Determinants of the Perceived Bargaining Power of Works Councils – Results of an Empirical Investigation in 1000 German Firms

  1. 1.  University of Duisburg-Essen

Abstract

The paper analyses the subjectively perceived bargaining power of the Works Councils, relative to the power of the management. We draw on a representative sample of telephone interviews with 1000 human resource managers and 1000 works councils in as many companies. First, we present empirical findings about the bargaining power relationship between works councils and management. Second, we identify causes for differences in power from firm to firm. All in all, our empirical results do not confirm hypotheses of the determinants of the perceived bargaining power derived from power-dependence theory. Only the degree of unionisation proved to have an impact on perceived power. Our findings indicate that subjectively perceived power is not sufficiently explained by resource control and its corresponding structural characteristics. Thus, other features of the bargaining relationship and its history as well as the characteristics of participants and the personal relationships should also be considered in future research.

1         INTRODUCTION

Power is a central factor in exchange between capital and labour. This applies to negotiations between unions and employers as well as to bargaining relationships between works councils and management. The factors determining union power have been both theoretically and empirically analysed (e.g. Mitchell 1996; Green and McIntosh 1998). Contrastingly, the power of works councils and the sources of various potential influences have to a much lesser extent been the subject of quantitative study (cf. amongst the few exceptions Witte 1980; Teulings 1987; Peetz and Pocock 2009, the latter have studied the power of workers perceived by workplace delegates). Nonetheless, the subject of works council bargaining power is significant: first, because negotiations over working and wage conditions in many countries are being increasingly shifted to company level (cf. Traxler et al. 2002) and because works councils are playing an ever greater role within collective bargaining. Second, the question of bargaining power determinants extends well beyond the specific works council institution and the negotiating relationship between works councils and management. As such, the existing procedures for measuring power remain unsatisfactory (i.e. Armstrong et al. 1977; Prasniskar and Gregoric 2002), which is also a result of the different conceptions of power (Clegg et al. 2006). Third, little study has been made of the relationship between the existing objective fundamentals of power and the subjective perception of one´s own power and that of other protagonists. Also open to question is whether the same determinants influence both objective and subjective power. Experimental findings underpin the fact that parties controlling resources important to an actor are also perceived as more powerful by the latter (Bacharach and Lawler 1976; Aguinis et al. 1994). Perceptions of power within a company and in long-term bargaining relationships are potentially less influenced by the objective or perceived control of resources, but rather are more influenced by the personal relationships of the actors and specific experiences resulting from long-term joint negotiation. This would mean, for example, that the degree of unionisation (which is evaluated here as a contributing factor of the control of resources by a works council) might have relatively little effect on the perceived power of works councils. A study answering such questions is, in our opinion, not only of interest from a theoretical and empirical perspective, but also in terms of bargaining policy. We draw on data from a survey of 1000 human resource managers and 1000 works councils in as many companies. That means, both sides within the same company were questioned on the power of works councils. In addition, we also assessed company characteristics (such as degree of unionisation, codetermination at the company level, size of works council etc.). Amongst other things, we would expect the degree of unionisation to have a significant impact on objective works council power. However, there is little clarification as to whether such variables are also linked to subjectively perceived power.

Thus, the aim of our paper is twofold. First, we present empirical findings as regards the bargaining power of the works council – power subjectively perceived by management and works council. We quantitatively measure the extent of bargaining power that each side attributes to the other, and we show differences between the perceptions of both protagonists. Second, we identify reasons for differences in perceptions of works council power from company to company.

Our contribution differs from other papers in more ways than one: First, we analyse subjectively perceived power and not objective power. In bargaining situations, perceived power is regarded as more important than objective power because decisions and behaviour are primarily influenced by perceived power (Bacharach and Lawler 1981; Walton, Cutcher-Gershenfeld and McKersie 2000). Second, we analyse the power relationship, i.e. the distribution of power between both protagonists (Emerson 1962; Lawler and Bacharach 1986). Third, in our study, power is assessed directly via interviews and not indirectly, for example, by measuring trade union density as a proxy variable for power or potential influence (e.g. Green and McIntosh 1998). Fourth, our analysis is based on a large quantitative survey and not on a small number of qualitative interviews (such as, for example, Stirling and Tully 2004 for European Works Councils). Finally, and of major importance: we assess the perceptions of both bargaining parties, whereas other studies are limited to only one side (e.g. Jirjahn and Smith 2006).

