility for opinions expressed in signed articles rests solely with their authors, and

publication does not constitute an endorsement by the ILO.

Productivity in German

manufacturing rms:

Does xed-term employment matter?

Sebastian NIELEN* and Alexander SCHIERSCH**

Abstract. Using a large data set of German manufacturing establishments and

various panel data models, the authors investigate the relationship between labour

productivity and the use of xed-term employment, taking account of the possible

distortions that may result from self-selection into the use of xed-term contracts.

Their empirical results provide no evidence for the expected inverse U-shaped re-

lationship between xed-term employment and labour productivity, and testing for

a linear relationship leads to mostly negative coefcients that are only signicant

in a few specications. Overall, their results thus indicate that there is no signi-

cant relationship between the use of xed-term contracts and labour productivity.

The importance of xed-term employment in Germany is constantly in-

creasing, with more and more rms using this form of employment. The

share of xed-term contracts for new hires increased from around 30 per cent

in 2000 to about 45 per cent in 2 010 (IAB, 2011). In 2010, one out of every

ten employees liable for social security contributions was employed under a

xed-term contract, an increase of 30 per cent compared to 2000 (Gundert

and Hohendanner, 2011).

There are, however, only a few empirical studies analysing the effect of

the increased use of xed-term employment on productivity. Using sector ag-

gregates, Damiani and Pompei (2010) analyse the effect of labour protection

on total factor productivity (TFP) growth in 18 European countries between

1995 and 2005. They also control for the effect of growth in temporary em-

ployment on TFP, nding a negative and signicant relationship. Also using

sectoral data, Auer, Berg and Coulibaly (2005) analyse the effect of employ-

ment tenure on productivity in 13 European countries from 1992 to 2002 . Their

results show that productivity increases with job tenure, although it decreases

after 13 years of tenure. It follows that rms with a lower share of xed-term

International Labour Review536

workers should have higher productivity. Making use of Spanish sectoral data

from 1987 to 200 0, Ortega and Marchante (2010) report a negative effect of

temporary contracts, but only in the manufacturing and energy sectors. No ef-

fects are found for the remaining sectors.

At the micro level, Cappellari, Dell’Aringa and Leonardi (2012) use

13,000 rm-level observations, covering all sectors of the Italian economy be-

tween 2004 and 2007, to analyse the effects of reforms deregulating the use

of xed-term contracts on labour productivity. They nd a small negative, but

only weakly signicant, effect and must, therefore, reject their hypothesis that

the reforms had increased labour productivity. This result is in line with the

ndings of the two previously mentioned studies. Finally, Kleinknecht et al.

(2006) analyse the effect of xed-term employment on 590 Dutch rms. They

nd that the percentage of personnel with xed-term contracts has no signi-

cant effect on sales growth. In order to check the robustness of this nding,

they split the data set into rms with active R&D and rms without active

R&D. Again, the use of xed-term employment is found to have no effect

on sales growth in either subgroup. With the exception of Kleinknecht et al.

(2006), previous empirical results thus point to a weakly negative relationship.

This study contributes to the limited literature on this issue in two ways:

(i) it is the rst to provide evidence of the effect of xed-term employment

on labour productivity in Germany, a major European economy with an

ever-growing proportion of atypical employment; and (ii) in contrast to the

aforementioned studies, we explicitly control for the potential problem of self-

selection into the use of xed-term contracts, given that some rms might sys-

tematically not use such contracts. Additionally, we apply dynamic panel data

models to soften the assumption of strict exogeneity of explanatory variables.

Our main results show that rms do not gain higher labour productivity

from using a modest number of xed-term workers. However, we also show

that there is no negative effect from using xed-term workers intensively. In

other words, contrary to previous studies, we nd that rms are not worse off

if they use xed-term workers instead of regular employees.

The remainder of the article is organized as follows. The rst section de-

scribes the system of industrial relations and the regulation of xed-term em-

ployment in Germany. The second section discusses the related literature and

derives our hypothesis. The third section introduces the data and discusses the

descriptive statistics. Our methodology and empirical strategy are introduced

in the fourth section, along with our empirical analysis. The fth section pro-

vides a concluding discussion.

Institutional backdrop

This section consists of two parts. The rst provides a general overview of the

system of industrial relations in Germany. The second describes the regula-

tion of xed-term employment in Germany and how this has changed since

the implementation of the Employment Protection Act in 1985.