Abstract
Drawing on questionnaire-based survey data comprising responses from the members of 14 trade union organizations based in 12 European countries, this article explores the impact of workplace representatives on perceptions of the performance of trade unions among members. The article shows that large numbers of members are dissatisfied with the quality of representation, even if a workplace representative is located at their place of work. When there is a workplace representative present, it is apparent that members view the performance of the union to be superior to that of management on a wide range of workplace tasks. When there is no workplace representative present, the performance of management is generally superior to that of the union.
For most trade unionists the workplace is the location of their unionism and the workplace representative is key to generating collective identities. Structural shifts in the labour market have contributed to declines in the coverage of workplace representation. Where workplace representation remains, the decentralization of bargaining has increased the range of issues handled by, and hence the workload of, workplace representatives. Furthermore, new management practices, including direct communication with employees that often bypasses workplace representatives, may directly challenge their interpretation of events and may undermine their position. This article charts the coverage of workplace representation and the views of unionists towards workplace representation in 14 union organizations based in 12 European countries. Running through the article are two arguments. First, the coverage of workplace representation is far from universal and where a representative is present many unionists argue that more representatives are necessary to represent members. Secondly, members with a representative present at the workplace rate the performance of the union to be superior compared to their counterparts with no representative and to be superior to that of management: that is, the presence of workplace representation is a prerequisite to any form of union renewal.
With the weakening of the political and bargaining position of unions two inter-related themes resonate throughout the literature on how unions may reverse the decline of recent years: renewal and the maintenance or regeneration of power. The activities of workplace representatives are central to union practices within these areas. The initial purpose is thus to highlight the influence of workplace representation in these areas.
The combination of institutions, state and employers’ policies, and union structures and identities account for the variation in union strategic choices generally, and on union renewal strategies specifically (Frege and Kelly, 2004: 31–44). Most union renewal strategies are underpinned by two assumptions. First, raising membership or density levels is important, but insufficient, for renewal. Secondly, it is necessary to develop workplace representation in order to ensure that membership gains may be secured and that workplace organization can become self-sustaining (Markowitz, 2000; Simms et al., 2013: 7–13). Workplace representation is thus both an object and a means of union renewal: an object in the sense that its development is a purpose of union renewal; a means in so far as workplace representatives are required to sustain renewed workplace organization.
The implications of union renewal policies for workplace representatives in single channel and dual systems are wide-ranging. Within single channel systems the independence from management of shop steward organization was initially derived, in part, from high levels of union density and strikes (Hyman, 1996) and consolidated in some instances by managerial incompetence (Gallie and Rose, 1996). The reassertion of managerial prerogative and the use of human resource management (HRM) techniques to bypass workplace representation subsequently exposed limitations in shop steward organization (Purcell and Ahlstrand, 1994). These limitations were accentuated in the UK where declines in the level of union density and strike activity, coupled to the terms of the legislation enacted during the 1980s and early 1990s, resulted in a greater dependence on managerial acquiescence for the continued presence of shop stewards, particularly in the private sector (Charlwood and Forth, 2009). Other single channel systems fared better, as legal underpinning offered more support to shop steward organization. The Ghent system facilitates the maintenance of high levels of union density, which, in turn, ensure employers engage with workplace representatives. Similarly, the Swedish Förtoendemannalagen (Workplace Union Representation Act) of 1974 specified the duties of workplace representation and allowed meetings in working time, thus consolidating workplace organization. Even with such legal support, the confidence of Swedish union members in trade unions fell markedly between 1980 and 1990 (LO quoted in Kjellberg, 1992: 123–124).
The legal underpinning of works councils in dual systems has also had a marked impact on the character of union renewal policies. In Germany, for example, it is unlawful for employers to prevent the establishment of works councils, if requested by workforces, and the majority of works council seats are occupied by trade unionists. The capacity of employers to resist a union presence is thus severely limited. Furthermore, the efficacy of works councils is shown to be dependent upon resources provided by trade unions, notably in the form of research, training and legal expertise (Müller-Jentsch, 1995). Although a more politically distant relationship between works councils and unions persists in the Netherlands (Visser, 1995), trade unions and works councils in Austria and Germany have developed a relationship of mutual inter-dependence (Traxler, 1998; Behrens, 2009). In countries where the establishment of works councils depends on an initiative taken by employees, works councils have always been established in only a minority of eligible workplaces. In Germany, for example, works councils are found at 10 per cent of eligible workplaces and cover between 38 per cent (East Germany) and 45 per cent (West Germany) of private sector employees. The proportion of works councillors in Germany that are not union members, however, has risen since the late-1960s (Ellguth and Kohaut, 2010), a trend reproduced in the Netherlands (Visser, 1995) and Austria (Traxler, 1998), which, as unions are reliant on works councillors to organize new members, presents a challenge in the context of union renewal initiatives. In short, in both single channel and dual systems of representation union renewal initiatives place demands on workplace representatives and raise questions about the capacity of workplace representatives to articulate effectively the interests of trade union members while also deepening workplace organization.