We proceed as follows: we first present our theoretical basis and develop our hypotheses, then we introduce the database and report on descriptive empirical findings and the results of bivariate and multivariate analyses. In the next step we discuss the results before finally concluding with a brief summary.

2         Theoretical background

We draw on the following definition of power: “By power is meant the ability of individuals or groups to make their own concerns or interests count, even where others resist. Power sometimes involves the direct use of force, but is almost always also accompanied by the development of ideas (ideology) which justify the actions of the powerful” (Giddens 1989: 52). The terms power and influence are applied with identical meaning in the following. Our general propositions are based on the power-dependence theory (Emerson 1962; Bacharach and Lawler 1981). The main assumptions of this theory can be formulated in the following sentences. The first sentence defines power by dependency: (1) The higher the dependency of actor A on actor B, the less power A is able to exercise over B. This definition corresponds with Giddens´ idea of power as being the capacity of an actor to realise his interests. The second and third one are hypotheses: (2) The dependency of actor A on actor B is the greater, (i) the more value A places on specific objects controlled by B, and (ii) the less opportunity A has to gain hold of these objects outside the scope of the A-B relationship. The same thing applies in B´s case. (3) The more A is dependent on B, the more likely B will be in a position to enforce his or her demands, thus overcoming A´s potential resistance to these (Emerson 1962: 32). Power is a function of interests, control and alternative resource sources. As such, we are able to establish that power is a relational construction. Therefore, it is not a question of how much power a given party generally has, but rather how relative such power is to the other respective party in a specific exchange ratio. An initial, more specific conclusion is: works council bargaining power is based on the dependency of the management on the works council and, respectively, the resources the works council controls. We expect (in the case of similar or the same interests) works council power to be greater in companies in which it controls more resources. We also presuppose that the greater the available means to control resources, the greater resource control will be. As such, it can be assumed that a greater degree of unionisation would make the mobilisation of employees easier and, as a consequence, contribute to better control of labour as a resource. Furthermore, we also assume that such potential is also readily employed. These assumptions presuppose, that there is a relationship between trade unions and works councils. This is the case in in Germany, where the proportion of works councillors’ organized in unions is very high and the trade unions need the works councils to mobilize the employees in a firm.

So far the conception of power and dependence applied here has many parallels in economic bargaining theories (e.g. Chamberlain 1951): It is the supply and demand of specific resources that play the dominant role here. But the question is to what extend subjective power plays a role. An “objectivistic” version would be unproblematic in terms of objective power, that is, if objective power alone determined the result of bargaining. However, the manipulation of perceptions is an important factor in bargaining processes (Bacharach and Lawler 1981; Walton and McKersie 1965, Walton, Cutcher-Gershenfeld and McKersie 2000). “Our perspective assumes that conveying an impression of power will yield the same consequences as having ´real´ power; manipulating an opponent´s perception of the power relationship will have the same effect as an actual change in that relationship” (Lawler and Bacharach 1986: 195). “Because fundamentally “bargaining power is subjective in nature” (Lawler and Bacharach 1986: 194 et seq.), with regard to power we also have to take perceptions and subjectivity into account (cf. for subjective measures of power also Peetz and Pocock 2009: 633f.).

The fact that we place special emphasis on perceived power has consequences for our hypotheses. We cannot assume beforehand that determinants of objective power also influence subjective power. For example, a high degree of unionisation in a firm may influence objective works council power; however, that is not to say that this power source is also perceived by the works council and contributes to the perception of greater power.

