Abstract
Are International Framework Agreements potential mechanisms of cross-border labour regulation? For some scholars they are no more than ‘window-dressing’, for others they are effective instruments of transnational regulation. We argue that their effectiveness depends on their interconnection with other mechanisms of labour regulation. Drawing on a broad review of research and our own empirical data, we conclude that different institutional frameworks, different dominant actor groups and configurations, and also different modes of regulation and bargaining, lead to varying and sometimes contrasting outcomes.
Introduction
International Framework Agreements (IFAs) have become increasingly widespread since the mid-1990s. Sometimes they are used by Global Union Federations (GUFs) to start social dialogue with multinational companies (MNCs); sometimes they are negotiated between management and labour at the company level, with GUFs only later invited as signatories. IFAs are global in scope and typically relate to ILO core conventions (Schömann, 2012). In the past decade there has also been a growth in European Framework Agreements (EFAs), narrower in geographical scope but usually broader in content (Telljohann et al., 2009a, 2009b). Transnational Company Agreements (TCAs) are either IFAs or EFAs in which European Works Councils (EWCs) play a key negotiating role. We refer to all these types generically as Transnational Framework Agreements (TFAs). The European Commission database (2013) lists 215 TFAs with varying content and scope of application. Our own research identified 89 IFAs signed between management and GUFs in 87 MNCs by the end of 2012. A further 14 agreements on international labour standards were signed between MNCs and EWCs, without involving GUFs. For some researchers, IFAs are no more than ‘window-dressing’, but for others they are effective instruments of transnational labour regulation, extending minimum labour standards into regions where labour would otherwise be too weak to represent itself.
We develop two arguments in this article. First, to analyse the effects of transnational agreements we must take systematic account of their different forms. To this end, we highlight the specific role of EWCs and the differing union strategies towards the content and negotiating actors in TFAs. We also refer to a specific group of TCAs that regulate core labour standards.
Our second innovative contribution is to explore the interconnections between cross-border agreements and other mechanisms and actor groups in labour regulation. We argue that cross-border agreements are not per se either effective or ineffective: the effect depends not only on the type of TFA, but also on the specific interaction with other regulatory mechanisms. From a broad review of research and our own empirical data we conclude that different institutional frameworks, different dominant actor groups and configurations and different modes of regulation and bargaining lead to varying and sometimes contrasting outcomes. Evaluation of TFAs also depends on the conceptual framework adopted: traditional industrial relations approaches concentrate on the collective actors of capital, labour and the state (Hyman, 1994; Poole, 1986), but the broader context of societal power relations and the role of other actors are frequently underestimated.
Our analytical focus is on the emerging web of transnational labour regulation. This approach integrates and extends ideas of classic industrial relation and organizational bargaining approaches (Walton and McKersie, 1991) with neo-institutionalist concepts of institutionalized expectations of behaviour. The dynamics and effects of specific types of TFAs depend not only on varying union strategies, but also on the strength of other collective actors, including civil society actors. Each actor group follows its specific logics; but all are connected within a broader organizational field that spans different ambits of regulation (local, regional, national, supranational, transnational, global), arenas of regulation (individual employment contract, collective agreement, labour law, discursive legitimation) and dominant actors (individuals, companies, unions, state, NGOs and social movements).
We outline a multi-actor, multi-dimensional and multi-local framework for analysing TFAs. Our theoretical framework derives from previous studies of IFAs and also our own research, which includes a standardized analysis of IFA texts. We identify actor groups formally involved in the negotiation of the IFAs and their main contents; then analyse other TFAs on core labour standards. In order to capture the complex nature of such cross-border agreements and the negotiation process, we present case studies of three IFAs in companies with different institutional, organizational and actor settings. Finally, we propose a general framework of transnational labour regulation.
IFA research
In industrial relations research, IFAs are defined as potential mechanism for unions and workers’ representatives to implement minimum standards for working conditions in MNCs and enable GUFs to become an accepted negotiation partner fighting for the right to organize at all company sites (Papadakis et al., 2008; Schömann, 2012). IFAs may be viewed as one form of transnational texts and agreements between MNCs and GUFs; other possible outcomes include EFAs. These studies usually focus on strategic decisions by management and labour and the corresponding power relationships (Müller et al., 2008; Schömann et al., 2008b; Telljohann et al., 2009a).