A second theme that resonates throughout recent literature on union responses to decline concerns the issue of power. One element of trade union power is institutional, which, to a degree, is a function of the extent and character of the legal underpinning of workplace representation (Hancké, 1993). The legal underpinning of workplace institutions is common to Austria, Germany, the Netherlands and Spain, where dual systems are in operation, but is not always present where single channel systems operate. Workplace representation is central to a second element of trade union power: organizational power. While a high level of union density is often viewed as prerequisite to the deployment of union power, issues of unity, mobilization and solidarity influence the ‘social capital’ of unionism whereby organized networks of unionists provide mutual support within and outside of the workplace and generate organizational power (Jarley, 2005). Workplace representation is essential to the generation of union social capital. Earlier studies demonstrated that workplace representatives in single channel and dual systems were able to represent and mobilize members, to organize collective and solidaristic workplace organization and, on the basis of their relations with members, to secure concessions from management (Korpi, 1978; Brandt et al., 1982).
The capacity of representatives to generate social capital among unionists may be questioned on four counts. First, shifts in the labour market effectively moved employment from areas of union strength to private sector services where workplace representation is more rudimentary (Dolvik and Waddington, 2005) and the structural power of unionized employees tends to be more limited (Wright, 2000). Associated with these shifts is the employment of more diverse workforces, which has generated a ‘crisis of interest aggregation’ (Müller-Jentsch, 1988: 177–178) as unions are required to accommodate diverging and often competing interests. Furthermore, as older white male manual workers are over-represented as shop stewards and works councillors, the assumption that workplace representatives could represent an increasingly diverse workforce has come under increasing scrutiny (Briskin, 1999; Healy et al., 2004).
Secondly, the decentralization of bargaining has effectively increased the workload of representatives. While managers have tended to ‘drive’ decentralization processes, the character of such processes varies markedly (Keune, 2011). Although industrial bargaining formally remains in place throughout much of western continental Europe, the range of issues handled at company and workplace levels has increased as managers have sought to take decisions as near as possible to their point of effect and companies have either left or failed to join employers’ associations. Although policy initiatives intended to establish industrial bargaining were implemented in several eastern European countries after 1990, employer resistance restricted the coverage of bargaining and, where it is present, bargaining is primarily limited to company level (Bohle and Greskovits, 2012; Meardi, 2012). In both single channel and dual systems, decentralization thus introduces additional challenges for workplace representatives, makes meeting members’ expectations more difficult, requires workplace representatives to handle a wider range of issues, limits the access of workplace representatives to central management decision-makers and ‘complicates’ articulation between workplace and union (Waddington, 2001).
Thirdly, the generation of social capital at the workplace relies on the responsiveness of workplace representatives to the concerns of constituents and on the participation of members in decision-making. Changes in labour markets have led many to question whether workplace representation was sufficiently sensitive to the concerns of diverse memberships. In single channel and dual systems member attendance at decision-making meetings is low or has declined and a vast number of elections to representative positions are not contested. Furthermore, even at the zenith of workplace organization elements of bureaucratization that contrasted with a democratic ideal were detected (Hyman, 1979). These elements ‘distanced’ workplace representatives from their constituents as the former assimilated a concern for the preservation of the institutions of workplace representation.
Fourthly, a ‘crisis of workers’ loyalty to trade unions’ was identified (Müller-Jentsch, 1988: 177–178) as arising from management policies designed to promote employee commitment and/or substitute management for trade unions as a source of advice (Bratton, 2001: Heery et al., 2004). Direct communication between managers and employees that bypasses workplace representatives is an additional feature associated with HRM that challenges the role of the workforce representative (Boxall et al., 2007: Dundon et al., 2005). More effective managerial communications challenge the capacity of local representatives to mobilize power resources, particularly when such communications promote workers questioning issues of solidarity and collective identity (Lévesque and Murray, 2010). In order to encourage loyalty to the union, workplace representatives are required to compete more intensely with management on communications. To examine these issues the article comprises three further sections, which review the methods used to collect the data, present the data and assess their implications.
