Works Councils and Plant Closings in Germany
| Published date | 01 March 2004 |
| Author | Arnd Kölling,Lutz Bellmann,John T. Addison |
| DOI | http://doi.org/10.1111/j.1467-8543.2004.00307.x |
| Date | 01 March 2004 |
Works Councils and Plant
Closings in Germany
John T. Addison, Lutz Bellmann and Arnd Kölling
Abstract
This paper is the first study to investigate the impact of workplace representa-
tion on plant closings in Germany, using data from a nationally representative
establishment panel. Across all establishments in our sample, we find evidence
of a positive association between works council presence and plant closings.
There is the contrary suggestion that industry-wide collective bargaining plays
a neutral to benign role.As for the interaction between collective bargaining and
workplace representation, this appears strongest for establishments with fewer
than 50 employees: such plants are much more likely to close if they have a
works council and are not covered by a collective agreement.
1. Introduction
Analysis of the effect of unions on plant closure has long lagged investiga-
tion of their impact on such outcome measures as labour productivity, finan-
cial performance and investments in physical capital. This is unfortunate
because, without some indication of the effect of bargaining on closings, our
estimates of union impact may be biased or incomplete. Thus, for example,
if the sample of union firms investigated is made up of long-run survivors,
the observation of a positive union effect on productivity might be a chimera.
By the same token, negative union effects on financial performance, where
observed, may simply reflect the capture of economic rents or rent seeking
that does not impair the joint surplus (or perhaps even increases it over some
range). Corroboration of this neutrality might be forthcoming from infor-
mation on plant closings. Again, analysis of union effects on firm survival
may assist our understanding of other empirical regularities, most obviously
the decline in union density in recent years.
The focus of the present inquiry is the impact of works councils on plant
closings. The vehicle of workplace representation in Germany is the works
British Journal of Industrial Relations
42:1 March 2004 0007–1080 pp. 125–148
John Addison is in the Department of Economics at the University of South Carolina and with
IZA, Germany. Lutz Bellmann and Arnd Kölling are at the Institut für Arbeitsmarkt- und
Berufsforschung der Bundesanstalt für Arbeit, Germany. Bellmann is also a research fellow with
IZA.
© Blackwell Publishing Ltd/London School of Economics 2004. Published by Blackwell Publishing Ltd,
9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
council, not the union. Collective bargaining per se is typically conducted at
industry or regional level.1Unless expressly authorized under the collective
agreement, the works council cannot formally bargain over pay. But the
co-determination rights of the works council (see below) bestow de facto
bargaining power, so that the workplace institution may influence not only
production but also wages and fringes. In investigating the effect of the works
council, therefore, we will have to take account of the system of industrial
relations within which it is embedded.
As we see it, that system is important in two respects. First, the potential
strictures of the industry agreement may independently influence the prob-
ability of plant failure. Second, the presence of a collective bargaining agree-
ment may discipline the exertion of power at the workplace and hence
circumscribe rent-seeking by the works council. In this way, the collective
agreement may help to ensure that the works council focuses on production
rather than distribution, squabbles over which can reduce the joint surplus
and impair the survival of the plant (see Freeman and Lazear 1995). This
latter argument is analogous to that sometimes encountered in the British
literature to the effect that ‘strong’ unions are necessary to realize the full
benefits of collective voice (see e.g. Bryson 2001).
Quite apart from its general industrial relations interest, the immediate
policy context of the present inquiry is the July 2001 revision of the Works
Constitution Act (Betriebsverfassungsgesetz). The latest changes to the law
seek to facilitate the formation of works councils, as well as to increase their
influence (see Addison et al. 2003a). Since works council coverage has his-
torically been spotty among smaller establishments, it is of policy interest to
determine whether works council impact on closings differs between large
and small plants.2The wider policy context is of course yet more recent
legislation, namely, the EU Directive on national systems for informing and
consulting workers (Official Journal of the European Communities, 2002), the
German institution providing the metric if not the template for representa-
tive employee involvement measures of this type.
The plan of the paper is as follows. We first rehearse some conjectures on
the association between worker representation and plant closings, and review
the local adaptations pertinent to German circumstances. We then very
briefly consider the results of a sparse and hitherto exclusively Anglo-Saxon
empirical literature on unions and plant closings.This literature helps to con-
textualize and further motivate the present empirical inquiry. Next, a discus-
sion of our unique dataset precedes our empirical analysis. A summary
concludes.
2. Conjectures on worker representation and plant closings
There are a number of grounds for expecting worker representation to influ-
ence the probability of workplace closure. The Anglo-Saxon literature has
broadly focused on union bargaining and collective voice. In the standard
126 British Journal of Industrial Relations
© Blackwell Publishing Ltd/London School of Economics 2004.
To continue reading
Request your trial