Another perspective comes from management and business research, especially from international HRM and business ethics, with a focus on management decision-making processes (Edwards et al., 2007; Egels-Zanden, 2009; Ungericht and Hirt, 2010; Sotorrio and Sanchez, 2008), institutional expectations (Fichter et al., 2011a; Preuss et al., 2009) or new models of labour relations along global value chains (Riisgaard and Hammer, 2011). Other authors discuss the legal status and enforceability of IFAs (Coleman, 2010; Ewing, 2011; Sobczak, 2007; Zimmer, 2008).
In an institutional approach, representation bodies are embedded in an institutionalized frame of societal expectations. These may encompass employee representation bodies accepted by management as legitimate dialogue partners at global level (Riisgaard, 2005; Stevis, 2010: 5) but may also include new social movements, NGOs and other collective actor groups (Tarrow, 1998).
These three perspectives need to be integrated in a broader conceptual framework. Our approach attempts to connect different theoretical perspectives to explain actor constellations, content and outcomes of IFA negotiations. There may be differences between formally outlined contents of TCAs and actual negotiation processes (Hertwig et al., 2011; Rüb et al., 2011). In considering IFAs in a broader framework of transnational labour regulation, not only GUFs and top management but more complex constellations of actor groups must be considered. This involves different levels (global and regional) and different actor strategies. These principles guide our empirical analysis of a broad number of existing IFAs. We focus on both dimensions, analysing the content of IFAs in 62 MNCs and three company case studies.
Research design
Our research analyses the various dimensions of labour regulation through content analysis of IFA texts and other TCAs on core labour standards. We selected 62 companies that signed an IFA or another TCA on core labour standards before the end of 2011 and have a EWC, so as to highlight the role of other employee representation bodies besides GUFs. A detailed content analysis of the IFA (norms, actors involved, suppliers mentioned, enforcement/monitoring mechanisms, etc.) was combined with company-related information (sector, size, headquarters). In addition, 11 interviews were conducted in spring 2010 with headquarters officials from six GUFs: UNI Global Union (UNI), the International Metalworkers’ Federation (IMF), the International Federation of Chemical, Energy, Mine and General Workers’ Unions (ICEM), Building and Woodworkers International (BWI), International Textile, Garment and Leather Workers Federation (ITGLWF) and the International Union of Food, Agricultural, Hotel, Restaurant, Catering, Tobacco and Allied Workers’ Association (IUF). The IMF, ICEM and ITGLWF have since merged to form IndustriALL.
In order to explore in detail the negotiation processes, outcomes, actor involvement and strategies as well as the overall company contexts, we analysed two company case studies in the metal sector and one in the service sector. Semi-structured interviews were conducted with union officials (trade union coordinators of all three companies), EWC members (secretary and other members of the steering committee) and HR management (with head of international HRM in the two metal sector companies).
GUFs and EWCs in 62 selected MNCs
By the end of 2012, around 100 MNCs had signed agreements on core labour standards with GUFs and other employee representation bodies (Global Unions, 2013). Some 100 TFAs on other topics were signed between managements and employee representation bodies. The actual number of IFAs is not easy to verify because they do not have to be officially registered; hence most researchers rely on data provided by the GUFs. However, this often underestimates the role of EWCs, which formally possess only information and consultation rights but in practice, often manage to negotiate agreements with management at European level; 29 have signed 56 EFAs and IFAs. As a result, they have enlarged their field of work and become partners in transnational negotiations with management that go beyond the regional area of the EU (Schömann et al., 2008a; Stevis, 2010). Therefore our analysis includes agreements with a transnational reach that are not negotiated by GUFs but by other employee representation bodies, ‘joint-texts’ (Commission of the European Communities, 2008) or TCAs (Rüb et al., 2011) but, on the other hand, is limited to companies that actually have an EWC.