Methods and sample composition
Surveys conducted between 2005 and 2010 are the source of the data that follow. The survey distribution within each union was conducted independently, as the negotiation of access and time management issues effectively precluded simultaneous distribution across all of the participating trade unions. The basic questionnaire design was sustained throughout the research with changes made only to accommodate national institutional variation. The questionnaires were piloted to ensure that potential respondents understood what was being asked of them and whether they had the information at their disposal to answer the questions in the survey.
Variation in the size of the distribution within each union was a function of cost, membership size and the quality of membership records. Within several participating unions sections of the membership were selected as target groups for the survey. The target groups were selected by representatives of the participating unions and the author on the basis that they were the subject of a union organizing initiative, were considered ‘difficult’ to organize, more information on them was deemed useful within the participating union, and/or they were employed in private sector services. No claim is thus made here that the results are representative of the entire unions that participated in the survey. For each participating union unemployed and retired members were excluded from the survey. Details of each distribution and a sketch of the participating unions are provided in Table 1.
The sample and survey distribution.
With two exceptions the questionnaires were distributed by post to union members at the address held by the union. Completed questionnaires were returned either directly to the author or to the head office of the participating union and then sent unopened to the author. The first exception to this general procedure was CC.OO where the questionnaire was distributed electronically to members of four of the federations that had submitted their email addresses. There is no doubt that this procedure resulted in a skewing of the sample of returns towards younger members: median age of CC.OO respondents was between 37 and 38 years whereas that for all participating unions was between 43 and 44 years. The second exception involved BBDSZ, KASZ and Solidarność. These organizations did not hold membership records and thus there was neither the means to create a representative sample of the members nor to distribute the survey centrally. In consequence, the questionnaires were distributed by students to members at their place of work, collected by the students and then returned to the author via the head office of the union. The implication of this approach was that the questionnaire returns tended to be drawn from large workplaces within large Hungarian and Polish cities. The sample thus cannot be regarded as representative regarding individual or workplace characteristics. To establish the priorities of members from eastern Europe was considered sufficient justification to proceed on this basis.
Two further remarks concerning trade union mergers and the participation in the research of union confederations are relevant to Table 1. Since participating in the survey three unions have merged with the result that they no longer exist as independent organizations: RBF merged in 2006 with Fagligt, Fælles, Forbund (3F); Sif merged with Handelastjänstemannaförbundet to form UNIONEN in January 2008; and TU merged to form Ammattiliitto PRO in January 2011. In each of these cases the confederal affiliation of the post-merger union is the same as that of the participating union. In addition, both CC.OO and Solidarność are union confederations. In the case of CC.OO four federations participated in the survey, each of which organized in one of the segments of the economy listed in Table 1. Solidarność comprises about 8300 company trade unions, 37 regional unions and 16 industrial unions that represent workers throughout the Polish economy. The survey was distributed among these three elements of Solidarność in the sectors mentioned in Table 1.
Here shop stewards in single channel systems and works councillors in dual systems are treated as workplace representatives. 1 Union workplace representatives established by some unions that operate in dual systems are thus excluded from the category ‘workplace representative’. 2 The countries regarded as operating single channel systems from which participating unions are drawn include Denmark, Finland, Hungary, Italy, Poland, Norway, Sweden and the United Kingdom, whereas those operating a dual system include Austria, Germany, the Netherlands and Spain. In Hungary and Spain provisions allow for shop steward and works council representation. Hungary is treated as a single channel system in so far as the works councils established under the Labour Code of 1992 had only information and consultation rights, were often integrated into local union organization, tended to be established only where union organization was already present, and, in the main, left bargaining with management to the local union (Tòth, 1997). In contrast, Spain is treated as a dual system in that the Comite de Empressa (workers’ committee) comprises representatives elected by the workforce in enterprises with 50 or more employees; has the right to negotiate company or workplace agreements, and to information and consultation; and the responsibility of ensuring compliance with external regulatory changes, such as on social security or employment legislation (Hamann and Martinez Lucio, 2007).