3         Data and hypotheses

We have access to data that will help us initially specify perceived works council power and identify structural determinants or correlations of power. With financial support from the Hans-Boeckler-Foundation (www.boeckler.de), in 2005 a representative telephone interview was carried out with 1000 human resource managers and 1000 works councils in as many companies. From the Hoppenstedt database, a stratified random sample categorized into four size classes (100-199, 200-499, 500-999, more than 1000 employees) was drawn up. Firms with at least 100 employees in all industries were interviewed (excluding agriculture and forestry, fishing and fish farming, education as well as suppliers of other public and personal services). Firms without a works council were excluded. In every company the human resource manager responsible for signing works agreements and the corresponding member of the works council were identified and interviewed.

The central focus of the survey was to record the opinions of managers and works councils on the decentralisation of collective bargaining. Additionally, we also have access to a number of variables that we are able to use as power determinant indicators. These variables have a common denominator in that they all impact on resource control. In our analysis we include the following variables, a plus or minus in brackets (+, -) indicates the supposed direction of its effect on perceived power.

The trade union density of the firm (+) and the works council (+); the existence of employees´ representatives on the supervisory board (in Germany the respective codetermination laws apply only for private and public limited companies with at least 500 employees), (+); age of the works council (+); the company´s commitment to a collective bargaining agreement (+); the personnel structure of the company (percentage of highly trained employees (+), percentage of males (+)); the size of the company (+); industry (manufacturing sector (+) vs. non-manufacturing sector); the location of the company (western part of Germany (+) vs. eastern part); and the company´s profit situation (+). All these factors may have an effect on the power relationship. The effects are interpreted as based on the control of resources and the dependence on these resources. To name but two examples: (i) The higher the trade union density of the firm, the more the works council is able to control an important resource: namely, “human capital” and employee motivation. (ii) In the eastern part of Germany with its higher unemployment rate, workers are easier to substitute, therefore the degree of dependency of the employer on workers and also their representative body is lower.

We not only predict the sign of the coefficients of our variables, but also differences in the relative magnitude. As perceived power is the subject, we expect a greater impact from those variables that the protagonists perceptively associate with power. Accordingly, a strong impact should be particularly apparent in the case of variables that influence power and/or are used as power indicators within the protagonists´ “lay theories”. We also anticipate that variables that are practically regarded as characteristics of the protagonists or which are directly associated with co-determination within the company will also demonstrate a strong impact. As such, the degree of unionisation (mobilisation capability), the size of the works council (degree of experience and expertise), age of the works council (the older the works council, the more experienced and, thus, the more influential it will be) and an employee-elected supervisory board member should all have a stronger impact on power than the size of the company, personnel structure and profit situation.

3.1        Dependent variable: power relationship between works council and management

We assessed the power (respectively influence) of the works council by asking both sides the following questions. First, we asked a summary question assessing the power in comparison to other firms: “How great is the influence of the works council compared to other companies?”.  The interviewees had the possibility of answering with “below average”, “average” and “above average”. Second, we assessed the power relation between works councils and managements: “How would you generally describe the bargaining power relation between the management and works council in your company? You can award a total of 100 points. For example, a strong power overbalance on one side would be 90 to 10, while 50 to 50 means that both sides have the same bargaining power.” Given that both the management and the works council were asked all the questions, we are able to include the level of agreement between both parties´ perceptions within our analysis.

3.2        Independent Variables

The trade union density of the firm and of the works council is measured as the percentage of union members within the entire company workforce (WC). (Whether the management (M) answered the question or the works council (WC) is illustrated in brackets.) The existence of employees´ representatives on the supervisory board is documented via a present/not present question (WC). The company´s commitment to a collective bargaining agreement is also answered through a negative/affirmative question (M). The personnel structure of the company is documented via the percentage of highly trained employees and males (WC). The size of the company is attained through a dummy variable (more than 199 employees = yes; 200 is the median size and also the threshold for at least one full-time released works council member) (M). The industry (manufacturing sector = yes) and the location of the company (western part of Germany = yes) are also measured using dummy variables. As regards the age of a works council, works councils founded after 1990 are classified as “young” and, as such, are deemed to have relatively little experience (WC). We also use a simple dummy variable in the case of profit (profit = yes) (M).