Out of 92 companies signing such agreements between 1989 and the end of 2010, 62 have a EWC installed. In 78 of the 92 agreements, one or more GUFs were signatories or co-signatories; they prescribe a global application and can therefore be regarded as IFAs (Schömann et al., 2008a; Telljohann et al., 2009b), although some exclude certain countries such as China or the United States. Three were EFAs, signed by a European Trade Union Federation (ETUF) and applying only to Europe. In 11, the EWC was the only signatory on the employees’ side; seven of these specify a global scope while four are limited to the European operations of the companies. Here the question of actor constellation and especially of the role of EWCs in relation to ETUFs or GUFs becomes crucial. Evidently a clear distinction cannot easily be drawn between European and global issues and actors. We therefore focus on the 62 companies with EWCs, so as to compare the role played by EWCs in different company contexts. To compare differences in the contexts of the agreements and the strategies of management and employee representatives, the database was supplemented by transnational texts on labour standards which were signed by EWCs without the involvement of a GUF.
The ILO core conventions are the main point of reference for all agreements, but they do not always include them all. Some also include other issues such as health and safety. In the 62 companies, 55 agreements have a transnational scope and seven apply only to European sites. In contrast to national regulation, IFAs are not governed by national or transnational legislation, thus conflicts must be resolved, if at all, by company-level bodies and individual actors. In 15 of the agreements there is no mention of any form of monitoring or conflict regulation mechanism; 21 state that a joint monitoring group should be installed.
Our content analysis shows that at the formal level, EWCs are a crucial factor. Formally, EWCs have a European reach rather than a global one. But in 45 of the 62 IFAs, the EWC is a signatory or is responsible for the implementation or monitoring of the agreement. In six out of these a World Works Council (WWC) also exists. Only in 17 companies with a EWC was it not formally involved in the negotiation of the agreement. In 35 of the IFAs, both GUFs and EWCs were involved in its implementation and monitoring. Four actor constellations could be distinguished in the negotiation process: first, the EWC or WWC was not involved (17 cases); second, the EWC or WWC has certain functions regarding the agreement but did not sign (15); third, the EWC or WWC signed the agreement together with a GUF (19); fourth, the EWC or WWC signed the agreement without GUF involvement (11).
Controlling systematically for sector (and the corresponding GUF) and for company headquarters, some interesting statistical correlations could be detected. In the metal sector, joint signature by the GUF and EWC (and WWC, if it exists) predominates; out of 21 agreements, in only one case was the EWC not a signatory, and in only one case did it sign alone. This is in line with qualitative findings of case studies showing that EWCs tend to develop into negotiation bodies which may extend their activities across a European reach (e.g. Dehnen and Rampeltshammer, 2010; Hauser-Ditz et al., 2010). In the service sector, the pattern is reversed: out of 17 agreements, eight are signed without the EWC and six by the EWC alone.
These contrasts are at least partly explained by the different approaches of the GUFs in each sector, confirmed in our interviews with their officials. In the metal sector, the IMF uses EWCs as a strategic instrument to start negotiations, and accepts them as negotiation partners and co-signatories as long as they cooperate with it (Papadakis et al., 2008; Platzer et al., 2009). By contrast, UNI in the service sector regards IFAs as its own field of labour regulation and as a global framework which should be adopted and specified by the affiliates at national and local levels. Since EWCs, by definition, are not trade union bodies, UNI makes a clear distinction between their roles and functions and those of trade unions, and it does not involve them in the IFA negotiation process.
In practice, relations between GUFs and EWCs can differ from these official policies; and they may simply be disregarded by EWCs which negotiate agreements with company management. The strategies of GUFs and EWCs, and the corresponding power relations between them, influence inter-organizational bargaining both between GUFs and EWCs and between employee representatives and management, and can explain different forms of involvement in the negotiation of IFAs (Dehnen, 2013).
All 62 agreements have one common topic of regulation, related to strengthening social dialogue in the companies. This marks a clear difference between these agreements and voluntary codes of conduct. The companies in the sample had what could be called a Eurocentric structure, an aspect also found in other research on IFAs (Egels-Zanden, 2009; Hammer, 2005; Stevis, 2010). All but three have their headquarters in the EU, and 45 also focus their core business activities (defined as more than 40% of employees and sales) on Europe. There are 16 companies with headquarters in Germany; 12 of these IFAs were negotiated by the GUF/EWC-actor constellation. Only six of the 62 companies signed an agreement on international labour standards before 2001, when the European Commission began to encourage IFAs.