The views of union members on workplace representation
The views of trade union members on workplace representation are examined in three stages. The first stage establishes the coverage of workplace representatives and the extent to which union members regard workplace representation as satisfactory where representatives are present. The second stage explores unionists’ views of union-management relations and the quality of the performance of the union at the workplace. The third stage compares the performance of workplace representation with that of management on a range of communication and related issues.
The analysis features comparisons between the views of unionists at workplaces where a representative is present and the views of unionists employed at workplaces with no representative. For the unions in the sample based in single channel systems 57.7 per cent of unionists reported the presence of a shop steward whereas 66.7 per cent of unionists in dual systems reported the presence of a works councillor. These figures do not indicate that the legal underpinning of works councils in dual systems necessarily results in a higher coverage, as there is marked sectoral variation in the composition of the two samples. Three limitations are acknowledged regarding the approach used here for shop stewards and works councillors. First, the categories shop steward and works councillor are analysed as undifferentiated: that is, no account is taken of the sex, age or other demographic characteristics of the representative and how these may influence relationships with unionists and the articulation of different interests. Secondly, distinctions are not drawn between unionized and non-unionized works councillors, although it is acknowledged that unionized works councillors are more likely to promote the union than their non-unionized counterparts. Union members with a works councillor present at their workplace reported that 77.8 per cent of works councillors were unionized. Thirdly, no attempt is made here to incorporate an analysis of the impact of workplace union representatives in dual systems, which may supplement that of works councillors.
The coverage and perceived quality of workplace representation
Table 2 shows the coverage of workplace representation and the extent of satisfaction among unionists with the performance of workplace representatives. The results from Hungary and Poland should be treated with some caution, as the method of survey distribution is likely to result in a disproportionately large proportion of questionnaires being distributed at sites where workplace representatives are present.
The coverage and quality of workplace representation.
a The ‘proportion of members content with workplace representation’ was calculated by multiplying the proportion of members that reported the presence of a shop steward or works councillor at their workplace by the proportion of members reporting that the shop steward/works councillor adequately represented the members.
Four points are immediately apparent from Table 2 regarding the coverage of workplace representation. First, the coverage of workplace representatives for most unions is in the range 56.3 per cent (CC.OO) to 91.7 per cent (GPA). For the majority of unions between 25 per cent and 30 per cent of members do not have a representative present at their workplace. Secondly, the lowest coverage of workplace representation is among the hotel and restaurant workers of RBF (31.8 per cent) and the bank and insurance workers of UNITE (32.2 per cent), suggesting that sector and workplace size are influential factors on coverage rates. Further evidence supporting this argument is available in Ver.di where the coverage rate of works councillors for fire fighters in the public sector is 90.1 per cent compared to 65.2 per cent for Ver.di members in retail, banking and insurance. Thirdly, it is only in TU, UILCA, BBDSZ and Solidarność that a larger proportion of women than men report the presence of a representative at their workplace. In all of the unions operating in dual systems and most of the unions operating in single channel systems men are more likely to report the presence of a representative at their workplace. Campaigns to ensure the presence of a representative at every workplace where members are organized are thus less developed in feminized segments of the labour force. Fourthly, the argument that workplace representation in dual systems has been more resilient than shop steward organization in single channel systems (Jacobi et al., 1992) receives only partial support. The direct comparison of UNITE and Ver.di members in banking and insurance illustrates the relative resilience of workplace representation in dual systems as twice as many Ver.di members (65.2 per cent) than UNITE members (32.2 per cent) report the presence of workplace representatives. Comparison with UILCA, where 66.6 per cent of members have a shop steward at their workplace and membership is also organized in banking and insurance, shows that workplace representation in single channel systems can be resilient. The distinction between single channel and dual systems is thus insufficient to explain differences in the coverage of workplace representation.
If a workplace representative is present, respondents were asked to indicate whether there are sufficient workplace representatives. The range of responses citing that there were sufficient representatives varied from 51.2 per cent in UNITE to 82.1 per cent in Bondgenoten. Even if a workplace representative is present substantial numbers of members thus think that there is an inadequate number of workplace representatives. There is no straightforward relation between the presence of a workplace representative and members’ views as to whether there are enough workplace representatives. Although the coverage of workplace representatives in RBF is only 31.8 per cent, for example, almost 80 per cent of members with a representative at their workplace thought that there were sufficient numbers, putting RBF towards the top of the range. In contrast, UNITE is at the bottom of the range on both the coverage and the numbers of representatives.