4         Descriptive Findings

Initially, we measured perceived power using a question that established the power of a works council in comparison to that of works councils from other companies. Then we measured the distribution of perceived power between works councils and managements using a percentage scale. Finally, we were ultimately in a position to record to what extent works councils and managements differed in their perception of power.

Management judges works council influence to be lower than its own influence (Table 1). 27.3% of managers state that their works council has more influence in comparison to other companies, while the proportion of works councils that see their power as being comparatively greater is almost ten percent higher at 36.5%. On average, the relative strength between the company parties is similar on both sides (however, even here a slightly higher works council power perception is evident): the interviewed managers award the works councils 37.2 points (and management 62.8 points), while the works councils award themselves 40.5 points.

 

Table 1: Perceived power of the works council (from the perspective of the management and the works council)

Perceived power of the works council

Management

Works council

Power of the works council compared to other companies
(percentage “above average”)

27.3

(n= 972)

36.5

(n= 986)

Power relationship between management and works council
(Points for the works council, out of 100; mean (in brackets median/standard deviation))

37.2

(40/10.5)

(n= 995)

40.5

(40/10.6)

(n= 997)

 

Consequently, there is consensus between management and works councils that the works councils have inferior power. On the other hand, on average, works councils give a higher assessment of their own power than the management does. In the following analyses we draw on bivariate correlations and multivariate OLS and logistic regression models.

 

Table 2: Determinants of the perceived power of the works council (correlations)

Independent variables

Influence above average, perceived by
Works council

Influence above average, perceived by
Management

Relative influence, perceived by
Works council

Relative influence, perceived by
Management

Trade union density of the firm (%)

.12**

.20**

.14**

.15**

Trade union density of the works council (%)

.06

.14**

.12**

.13**

Number of works council members

.19**

.20**

.10*

.06*

Employees´ representatives on the supervisory board (yes=1)

.18**

.15**

.07*

.02

Age of the works council (founded before 1990=1)

.11**

.09**

-.07*

-.01

Binding collective agreement (yes =1)

.10**

.09**

.05

.05

Location of the company in western part of Germany (yes=1)

.11**

.06

.03

.06

Percentage of highly trained employees

-.01

-.01

.06

.12**

Percentage of male employees

.06

.03

-.03

.12**

Manufacturing sector (yes=1)

-.09**

.03

-.03

.09**

Company´s profit situation (profit=1)

-.02

-.01

-.02

-.02

Size of the company (more than 199 employees = 1)

.13**

.15**

.02

.04

**= p ≤ 0.01; * = 0.01 < p ≤ 0.05

 

All correlations are relatively low (Table 2). Only one correlation achieves r = .20, while all others are lower. Although the directions of the correlations correspond predominantly with our hypotheses, some unanticipated coefficients were also observed. Significant differences also exist between the correlations depending on the form of power variable and who assessed the power – the managers or the works councils. However, a uniformly interpretable pattern is scarcely identifiable. The relatively highest correlation levels are between the self-assessments of works councils and the degree of unionisation (r = .12 and .20), the size of the works councils (r = .19 and .20) and the presence of an employee-elected representative on the supervisory board (r = .18 and .15). The degree of unionisation in the works councils plays, at most, a minor role. As regards the manager assessments, the unionisation of employees as a whole, unionisation in the works councils and also the personnel structure had some impact.

The question must now be asked whether multivariate analyses show similar results. We carried out two groups of regression analyses: first, OLS regressions with relational power as a dependent variable, then a binary logistic regression in which the dependent variable is an assessment of whether the works councils possessed above-average influence in comparison with other companies. We differentiate between the management and works council assessments in the case of both dependent variables.