As our case studies show, the actor constellations which are stated in the agreements may not correspond to actual power relations and the negotiations between the bodies of employee representation in all case studies. Furthermore, both formal content and actual negotiation processes need to be analysed in order to understand the dynamics of transnational labour regulation.
Institutional pressures and modes of negotiation: Three case studies
As we have shown, the approach of each GUF towards EWC involvement is important in explaining bargaining tactics and outcomes. More generally, the topics of the agreements and the scope of application are influenced by the actors’ strategies. We explore the different dimensions that shape transnational forms of labour regulation through three company case studies. All three companies signed their IFAs after 2000 and have their headquarters in Europe. Historically, they are European in the sense that their core business activities were developed at European sites. However, they have become global players and have production facilities around the world, and changes in business structure led to new forms of labour regulation. All companies have a corporate structure which includes corporate human resource management. As part of our access agreements, we are obliged to refer to the companies anonymously. Details of each are summarized in Table 1.
The three company cases.
Company A
The company began to internationalize around 1998, with the emergence of new markets in South America, Africa and Asia; business units are organized along product divisions. The initiative to negotiate an IFA came from the IMF German coordinator, an IG Metall official; he is a member of the company’s supervisory board, giving him a powerful position ensuring frequent contact with the head of HRM. However, management decided to use the IFA as a tool to strengthen the company’s internal social dialogue with the EWC, so the latter gained more influence during the negotiation process. The IMF coordinator and EWC members had to find a common mode of negotiation. Their relationship is close, with each keeping the other informed about their work. The IMF coordinator convinced the EWC members of the importance of an IFA and took into account their views in negotiating with management.
Management agreed to negotiate an IFA at corporate level with the chair of the EWC and the IMF coordinator. The head of corporate HRM was the negotiation partner on the management side. The process was one of integrative bargaining (Walton and McKersie, 1991), since both negotiation partners regarded the IFA as a ‘positive-sum-game’ which would benefit both sides. Management agreed to negotiations for two reasons. First, the IMF coordinator introduced the idea to the board, and other board members decided to support an IFA. Second, the head of HRM saw an IFA as a management mechanism to transfer HR practices throughout the company. Because HR had an interest in using the IFA for its international strategies, the IFA was jointly implemented by management and employee representatives. The IFA has a global scope and relates to ILO core conventions as well as other aspects of employment relations, like health and safety, living wages and working time. Although studies of IFAs are generally rather pessimistic when it comes to actual implementation (Hammer, 2008), the company provides a good example of successful transnational conflict regulation and implementation. In 2008 conflict arose over the right to organize at a company site in Turkey; the unions there contacted IG Metall, which then informed the IMF coordinator and the EWC. The EWC chair discussed the issue with the head of HRM, and they went to Turkey together to resolve the conflict. Although the dispute arose again in 2011, the IFA could be used by employee representatives to start discussions with management.
Company B
This formerly European firm, with major subsidiaries in France, Belgium and Spain, was taken over by with an Indian company in 2007. The merger coincided with the opening of new production facilities in countries outside Europe, especially in Asia. An IFA was negotiated in 2005, before the merger; later two other agreements, an EFA and an IFA, were signed with the new management. The original IFA has not yet been renegotiated.
The IMF was involved in the original negotiations, but it was the EWC and the joint social dialogue group which led the initiative: the IFA was a product of a more general vision shared by the EWC chair and corporate management. It was one outcome of good relationships with the different national trade unions and company’s EWC. The company became a member of the UN Global Compact in 2003 and developed a voluntary code of ethics for its employees. Although the company had production facilities in Brazil, its operations were predominantly European; this put the EWC in a strong position and provided direct contact with headquarters. While formally the IFA had global application, its practical scope was primarily European.
Following the company merger, the IFA on core labour standards signed in 2005 has become ineffectual, since it is not accepted by the new top management. The 2008 IFA, on the other hand, was negotiated under different circumstances. The mode of integrative bargaining between management and trade union bodies was largely the same, but the aims and strategies of the different actors had changed. The agreement was narrower in scope, aimed specifically at monitoring and improving health and safety standards throughout the company; unlike other IFAs, it does not include ILO core conventions. Implementation and monitoring are performed by joint monitoring groups at all levels, from local to corporate. Dispute resolution mechanisms exist at local, national and transnational levels.