A means to ensure the presence of a sufficient number of workplace representatives is most pressing for women unionists. In only three of the participating unions (BBDSZ, KASZ and Bondgenoten) did a greater proportion of women than men think that there were enough representatives. Women were thus less likely to have a workplace representative present at their workplace and were less likely to think that there were enough representatives if one was present.
Given the number of members who work alongside workplace representatives, but think that there is an insufficient number of representatives, it is not surprising that a substantial minority of members take the view that workplace representatives do not adequately represent unionists. Between 50.8 per cent (TU) and 82.7 per cent (Solidarność) of members with a representative at their workplace thought that representatives adequately represented members. A range of factors, that are not mutually exclusive, may explain this result, included among which are the unwillingness of members to take on representative responsibilities resulting in low numbers of workplace representatives, the availability and/or willingness of representatives to undertake training, a reluctance among managers to allow sufficient facility time, and high expectations among members about what can be achieved in current economic and political circumstances.
In 12 of the 14 participating unions (exceptions BBDSZ and Solidarność) men are more likely than women to report that representatives adequately represent members. In the majority of unions women were also less likely to have a representative present at the workplace and were less likely to think that there were enough representatives if one was already present. Women emphasize support from unions more than their male counterparts as a reason for both joining and retaining union membership (Klandermans, 1984; Waddington and Whitston, 1997). The dissatisfaction among women compared to men with representation, even if a representative is present at the workplace, introduces an additional feature to the issue of workplace support from unions and raises questions about the nature of this support.
The final row of data in Table 2 reports the proportion of members that are content with workplace representation defined as there is a representative present at the workplace and members think that they are adequately represented. Putting aside the results from BBDSZ, KASZ and Solidarność for methodological reasons, in only three unions (UILCA, GPA and Bondgenoten) were more than half the members content with workplace representation. In five unions operating in single channel systems (RBF, TU, NITO, HK and UNITE) and the four federations of CC.OO that operate in a dual system fewer than 40 per cent of members are satisfied with workplace representation. In only three unions (TU, BBDSZ and Solidarność) is a greater proportion of women than men content with workplace representation. The extent of dissatisfaction among trade unionists with workplace representation raises questions about the issues on which satisfaction or dissatisfaction is expressed. Stages two and three of the analysis address these questions.
The impact of representatives on perceptions of workplace industrial relations
Stage two compares the views of unionists with and without a representative at their workplace on issues concerned with union-management relations and on the performance of the union on a number of basic tasks. Five statements on union-management relations are identified in the upper panel of Table 3. Overall, members are more positive in assessing these statements when there is a representative present at their workplace: in only 16 of the 70 cases were members with a representative less positive than their counterparts with no representative present. 3 Furthermore these 16 cases were distributed across the range of industrial relations systems represented here. Two such cases were in CC.OO based in a dual system, seven cases occurred in Nordic unions, four in the unions based in eastern Europe, two in UILCA and one in UNITE. The greater political ‘distance’ between Dutch works councils and trade unions compared to that in Austria and Germany is thus not reflected in the perceptions of Bondgenoten members.
The impact of representatives on the perceptions of members on workplace industrial relations single channel systems.
Note. For the panel of the table on union-management relations respondents were asked to answer ‘strongly agree’, ‘agree’, ‘disagree’ or ‘strongly disagree’ to each statement. The data reported are calculated by subtracting the sum of the ‘disagree’ and ‘strongly disagree’ percentage scores from the sum of the ‘strongly agree’ and ‘agree’ percentage scores. For the panel entitled ‘how good is the union at’ a category ‘intermediate’ was included in the question alongside options of ‘very good, ‘good’, ‘weak’ and ‘very weak’. The data reported for this panel were calculated by subtracting the sum of the ‘weak’ and ‘very weak’ percentage scores from the sum of the ‘very good’ and ‘good’ percentage scores. The ‘intermediate’ scores were discounted in the second section of the table.
In every union, members with a workplace representative were more likely to view the union as ‘well organized’ than members with no representative and in 10 of the 14 unions the score with a representative present was positive whereas that with no representative present was negative. While there may be an impact linked to the non-unionization of some works councillors in dual systems, it is insufficient to change the direction of the relationship between the presence of a works councillor and the perceptions of members of union organizations. The differences in the percentage scores on ‘well organized’ between members with and without workplace representation average 69.1 percentage points, which are high compared to the differences for the other statements. Only in UILCA (9.8 percentage points), BBDSZ (24.4 percentage points) and KASZ (35.8 percentage points) were the differences in the scores between members with and without workplace representation less than 50 percentage points. From the perspective of members, union organization is thus markedly enhanced through the presence of a workplace representative.
