Introduction
Economic globalization has challenged the roles and structures of traditional industrial relations actors. Firms are gaining considerable rule-making power in the global arena at the same time when national trade unions are weakening and collective bargaining is declining. The governments seem to allow for and accommodate the free operation of capital and product markets while demonstrating strong preferences for flexible labour markets. Since there is not much scope for ‘corrective measures’ at the national level, traditional actors have engaged in a number of labour initiatives with a worldwide scope accompanied by new actors (i.e. NGOs). Taken together, these labour initiatives seem to be evolving into a global labour governance regime (Hassel 2008) which involves different transnational industrial relations arrangements (Egels-Zandén 2009). International Framework Agreements (IFAs) are one of them.
IFAs are negotiated documents between MNCs and global union federations (GUFs) and stipulate compliance with core labour standards in company operations worldwide and their supply chain. Although literature on IFAs is proliferating, significant gaps remain mainly regarding local practice and compliance. The case studies of Wills (2002), Riisgaard (2005) and Stevis (2009) on the Accor, Chiquita and Daimler agreements respectively are so far the only attempts to fill this gap and have raised important questions on effectiveness. The paper addresses issues of local practice and enforcement of compliance employing a global governance conceptualization. The empirical focus is on the Telefonica IFA. We begin by reviewing the core concepts and critiques on global governance and further discussing the relevance of a global governance approach for industrial relations as a field of study. We proceed with our argument and methods, which are followed by the presentation and the discussion of our findings where we examine how multi-level dynamics ultimately impact on compliance. We conclude with key findings and their implications for global labour governance.

Global Governance
Definitions and critique
The discourse on global governance originates in political science and it has resulted in a vast literature from different fields including (albeit not limited to) international relations, international political economy, environmental studies and European Union studies. The review of the literature reveals overlapping areas, a number of commonalities and even convergence regarding the emergence of the discourse. Debates on globalization have generated and informed debates on global governance while the relationship between
1
them is mainly two-fold. Global governance has been conceptualized either as a response to globalization (Gereffi and Mayer 2004) or as a means to further and consolidate it (Overbeek 2005). However, these views are not necessarily mutually exclusive. In that respect, a further distinction can be made on whether global governance is perceived as ‘summative’ or ‘issue-specific’ (Whitman 2009). The former refers to the sum of all the different global governance arrangements and the latter to the global governance mechanisms of a specific area. Differences in perceptions have in turn led to a proliferation of definitions that have provoked severe criticisms in both conceptual and empirical terms: as Murphy (2000) put it, global governance seems to be ‘poorly understood and poorly done’.
Central to the debate is the departure from the notion of ‘government’. The developments of ‘governance without government’ at the global level (Rosenau 1992) have triggered definitional issues regarding the meaning of both ‘global’ and ‘governance’. The two terms are often used in a confusing manner: global as connoting either international or transnational, and governance as referring to regulation, participation or cooperation. With regard to the ‘governance’ element, regulation is perceived as only one function of governance (Gereffi and Mayer 2004) while the very notion of governance seems to imply that participation and cooperation are essentially institutionalized (Kahler and Lake 2003). With regard to the ‘global’ element, for Finkelstein (1995), global governance is about governing relationships and ‘government-like’ events that transcend national boundaries. In other words, global governance is in essence ‘doing internationally what governments do at home’ (ibid: 369). Yet, such a formulation does not depart from the convention that situates states and governments (or else the national/international divide) as the organizing focus of analysis (Rosenau 1999). Further narrow readings of the term ‘global’ have given rise to extreme claims that global governance is itself an aspiration. Referring to the absence of a global overarching authority, Whitman (2009: 138) reminds us that ‘there is no global governance of globalization as such’ and therefore global governance remains essentially ‘aspirational’. Ironically, the same author acknowledges that global governance comprises pre-existing arrangements that have been globalized to varying degrees as well as novel arrangements that have emerged in response to certain effects of globalization.