Negotiations were initiated for two reasons. First, when the company internationalized, both management and employee representatives regarded health and safety standards at the company’s production sites as inadequate, especially in Asia. Second, a health and safety committee already existed in the (European) company before the merger, so that this body could be charged with monitoring health and safety standards. The CEO, for whom health and safety at the company is a priority, regarded negotiations with trade unions as important to ensure implementation of the standards.
Power relations between EWC and IMF shifted after the 2007 merger: the new ownership structure and new internationalized HRM strategies strengthened the position of the global body. The subsequent EFA and IFA were negotiated and signed by trade union bodies only, to the exclusion of the EWC. Nevertheless, the IMF and its affiliates cooperate with the EWC, including on the topic of IFAs. Bargaining between the EWC and IMF reflects shared attitudes which define different roles for each actor: the EWC has information and consultation rights while the trade union has the right to negotiate.
Company C
This company is a major player in the retail industry, with its main markets in south-west Europe, Asia and South America. It is a formerly European-centred company with a strong presence in its home country, France. In 2005, two private equity funds bought key share holdings, and the change of ownership brought a new management structure and affected the company’s approach to social dialogue.
In contrast to the two previous cases, the EWC was not involved in the negotiation of the IFA in 2001, nor in its implementation and monitoring. The agreement was negotiated by UNI, which was able to gain the assent of management because of the strong social dialogue structures. Given French dominance of the corporate structures at the time, the French trade unions – which had strong ties with management – were closely involved in the whole process and agreed a common position with UNI. There was a complex dynamic of inter-organizational structures and dialogue between UNI, national unions and the EWC. Key positions in the latter are held by members of UNI affiliates, and the EWC was used as a channel to involve trade unions from other European countries and as a platform for trade union networks. However, it was excluded from the negotiations with management.
The IFA provides that the implementation and monitoring is the responsibility of UNI, without any role for company-level bodies or bipartite regulation by management and trade union organizations. Although UNI tries to use the IFA to strengthen workers’ rights, corporate management refuses to implement it at national and local level where social dialogue structures are not well established. Although the IFA specifies that employees have a right to organize, UNI still faces a problem with ‘yellow unions’ at Spanish sites. Royle and Ortiz (2009) conclude that – at least for the Spanish subsidiaries – the IFA has minor, indirect effects, and has not led to implementation of the right to organize. The distinctive characteristics of the Spanish commerce sector undermine transnational forms of labour regulation.
When the international private equity funds secured key holdings in 2005, the formerly integrative modes of bargaining and ‘social partnership’ between management and unions at the headquarters turned into a more adversarial relationship. This shift is evident in a change in trade union strategy from negotiation with management to public campaigning against the company. Furthermore, attempts by UNI in 2010 to renegotiate the IFA resulted in a unilateral company declaration of a code of conduct, issued without prior notice or negotiation with employee representation bodies. UNI’s organizational capital is the dominant power resource for labour regulation in this case, given that UNI has a strong union network and uses the EWC as a mechanism for discussion with management.
Case comparison
Although the IFAs of the two metal sector companies are similar in their content, they differ in terms of the dominant actors in the negotiation processes and conflict regulation procedures. Both reflect models of transnational labour regulation in an actor setting of collective bargaining, in which negotiations take place between employee representatives and management in a manner aimed at cooperation and social partnership. Negotiations in the third case took place without the involvement of the EWC, although management and the EWC also acted according to a common ideology of social partnership. All three IFAs relate explicitly to the ILO minimum standards. Beyond this, their contents differ in terms of implementation and monitoring mechanisms. None of the agreements provides for joint monitoring machinery, except for the 2008 agreement at Company B, which has the narrower remit of international health and safety requirements.
Different factors could be detected which were crucial for the negotiation processes. Especially important were organizational characteristics of the companies, including institutional embeddedness in national industrial relations systems and changes in business structures. These factors help to understand the negotiation strategies and outcomes on the management side. Conflict resolution and modes of negotiation between different bodies of employee representation, as well as between ‘capital and labour’, framed the negotiations. We discuss these factors further in a more general framework for analysing transnational labour regulation in the concluding section.