For ‘my workplace is well managed’ and ‘management-employee relations are reasonable where I work’ in 10 of the 14 unions members were more positive when there was a representative present at the workplace. In five unions, four of which are Nordic (Sif, TU, NITO, HK), members with a workplace representative were more likely to cite that ‘management has prevented the union from influencing key decisions’ than were members without a workplace representative. The average difference on ‘my workplace is well managed’ is 16.4 percentage points; on ‘management-employee relations are reasonable where I work’, 15.6 percentage points; and on ‘management has prevented the union from influencing key decisions’, 27.6 percentage points. 4 These narrower differences compared to ‘well organized’ are anticipated in so far as each of these statements refers to relations between management and workplace representatives, which are not determined solely by the actions of workplace representatives. In the majority of participating unions the presence of a workplace representative was sufficient to influence unionists to take more positive views of the performance of the union.
In the 11 participating unions based in western Europe, members with representation were less likely to think that ‘within my workplace my union has little power and influence’ than unionists with no representative present. In contrast, more members with a workplace representative from the three unions based in eastern Europe (BBDSZ, KASZ, Solidarność) thought that ‘within my workplace my union has little power and influence’ compared to their counterparts with no union representative. While this result suggests that union influence at the workplace is more constrained in eastern Europe than elsewhere, Hungarian and Polish unionists also thought the presence of a workplace representative limited the capacity of ‘management to prevent the union from influencing key decisions’ and that the union was more likely to be ‘well organized at the workplace’, suggesting that the constraints on union influence are mitigated.
The lower panel of Table 3 presents the results on the perceptions of members on tasks undertaken by the union at the workplace. Table 3 includes two tasks on communications, two on representation, two on internal union relations and finally one on recruitment and retention. Treating the lower panel as a matrix of unions and tasks, members with a representative rated the performance of the union to be superior on 83 of the 98 cases than members without a representative. In eight of the unions operating in single channel systems (RBF, NITO, HK, BBDSZ, KASZ, Solidarność, UILCA, UNITE) members regard the performance of the union to be superior on every one of these tasks when there was a shop steward present at their workplace. Although differences between members with and without a workplace representative were small, in Sif and TU members without a representative rated the performance of the union to be superior in ‘encouraging participation in the democratic processes of the union’ (hereafter participation). Additionally, the same group of members in TU viewed the union to be superior at ‘consulting members on matters that affect them at work’ (hereafter consulting members), although both members with and without workplace representatives scored the union negatively on this issue. For the unions operating in dual systems, members without a works councillor rated the performance of the union on at least one of the tasks to be superior compared to their counterparts with a works councillor and in CC.OO this situation prevailed for four of the seven tasks. The situation in CC.OO may result from the peculiarities in the division of labour and the constitutional relationship between works councillors and shop stewards.
Regarding the communication tasks, only in TU and CC.OO did members without a representative rate the performance of the union on ‘consulting members’ superior to members with a workplace representative and in CC.OO the same situation prevailed on ‘regular communication with members’ (hereafter regular communication). Elsewhere members with a workplace representative rated the performance of the union to be superior on both tasks. Furthermore, the scores recorded by members without a workplace representative are negative in five unions (TU, NITO, HK, UNITE, Bondgenoten) on ‘consulting members’, and in eight unions (Sif, TU, NITO, HK, UNITE, Ver.di, GPA, Bondgenoten) on ‘regular communication’, indicating that more members thought the performance of the union to be weak than viewed the performance as good. The average difference between the scores for members with and without a workplace representative on ‘consulting members’ is 25.7 percentage points and on ‘regular communication’ is 26.4 percentage points. 5 In other words, the presence of a workplace representative has a marked and positive impact on members’ perception of the union. This point applies to the unions that operate in dual systems even though more than 20 per cent of works councillors in the sample are not unionized.