 

Table 3: Determinants of the perceived power of the works council (OLS regression)

Independent variables

Relative influence, perceived by
Works council

 

Relative influence, perceived by
Management

 

 

B

Beta

B

Beta

Trade union density of the firm (%)

.048**

.12

.033

.08

Trade union density of the works council (%)

.008*

.02

.007

.02

Number of works council members

.114

.04

.127

.05

Number of works council members

.114

.04

.127

.05

Age of the works council (founded before 1990=1)

1.071

.05

.735

-.03

Binding collective agreement (yes =1)

.094

.00

.587

.03

Location of the company in western part of Germany (yes=1)

1.076

.03

-1.662

-.05

Percentage of highly trained employees

.031

.07

.031

.07

Percentage of male employees

.016

.03

-.021

-.04

Manufacturing sector (yes=1)

-2.394*

-.09

.385

.02

Company´s profit situation (profit=1)

-.265

-.01

.092

.00

Size of the company (more than 199 employees = 1)

-1.088

-.05

.513

.02

Constants

37.490**

 

33.233**

 

R-squared

.037

 

.037

 

Adj. R-squared

.022

 

.022

 

n=

811

 

810

 

**= p ≤ 0.01; * = 0.01 < p ≤ 0.05

 

The results relating to the relative influence perceived by works councils (Table 3) show only one single effect that is worth mentioning on account of the magnitude and the significance of the coefficients,: namely, the degree of unionisation. All of the other variables have hardly any (significant) influence on perceived power.

A similar pattern was also observed in the case of the second power variable (above-average influence, Table 4). The impact of the degree of unionisation is positive and significant for both management and works council. Otherwise, there were no significant relevant effects that were similar for both parties. However, the size of the works council (from a works council perspective) and the existence of co-determination in the supervisory board (from a manager perspective) show a positive and significant connection to perceived power.

The proportion of explained variance is low in all the models.

 

Table 4: Determinants of the perceived influence (above average) of the works council (logistic regression)

Independent variables (Coefficients: odds ratio)

Influence above average, perceived by
Works Council

Influence above average, perceived by
Management

Trade union density of the firm (%)

1.012**

1.020**

Trade union density of the works council (%)

.996

.998

Number of works council members

1.042

1.060*

Employees´ representatives on the supervisory board (yes=1)

1.767*

1.383

Age of the works council (founded before 1990=1)

1.268

1.180

Binding collective agreement (yes =1)

1.232

.964

Location of the company in western part of Germany (yes=1)

1.616

1.067

Percentage of highly trained employees

1.000

.997

Percentage of male employees

.997

.999

Manufacturing sector (yes=1)

.546*

1.030

Company´s profit situation (profit=1)

.890

.998

Size of the company (more than 199 employees = 1)

1.055

1.503

Constants

.451*

.079**

Nagelkerke´s R-squared

.103

.125

n=

804

790

**= p ≤ 0.01; * = 0.01 < p ≤ 0.05

 

In addition to the models presented we estimated several other models with different specifications. We did this because company size is correlated with the number of works council members (r = 0.51) and also with the existence of an employee representative on the supervisory board (r = 0.31) (the correlation between the number of works council members and an employee representative on the supervisory board is r = 0.46). But all in all, the results of these regression models were nearly identical.

5         Discussion

To all intents and purposes, our empirical results do not confirm our hypotheses. Only the degree of unionisation proved to have a recognisable, easily interpretable impact on perceived power that was valid for both sides – works councils and managers. Although other factors also have an impact (for example, the size of the works council and presence of employees´ representative on the supervisory board), these effects are only perceived by one respective side (mainly on the works council side). A cautious statement could be made that the correlations concerning features of the works council itself (or that are directly linked with codetermination, such as a codetermining member on the supervisory board) are higher; however, such confirmation is somewhat fragile.

A range of explanations exist concerning the non-confirmed hypotheses. First, measurement problems on account of interviewee errors cannot be excluded. Second, when the implication is made – as is the case here – that the manipulation of perceptions is part and parcel of the bargaining process, then it stands to reason that the interviewees would not portray either the power of the other parties nor their own power correctly – particularly if they assume that the evaluations could become known to the other side. Third, perhaps perceived power is also heavily determined by bargaining process history, the longstanding experience of the negotiating parties or the characteristics of principal participants in the negotiation. We do not have the data required to test such assumptions.