What is also clear from the cases is that the bargaining relationships which facilitated initial negotiation on the IFAs may be disrupted by changes in business structures and consequential changes in HRM strategies. This was particularly apparent in Companies B and C; whereas greater ownership stability in Company A may help explain the greater effectiveness of the IFA.
Discussion and conclusion: A model of transnational labour regulation
We now integrate our own research results with empirical studies and theoretical considerations by other authors in order to explain the negotiations, contents and effects of IFAs. In order to analyse the negotiations and outcomes of IFAs and of TFAs in general, a theoretical framework is needed which combines the strategies of different actors with organizational and institutional factors. As we have indicated, an agreement which appears to be an IFA could be the result of intensive negotiations by an EWC. An agreement which appears to be a (regionally limited) EFA could actually be extended to other regions of the world by additional negotiations or even further formal provisions that enable EWCs to intervene outside the formal legal terrain of the EU. Any kind of TFA also could be primarily the outcome of management initiatives and strategies to harmonize and standardize labour regulation around the world. Case- and event-oriented NGOs and international media activities could, in conjunction with the long-term activities of GUFs, open new windows of opportunity as was the case after the collapse of a factory complex in Bangladesh in April 2013 (Greenhouse and Yardley, 2013; Solidarity Center, 2013). All the examples discussed above underline that TFAs are just one element in the broader emerging and strengthening texture of transnational labour regulation. IFAs exist besides other TFAs, besides codes of conducts and Global or European bodies of employee representation like EWCs. But, as was shown in the empirical sections, EWCs are not independent from the logics of TFAs. Thus the theoretical model presented in this article aims to combine different dimensions of labour regulation to explain negotiations over TFAs, and especially IFAs, and integrate these negotiations into a wider frame of transnational forms of labour regulation (Pries, 2010; Rüb et al., 2011; Schömann et al., 2012).
The first aspect of labour regulation involving IFAs is the subject or content of regulation. Out of the broad scope of possible topics of employment relations (pay, working time, external mobility), working conditions (qualifications, internal mobility, job control), and employee participation (individual or collective participation, codetermination), only a few elements are normally addressed in IFAs: they mainly relate to minimum conditions and norms. As the empirical analysis showed, some IFAs relate to other topics besides or instead of ILO core conventions. As Telljohann et al. (2009b) and Schömann et al. (2008a) have shown in their analyses, certain aspects depend on the differing strategies of the various GUFs. Rüb et al. (2011) have analysed the strategies of EWCs in respect of TCAs on European topics; but as we have shown, EWCs may have their own strategies in matters towards IFAs as well.
Thus, as a second dimension, different arenas and actor constellations affecting labour regulation by IFAs must be taken into account. Arenas of regulation are understood as the specific settings of normative frameworks in which certain actor constellations are socially established and taken for granted. They may range from direct legal contractual relations to the arena of legitimacy to which a company as a profit-making organization is exposed, and this may be almost wholly independent from strict formal-legal aspects. Since IFAs are negotiated between management and employee representation bodies, they are located in between the arenas of company agreements, collective agreements and discursive legitimacy; to identify the specific arena in which they are embedded is an empirical question, as was shown by the various forms of EWC involvement we have discussed.
A third dimension relates to the spatial-regional scope covered by labour regulation mechanisms. Labour law normally applies to nation-states but – as in the case of the EU – it may be valid at supranational level. Collective agreements could apply to whole economies, sectors, regions or companies. ILO minimum standards aim at a global level of regulation. IFAs are agreements with a transnational scope of application because they should apply to all parts of a company, anywhere in the world. But in practice they may formally exclude selected parts of companies (as in China, where cooperation with unions is difficult) and the actual scope of application could even be further reduced in practice (Stevis, 2010). Furthermore, depending on GUF policies, certain parts of the companies or their global value chains might be part of the IFA negotiations. Besides this, the cases presented above highlight the importance of including companies’ organizational characteristics, like business structures or the scope of internationalization, into the analysis (Hauser-Ditz et al., 2010).