The representative tasks included in Table 3 are ‘individual representation on grievances, discipline and other matters’ (hereafter individual representation) and ‘ensuring health and safety’ (hereafter H&S). With the single exception of CC.OO, where the scores are equal, members with a workplace representative rate the performance of unions to be superior on ‘individual representation’ compared to members with no representative. Furthermore, there are no negative scores recorded, indicating that even when there is no representative present at the workplace unions have implemented systems to ensure members have access to representation. In many unions full-time officers are deployed to meet the shortfall in workplace representatives. On ‘H&S’ it is only in Bondgenoten that members without a representative rate the performance of the union to be superior than members with a representative. Elsewhere representatives have a positive effect on members’ perceptions of ‘H&S’. The differences in the average scores for members with and without a representative at their workplace were narrower for both ‘individual representation’ (20.9 percentage points) and ‘H&S’ (19.7 percentage points) compared to the communication tasks.
The presence of a representative also affects members’ perceptions of how unions articulate workplace activities with the wider union represented in Table 3 by ‘encouraging participation in the democratic processes of union’ (hereafter encouraging participation) and ‘representing your interests within the union’ (hereafter representing interests). In all of the participating unions, members with a representative were more positive than members with no representative on ‘representing interests’, a situation also present in 10 of the 14 participating unions on ‘encouraging participation’. It should be noted, however, that in Sif, TU and HK and GPA negative scores on ‘encouraging participation’ are recorded for members with and without a representative whereas in NITO, KASZ and UNITE only members without a representative record a negative score, a situation reproduced in TU and UNITE for ‘representing interests’. The large number of unions recording negative scores suggest that the impact of workplace representatives is not as marked on articulating workplace activities as it is on communication and representation, a point supported by the lower average scores with and without a representative for ‘encouraging participation’ (17.2 percentage points) and ‘representing interests’ (17.8 percentage points) than the comparable scores recorded on communications and representation tasks. In so far as the organizational power of trade unions requires internal union democracy (Lévesque and Murray, 2010), however, the presence of a workplace representative contributes to the maintenance or generation of organizational power.
Comparing union performance with that of management
Direct communication between management and employees is a key theme of much recent literature on HRM (Boxall et al., 2007). Such communications can be formal or informal, may be by email, work-of-mouth or one-to-one meetings, may be effective in generating higher levels of employee commitment to the goals of the enterprise, and may lead employees to regard management as the means whereby problems at the workplace can be addressed and as the primary source of workplace information (Dundon et al., 2005; Wilkinson et al., 2013). In some circumstances direct communication from managers is associated with union substitution and the bypassing of workplace representatives (Bratton, 2001; Heery et al., 2004). In the context of sophisticated managerial practices intended to generate employee commitment, a challenge for representatives is to ensure effective communications with members, thereby ensuring that the union message and a union purpose is sustained at the workplace (Gumbrell-McCormick and Hyman, 2013: 152–153) and mitigating, if not avoiding, a ‘crisis of workers’ loyalty to their unions’ (Müller-Jentsch, 1988: 177–178). Stage three examines the efficacy of management and unions on communications and related workplace issues. The statements listed in Table 4 refer to issues where unions and management provide information and may contest the interpretation of events. Communication on these issues is thus central to securing influence.
The relative performance of the union and management.
Note. For the questions reported in Table 4 respondents were asked to answer ‘very good’, ‘good’, ‘intermediate’, ‘weak’ and ‘very weak’ to each of the statements. The reported data are calculated by subtracting the sum of the ‘weak’ and ‘very weak’ percentage scores from the sum of the ‘very good’ and ‘good’ percentage scores.
Four points are immediately apparent from Table 4. First, and replicating the evidence presented in Table 2, for 96 of the 98 statements union members thought that the union was more effective when a workplace representative was present. Only on ‘ensuring security of employment’ in Sif, and ‘dealing with work problems you or others may have’ in Bondgenoten did members without a workplace representative think that the union performed better. Secondly, in 86 of the 98 cases the union performance was superior to that of management when a workplace representative was present. This number fell to 22 when no workplace representative was present. Trade unionists are thus more likely to regard the union as superior to management at the tasks listed in Table 4 when a workplace representative is present. Thirdly, in 59 of the 98 cases the performance of the union in the presence of a workplace representative was superior to that of management whose performance, in turn, was superior to that of the union where no workplace representative was present. In other words, the absence of a workplace representative allowed management to gain a superior position in almost 60 per cent of cases. Fourthly, for only one case, ‘dealing with problems you or others may have’ in Bondgenoten, was the performance of the union in the absence of a workplace representative superior to that of the union in the presence of a representative and that of management. In the majority of cases the presence of a representative is central to a positive perception of the performance of the union among members and to the perception of a superior performance of the union compared to management.
