Industry‐Level Wage Bargaining: A Partial Rehabilitation—The German Experience
| Author | Bernd Fitzenberger,Wolfgang Franz |
| Date | 01 September 1999 |
| DOI | http://doi.org/10.1111/1467-9485.00142 |
| Published date | 01 September 1999 |
{Journals}sjpe/sjpe46-4/q193/q193.3d
Scottish Journal of Political Economy, Vol. 46, No. 4, September1999
#Scottish Economic Society 1999.Publ ishedby Blackwell Publishers Ltd, 108 Cowley Road, Oxford OX4 1JF, UK and
350 Main Street, Malden, MA 02148, USA
INDUSTRY-LEVEL WAGE BARGAINING:
A PARTIAL REHABILITATION
ÐTHE GERMAN EXPERIENCE
Bernd Fitzenbergerand Wolfgang Franz
ABSTRACT
In order to reduce unemployment, it is often recommended that industry-level wage
bargaining in Germany should be replaced by a more decentralized system. This
paper provides a critical assessment of the current wage bargaining institutions and
re-examines the case for a more decentralized system. Based on a theoretical model
integrating Insider± Outsider aspects into the comparison, the uniformly superior
employment performance of a decentralized wage bargaining system is questioned.
We conclude that, rather than solely trying to decentralize wage bargaining, a
promising policy option may be to improve the skills of the unemployed by efficient
labour market policies and to foster institutional reforms such that wage
bargaining takes account of the long-run employment consequences of wage
setting.
II
NTRODUCTION
According to conventional wisdom fully decentralized wage bargaining systems
out-perform wage settlements at a medium centralized level such as sectoral or
industry-level wage bargaining. If so, one promising candidate for explaining
persistent unemployment can be identified for those countries with industry-
level wage bargaining, such as Germany. Moreover, the obvious cure for solving
parts of the problem of joblessness is to give room for wage bargaining
exclusively on the firm level.
Our paper challenges this view to some extent. The argument is that fully
decentralized wage bargaining may be superior in some cases, however, this
result is anything but certain in others. Therefore, the demands for a radical
change of the (German) wage system, which are frequently announced in public,
are often too far reaching.
The motivation of our paper is twofold. First, the German wage bargaining
system has come under severe attack. The institutional setting of industry wage
bargaining is viewed as sharing major responsibility for wages being too
inflexible both with respect to their level and their dispersion. Indeed, there is
437
Dresden University of Technology
Centre for European Economic Research (ZEW), Mannheim and University of Mannheim
{Journals}sjpe/sjpe46-4/q193/q193.3d
virtually no disagreement that more flexibility in the wage bargaining outcomes
is in order. While this goes without saying, it is another question whether
achieving this flexibility requires wage bargaining at the firm level. More
precisely, as long as it can be shown that, on theoretical grounds, decentralized
wage bargaining does not always outperform (slightly) more centralized wage
settlements, an alternative, if not better, strategy would keep the existing
industry-level wage bargaining system and try to render it more flexible.
Second, a serious deficiency of the well known Calmfors± Driffill (1988) U-
hypothesis and its variants, that both decentralized and centralized wage
bargaining out-perform industry-level bargaining, is the lack of, or at least
unsatisfactory, integration of Insider± Outsider aspects. Our paper attempts to
enrich the existing theoretical literature. It comes as a surprise that analyses of
the optimal degree of centralization of wage bargaining did not take into
account more seriously the Insider± Outsider aspect because even proponents of
decentralized bargaining admit that Insider power might be more relevant at the
firm level than at the industry level. 1Moreover, previous work by the authors
has shown that some implications of the theoretical literature on the Calmfors±
Driffill-hypothesis, which can be tested empirically, are inconsistent with facts:
On the basis of estimated wage equations for several EU-countries no significant
differences could be found which can be attributed to low or high centralization
of wage bargaining. 2Despite the limits of that approach some doubts on the
validity of the Calmfors± Driffill-hypothesis could be established. Hence, the
obvious question is whether the failure to integrate Insider± Outsider aspects can
serve as an explanation of the limited empirical evidence concerning the
superiority of decentralized bargaining. Indeed, this is a major argument put
forward in this paper.
The remainder of this paper is organized as follows. As a prerequisite for the
theoretical considerations the next section engages in setting the scene by very
briefly describing the institutional framework and some basic facts with respect
to wage bargaining in Germany. In addition, this section also raises some doubts
as to whether fully decentralized wage bargaining always does better than
industry wage settlements. Section III is devoted to the theoretical analysis.
Section IV puts the theoretical results in perspective and concludes.
II INSTITUTIONAL BACKGROUND AND BASIC FACTS
As a prerequisite for the following analysis, it is necessary to review briefly the
institutional regulations concerning the wage bargaining process and to add
some basic quantitative facts.
Wage bargaining in Germany typically takes place at the sectoral or industry
and regional level such as for the metal and electrical industry in northern
Baden-Wu
Èrttemberg (one of the 16 states in Germany). While the outcome of
this wage settlement concerns, strictly speaking, only union members in firms
1See, for example, Berthold and Fehn (1996).
2Fitzenberger and Franz (1994).
438 BERND FITZENBERGER AND WOLFGANG FRANZ
#Scottish Economic Society 1999
To continue reading
Request your trial