A fourth dimension of labour regulation concerns the mode of regulation. Here the classical distinctions between substantive and procedural regulation and between formal and informal regulation are crucial. Substantive regulation refers to the direct definition of such elements as conditions of work and employment (for example, weekly working hours and wages), whereas procedural regulation is designed to define the mechanisms and actors involved in determining substantive issues. Formal modes of regulation are explicit statements, contracts and rules; informal modes of regulation include ‘customs and practices’ that do not necessarily have a formal framing. As IFAs are written agreements, they are formal regulations, but their informal significance and impact can vary to an extreme degree between companies, and even in one company between different sites as was shown in the case of Company C. IFAs normally have a mixed logic of substantive regulation (referring to the ILO minimum standards) and procedural provisions (which could be related to mechanisms of enforcement or monitoring).
The way in which individual and collective conflicts are regulated is a fifth dimension of labour regulation. Conflict resolution takes place when one actor tries to establish or change norms against the position of another actor. The ways of dealing with such conflict differs widely from country to country and from company to company. This is a critical point in negotiating IFAs, in that actors from different countries (and thus different institutional backgrounds) have to find a common solution. Conflict resolution during IFA negotiations takes place both intra-organizationally within employee representation bodies (mainly from the different plants and countries, but also between different political-ideological lines); and inter-organizationally between employee representatives and management, and between the different employee representation bodies (like GUFs and EWCs) (Dehnen, 2013). Conflict resolution and bargaining over labour regulation depend on the dominant actors involved, as was shown in the case studies. Regulation of the conditions of employment, work and participation is influenced by individual and by collective actors. Besides management and labour, other collective actors (like NGOs or public media) may also play an important role, even though they have no legal authority to become involved in labour regulation. In the case of IFAs, GUFs and companies are necessarily important actors, but so are other actor groups: EWCs, employee representation bodies at local or national level or suppliers (Fichter et al., 2011b; Pries, 2010; Stevis, 2010). Our own research highlights that EWCs may not play merely an ancillary role, but may be the driving force behind IFA negotiations.
A seventh dimension of labour regulation reflects the dominant power resources which can be activated by the different actors. These resources can be distinguished between social, cultural, economic, organizational and political capital. In negotiating IFAs, both economic and organizational capitals are crucial for the employee side. The actors may require economic capital to start campaigns if the company is not willing to negotiate an IFA. Internally, EWCs and/or GUFs require organizational capital to mobilize all national and local actors to implement an IFA. Probably more important, many studies indicate the crucial role of social capital in the sense of personal networks of trust. The configuration of actors is a further dimension which may help to explain different forms of (transnational) labour regulation. This dimension encompasses the taken-for-granted mutual expectations and roles of the main actor groups. Obviously, in a framing of liberal-individualistic or of unilateral-paternalistic actor configuration negotiating an IFA will be more difficult than in a configuration of collective bargaining or political concertation. As the case studies highlight, a change in these actor configuration can lead to new bargaining forms, as has been the case in Company B, or can lead to difficulties in re-negotiating an IFA, as has been the case in Company C.
Finally, an eighth dimension of labour regulation relates to the ideology of the actors, by which we mean common beliefs and values which are the basis of the institutional framework and the rules of labour regulation. To some degree, such a common ideology normally exists at national level, but not at transnational level. An international company may encompass both countries with a social partnership climate (like Germany) and those with antagonistic class cultures. Correspondingly, IFA negotiations may be dominated by a zero-sum logic or by a search for win-win outcomes. For IFA negotiations, our case studies showed that a social partnership climate helped in persuading company management to conclude an agreement. Case studies by other researchers show that different ideologies may cause difficult negotiations and lead to trade union campaigns against the management (Hoffman, 2011). Conflicting ideologies within one international company can also lead to difficulties in implementing IFAs (Hessler, 2012; Niforou, 2012).