Overall, our findings indicate that measurements of subjective power assessment should be approached with great caution. The question regarding the determinants of perceived power also remains open. Although theoretically well-founded hypotheses regarding objective power can be developed using the power-dependence theory, according to our findings subjectively perceived power is not sufficiently explained purely by resource control and its corresponding structural characteristics. There is every indication that consideration also has to be afforded to the features of the bargaining relationship and its history as well as the characteristics of participants and the personal relationships involved.

References

Armstrong, K.J.; Bowers, D.; Burkitt, B. (1977): The Measurement of Trade Union Bargaining Power. In: British Journal of Industrial Relations, 15: 91-100.

Aguinis, H.; Nesler, M.; Quigley, B.; Tedeschi J., (1994): Perceptions of Power: A Cognitive Perspective. In: Social Behavior and Personality: An International Journal, 22: 377-384.

Bacharach, S.B.; Lawler, E.J. (1976): The Perception of Power. In: Social Forces, 55: 123-134.

Bacharach, S.B.; Lawler, E.J. (1981): Bargaining: Power, Tactics, and Outcomes. San Francisco.

Chamberlain, N.W. (1951): Collective Bargaining. New York.

Clegg, S.R.; Courpasson, D.; Phillips, N. (2006): Power and Organizations. London et al.

Emerson, R.M. (1962): Power-Dependence Relations. In: American Sociological Review, 27: 32-41.

Giddens, A. (1989): Sociology. Cambridge.

Green, F.; McIntosh, S. (1998): Union Power, Cost of Job Loss, and Workers´ Effort. In: Industrial and Labour Relations Review, 51: 363-383.

Jirjahn, U.; Smith, S.C. (2006): What Factors Lead Management to Support or Oppose Employee Participation: With and Without Works Councils? Hypotheses and Evidence from Germany. In: Industrial Relations, 45: 650-680.

Lawler, E.J.; Bacharach, S.B. (1986): Power Dependence in Collective Bargaining. In: Advances in Industrial and Labor Relations, 3: 191-212.

Mitchell, N.J. (1996): Theoretical and Empirical Issues in the Comparative Measurement of Union Power and Corporatism. In: British Journal of Political Science, 26: 419-429.

Peetz, David; Pocock, Barbara (2009): An Analysis of Workplace Representatives, Union Power and Democracy in Australia. In: British Journal of Industrial Relations 47: 623-652.

Prasnikar, J.; Gregoric, A. (2002): The Influence of Workers´ Participation on the Power of Management in Transitional Countries: The Case of Slovenia. In: Annals of Public and Cooperative Economics, 73: 269-297.

Stirling, J.; Tully, B. (2004): Power, Process, and Practice: Communications in European Works Councils. In: European Journal of Industrial Relations, 10: 73-89.

Teulings, Ad W.M. (1987): A Political Bargaining Theory of Co-determination - An Empirical Test for the Dutch System of Organizational Democracy. In: Organization Studies, 8: 1-24.

Traxler, F.; Blaschke, S.; Kittel, B. (2002): National Labour Relations in Internationalized Markets: A Comparative Study of Institutions, Change, and Performance. Oxford u.a.

Walton, R.E.; Cutcher-Gershenfeld, J.E.; McKersie, R.B. (2000) Strategic Negotiations: A Theory of Change in Labor-Management Relations. London.

Walton, R.R.; McKersie, R.B. (1965): A Behavioral Theory of Labor Negotiations. New York, London.

Witte, E. (1980): Das Einflusspotential der Arbeitnehmer als Grundlage der Mitbestimmung. Eine empirische Untersuchung. In: Die Betriebswirtschaft, 40: 3-26.

Reviews

License

This article and its reviews are distributed under the terms of the Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and redistribution in any medium, provided that the original author and source are credited.