Examining the data by reference to the statements further demonstrates the impact on the perception of union performance among members of the presence of a workplace representative. Regarding the statements in Table 4: in every union the performance of the union when a representative was present in the workplace of the member was superior to that of management on ‘treating members/employees fairly’. Within the single exception of NITO the same situation applies to ‘dealing with work problems you or other may have’. Management outperformed the union where a workplace representative was present in only one union (Sif) on ‘providing opportunities to comment on proposed changes’, and in two unions (NITO and BBDSZ) on ‘solving disputes that flare up between management and employees’.
Similarly in Sif and NITO management was rated more highly than the union when a workplace representative was present for both ‘keeping everyone up to date on proposed changes’ and ‘responding to suggestions from members/employees’. Putting aside the statement ‘ensuring security of employment’ in only 10 of 84 cases recorded for the remaining six statements does management outperform the union when there is a workplace representative present from the perspective of union members. Seven of these 10 cases are concentrated in two unions: Sif and NITO. Irrespective of whether a single channel or a dual system is in operation, the union with a workplace representative present is generally superior in performance to management in the other 12 unions.
‘Ensuring security of employment’ is the only issue in Table 4 where trade unionists think that management outperforms the union when a workplace representative is present in four unions (Sif, RBF, NITO and HK). It is thus only in Nordic unions that workplace representation may be inferior to management on the issues included in Table 4, and even among Nordic unions management is superior only in a minority of cases.
In contrast, management outperformed the union when there was no workplace representative present on every issue listed in Table 4 for seven unions (Sif, RBF, TU, NITO, HK, UNITE and GPA). In only two unions (UILCA and CC.OO) did the union outperform management when there was no representative. Elsewhere management outperformed the union on six of the seven issues listed in Table 4 in BBDSZ, five issues in Ver.di and Bondgenoten, three issues in KASZ and two issues in Solidarność. In short, management is much more likely to outperform the union when the member is not supported by a representative at his/her workplace.
Conclusions
This study has demonstrated four principal points. First, the performance of the union is superior at workplaces with representatives compared to workplaces where there are no representatives. Secondly, the union is more likely to outperform management on a range of workplace tasks when a workplace representative is present, but is unlikely to do so in the absence of workplace representation. Thirdly, a substantial minority of members with a representative at their workplace think that there are either not enough representatives or that representatives do not adequately represent members. When these data are combined with those on the number of members who do not have a workplace representative at their place of work, a majority of members in eight of the 14 participating unions were not content with union representation, with women members more likely to be discontented. Fourthly, the three points mentioned above apply to shop stewards in single channel systems and works councillors in dual systems of representation. Although more than 20 per cent of works councillors were not unionized, there were no marked differences between the two systems of workplace representation on the core issues of this analysis.
While there is no doubt that many managements have implemented more sophisticated workplace and communication practices in recent years, these have not generated a ‘crisis of workers’ loyalty to unions’ at workplaces where representatives are present. To the contrary, members view the performance of the union to be superior to that of management in these circumstances. The opportunities for management to generate a ‘crisis of workers’ loyalty to unions’ are more evident at workplaces with no representative where members are more likely to view the performance of management to be superior to that of the union. It is also the case that the presence of a workplace representative averts aspects of the ‘crisis of interest aggregation’ as once unions have established a workplace presence in private sector services the impact of a workplace representative is similar to that in sectors of more longstanding union organization.
The implications of these findings for union renewal are wide-ranging. Without maintaining, or preferably increasing, the number of workplace representatives, union renewal is unlikely. Furthermore, a reason cited by many members that leave trade unions is the inadequacy of support and representation at the workplace. In the context of deregulation and the decentralization of bargaining it is also difficult to envisage union members either formulating and delivering a bargaining agenda or mobilizing in support of such an agenda in the absence of workplace representation. A host of studies demonstrate that support from the union at the workplace is the principal reason for joining a union (Freeman and Rogers, 1999: 40–43; Jørgensen et al., 1992; Waddington and Whitston, 1997). Similarly, non-members often report an unwillingness to join a union if they perceive the union is ineffective. In short, the workplace representative is central to union joining and membership retention. The impact of representation on the generation of ‘social capital’ at the workplace is also evident. The union is perceived by members as better in communicating, articulating between workplace and the wider union, and the representation of members in the presence of a workplace representative, all of which contribute to the generation of ‘social capital’.
Footnotes
Funding
This research was funded by the European Trade Union Institute. The European Trade Union Institute receives funding from the European Commission.