Collective Bargaining and Innovation in Germany: A Case of Cooperative Industrial Relations?

• Date: 01 January 2017
• Author: John T. Addison,Lutz Bellmann,Katalin Evers,Paulino Teixeira
• DOI: http://doi.org/10.1111/irel.12165
• Published date: 01 January 2017

Collective Bargaining and Innovation in

Germany: A Case of Cooperative Industrial

Relations?

*

JOHN T. ADDISON, PAULINO TEIXEIRA, KATALIN EVERS and

LUTZ BELLMANN

At the level of theory, the effect of collective bargaining on innovation is con-

tested. The large proponderance of the U.S. evidence clearly points to adverse

effects, but other-country experience suggests that certain industrial-relations sys-

tems, or the wider regulatory apparatus, might even tip the balance in favor of

unions. Our pooled cross- section and difference-in-differences estimates provide

some weak evidence that German collective bargaining inhibits innovation. How-

ever, in conjunction with workplace representation, there is the suggestion that it

might actually foster innovative activity.

Introduction

The topic of collective bargaining and investment in intangible (and tangi-

ble) capital has been the subject of considerable controversy for a number of

years. The debate remains unsettled, although theory has tended to look with

more favor upon the union entity if it is located in an “appropriate”institu-

tional setting. Theory has in one sense been channeled in this direction by

empirical research pointing to a sharp dichotomy between North American

findings that are almost invariably negative in respect of the union impact on

innovation capital and European research that generally points to an absence

of significant associations once one proceeds much beyond the raw correla-

tions in the data.

*The authors’affiliations are, respectively, Durham University Business School, Durham, England, and Moore

School of Business, University of South Carolina, Columbia, South Carolina. E-mail: ecceaddi@moore.sc.edu;

Universidade de Coimbra, Coimbra, Portugal. E-mail: pteixeira@fe.uc.pt; Institut f€

ur Abeitsmarkt- und Berufs-

forschung, Bundesagentur f€

ur Arbeit, N€

urnberg, Germany. E-mail: Katalin.Evers@iab.de; Friedrich-Alexan-

der-Universit€

at Erlangen-N€

urnberg, and Institut f€

ur Arbeitsmarkt- und Berufsforschung, Bundesagentur f€

ur

Arbeit, N€

urnberg, Germany. E-mail: Lutz.Bellmann@iab.de.

Addison gratefully acknowledges financial support from the Riegel and Emory Human Resource Center of the

University of South Carolina.

INDUSTRIAL RELATIONS, Vol. 56, No. 1 (January 2017). ©2016 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.

73

In the present paper, we focus on the innovative activities of German estab-

lishments over the 6-year observation window 2007–2012. Our measure of

innovation is the actual (or successful) introduction of some product or process

innovation (although we shall also investigate failure to innovate). Apart from

allowing us to consider a new output indicator, our choice of Germany was

predicated on that nation’s unique structure of cooperative industrial relations,

early research seeming to offer some confirmation of the benefits of coopera-

tion provided the level of union density is not “excessive.”

We use both extensive descriptive analysis and regression techniques to evalu-

ate the role of different institutional arrangements on innovation, while controlling

for a wide array of establishment-level observables. We also tackle unobserved

establishment heterogeneity by constructing different establishment subsamples

and then investigating differences in changes in the incidence of innovation using

appropriate comparison groups of innovating and noninnovating establishments in

combination with collective bargaining (and works council) switchers and collec-

tive bargaining stayers in a difference-in-differences framework. This approach to

isolating the causal effect of labor institutions, using changes in collective bargain-

ing status as the main identification vehicle, is new in this literature. (For another

approach, see the discussion of Bradley, Kim, and Tian [2015] below.)

The structure of the paper is as follows. First, we provide a comprehensive

statement of the theory of collective bargaining/unionism and innovation with

a view to justifying consideration of the German case, while deriving a set of

more targeted hypotheses related to that nation. Second, we examine the

empirical evidence on innovation, with the goal of drawing a distinction

between the North American evidence and the rest before examining a still

sparse extant German literature by way of scene-setting. Third, we review the

unique dataset used in this inquiry and introduce the key innovation measures

and explanatory variables. Fourth, we present a set of descriptive results on

the frequency and continuity of the different types of innovative activity and

describe the unconditional and conditional (on union/worker representation)

probabilities of an establishment having a particular type of innovation. Fifth,

our detailed ceteris paribus results are presented, together with robustness

checks. A summary concludes, the burden of which is that German collective

bargaining is not generally to be construed as inhibiting innovation and may

indeed prove beneficial when accompanied by workplace codetermination.

Theoretical Considerations

Theory suggests that collective bargaining can have positive as well as

negative effects on innovation. In the traditional model, the union-set wage is

74 / JOHN T. ADDISON,PAULINO TEIXEIRA,KATALIN EVERS AND LUTZ BELLMANN

represented as an exogenous change in the price of labor, the firm in response

adjusting employment along its labor-demand curve. In this case, the union

premium or tax is levied on labor. Union firms duly substitute away from

expensive labor. The net effect is unclear. It depends on the degree of substi-

tutability between capital and labor and the magnitude of the scale effect as

the premium filters through into higher product prices and output falls.

By contrast, the more modern view is that unions tax capital, that firms

respond unambiguously by cutting tangible and intangible capital investments,

and that the wage is endogenous. The idea is that unions expropriate part of

the quasi-rents that form part of the normal (i.e., competitive) returns to capital

but which are vulnerable to capture once investment in specialized plant and

equipment and research and development (R&D) has been made. We note par-

enthetically that R&D expenditures have been used in the literature as a key

indicator of the asset specificity of an investment (see Cavanaugh 1998).

Familiarly, such assets will continue in use as long they earn a return above

their alternative use; the more specific the asset, the bigger the scope for union

rent seeking. Of course, with the relation-specific capital in situ, higher wages

are unlikely to influence the use of the asset, but firms will anticipate reduced

returns to such capital and invest less.

This is the so-called “hold-up”problem, first analyzed by Grout (1984). In

the simple one-shot two-stage game summarized by Menezes-Filho and van

Reenen (2003: 296), the firm first chooses a level of capital (either high or

low) and in the next round the union chooses the wage (high/low). By back-

ward induction, the union will always choose a high wage in the second stage

and, knowing this, the firm will always choose a low investment strategy at

the first stage. Further, the union tax on investment will vary directly with the

specificity of the asset and its longevity. The tax would vanish were the union

able to commit itself to a low-wage strategy by posting a bond or hostage to a

third party, or where there is bargaining over investment as well as wages.

1

However, collective bargaining is repeated over time rather than being a

one-shot exercise and, abstracting from an end-game scenario (Lawrence and

Lawrence 1985), repeated games offer a solution to the hold-up problem

because opportunistic behavior can in principle be appropriately punished

(e.g., van der Ploeg 1987). An important issue in the literature has been the

degree to which unions discount the future. In particular, it has been argued

that because union members do not have property rights in the union they will

be rationally myopic and discount the future at a higher rate than shareholders.

This tendency will be reinforced by the greater influence of older workers in

1

We ignore for the moment the potential hold-up problem on the part of the employer.

Collective Bargaining and Innovation in Germany /75