These eight dimensions integrate elements of traditional industrial-relations research, actor-focused negotiation theory and institutional/legitimacy approaches. As was shown in the case studies and the quantitative analysis, all these dimensions are important for analysing IFAs. Any form of transnational agreement is embedded in specific company traditions and cultures, is linked (implicitly or explicitly) to existing local, national and transnational regulations and involves specific dominant actor groups and strategies as well as different modes of regulation and bargaining. Some authors discuss the problem of implementation and monitoring of IFAs as related only or predominantly to insufficient instruments to enforce compliance through legal norms (Hammer, 2008; Sobczak, 2007). However, the absence of legal enforcement mechanisms for implementation is not the only problem. It is often intra-organizational negotiation on the employee side – with quite opposing views on such instruments as IFAs – and inter-organizational negotiation between GUFs and other employee representation bodies (national unions, shop stewards or works councils) that make the negotiation and implementation of IFAs so difficult. Furthermore, if the agreements lead to improved conditions of labour, work and participation, the entire value chain of the company has to be integrated in the agreement (Hessler, 2012; Riisgaard and Hammer, 2011).
Considering IFAs in the broader framework presented here has important consequences. Not only GUFs and top management, but complex constellations of actor groups have to be considered. We need to take account not only of the global reach of such agreements, but also of other regional effects. These may result in different actor strategies, mainly on the labour side (GUFs and EWCs), which must be analysed.
The empirical findings presented in this article reveal the complex multi-dimensional multi-level framework in which the negotiation and implementation of IFAs are embedded. Different institutional frameworks, different dominant actor groups and configurations, as well as different modes of regulation and bargaining lead to contrasting outcomes. The actual meaning of IFAs as a regulatory mechanism has to be analysed in the context of other dimensions of transnational labour regulation. Almost all of the dimensions presented are crucial for explaining the negotiation, content, actors and outcomes of IFAs. The case studies showed that actor constellations, the meaning of IFAs and even the mode of regulation and actors’ ideology can change, for example following mergers and acquisitions.
Although research on IFAs usually highlights the actions of the GUFs, our evidence reveals the important role which EWCs and company-internal trade union networks can play. Hence national institutional settings and company-specific norms, actor constellations and conflict regulation modes can strongly influence both the negotiation and the implementation of IFAs. In some cases, IFAs may indeed be little more than ‘window-dressing’ for either GUFs or management; in other cases, they may be used effectively by either management or employee representation bodies. IFAs can, but need not, be in line with a company’s international HRM strategy; they also can be an outcome of unilateral pressure from GUFs or EWCs to strengthen their position within the company.
We have shown that it is helpful to start with an extended notion of dimensions of labour regulation, analysing in particular a broader set of dominant actors, arenas of regulation and power resources. In the case of IFAs, logics of collective bargaining are intertwined with logics of company agreements. The strength of the approach presented lies precisely in offering a framework for integrating the different threads of the emerging transnational web of labour regulation. As the case studies have shown, the dynamics and effects of each specific type of IFA depend on the different overall union strategies, but also on the strength of other collective actors, particularly EWCs. In order to analyse the interrelations between different levels and arenas of labour regulation, the separate ‘tunnel visions’ of actor arrangements have to be overcome. This argument also points to some of the limitations of our research agenda. The complexity of different levels or scopes of regulation (local, regional, national, supranational, transnational, global), arenas of regulation (individual contract, company agreement, multi-employer agreement, labour law, discursive legitimation) and dominant actors (individuals, companies, unions, the state, NGOs and social movements) makes empirical research particularly challenging: it is extremely difficult to control for some factors in order to measure the influence of others.
Whereas the approach to the emerging web of transnational labour regulation is dynamic and allows for capturing processes of bargaining, there is a lack of solid structures on which to rely: for example, the date of signing an IFA can readily be identified, but should it be taken as the outcome of a longer process of negotiations, as a by-product of a ‘package deal’ which was made at the same time, or is it just the beginning of processes of transnational interest regulation? Furthermore, the differentiation between representatives of trade union bodies such as GUFs, and other employee representation bodies such as EWCs, may be an artificial distinction. In reality, most EWC members are also union representatives and act as such. Pointing at an emerging web of transnational labour regulation should not be taken as an invitation to argue that ‘everything is related to everything else’, however. But most research related to transnational labour regulation deals with just one of the interwoven threads. Focusing on specific research topics and questions – in this case the relation between GUFs and EWCs in the negotiation and implementation of IFAs – is the best way to detect the actual entanglement of one thread with the others.
Footnotes
Acknowledgements
The authors appreciate the very helpful comments and proposals of the anonymous referees and the Editor; nevertheless, all responsibility remains ours.
Funding
This research received a grant from Hans-Böckler-Foundation/Germany.