The lack of a common definition has prompted commentators to argue that global governance ‘appears to be virtually anything’ (Finkelstein 1995: 368). In this sense, it is perceived as being largely ‘atheoretical’ with little analytical and explanatory potential. Critics emphasize the need for parsimony and a clear focus: not all the forms or combinations of steering mechanisms can be assigned the label of global governance. However, in the critique of global governance as atheoretical no compelling case has been made on why a single or narrower and more parsimonious definition would add conceptual clarity. Advocates argue that although global governance is not a theory in its own right, nevertheless each of the different usages of the term serves to illuminate distinct aspects and processes of global change (Hewson and Sinclair 1999; Dingwerth and Pattberg 2006). Global governance is hence viewed as a concept at the making, which has been employed in different ways throughout the literature: analytical, normative and critical (Pattberg 2006). All three perspectives are welcome and useful in 2
that, taken together, they provide a holistic understanding of global governance. The critical perspective views global governance as a hegemonic discourse whose purpose is to disguise the negative effects of globalization (Overbeek 2005).

The normative one is mostly informed by the work of practitioners who perceive global governance as a political project that should be concerned with problem solving in light of the advent of globalization (Commission on Global Governance 1995). The analytical perspective comes closest to a theory since it uses global governance in order to understand and describe observable phenomena (Rosenau 1997; 1999).
Critiques however are not limited to definitional issues. Global governance is accused of having a normative bias, which in turn assumes a strong ‘cooperative’ or ‘accommodating’ function and hence largely neglects issues of power (Overbeek 2005). We agree only to some extent. The ‘normative’ perspective of global governance is by nature associated with a very strong normative bias given that it merely discusses how global governance should be done and how its shortcomings should be accommodated. On the contrary, issues of power are inherent in the ‘analytical’ understanding of global governance since it accepts that governance can be exercised by a multiplicity of actors whose interests are not always easily accommodated (Rosenau 1997). In this sense, the very critique on the normative bias of global governance confirms the position discussed above that rich insights can only be generated by the employment of distinct perspectives and usages. Moreover, the issue of power in global governance is increasingly attracting a lot of academic attention with Barnett and Duvall (2005: 8) suggesting that by identifying the different dimensions of power we are able ‘to understand how global outcomes are produced and how actors are differentially enabled and constrained’. However, according to further criticisms, an actor-centered focus of global governance (at least of the analytical stream) would expose the concept to the ‘pitfalls of pluralism’ (Overbeek 2005). The ‘plurality’ of actors, interests and structures seems to be ‘essentially undetermined, unbiased, ‘neutral’’ (ibid: 40). Opponents therefore emphasize the absence of context from the global governance discourse. Yet, as it will be discussed further below, a closer look at the analytical perspective of global governance reveals that the behaviour of actors both shapes and is shaped by the environment within which they operate.

Relevance for industrial relations
Economic globalization has challenged the roles and structures of traditional industrial relations actors. Firms are gaining considerable rule-making power in the global arena at the same time when national trade unions are weakening and collective bargaining is declining. The governments seem to allow for and accommodate the free operation of capital and product markets while demonstrating strong preferences for flexible labour markets. Since there is not much scope for ‘corrective measures’ at the national level, traditional actors have engaged in a number of labour initiatives with a worldwide scope accompanied by new actors (i.e. NGOs). Taken together, these labour initiatives constitute an evolving global labour regime (Hassel 2008) that involves transnational industrial relations arrangements as for instance the IFAs (Egels-Zandén 2009). The regime is characterized by the absence of ‘government’ (i.e. a number of overlapping public and private arrangements), decentralized responsibilities (i.e. from the UN or 3
OECD to the MNCs) and a shift from hard law towards soft regulation based on voluntary standards and benchmarking (Hassel 2008) –developments that industrial relations academics arguably cannot ignore. There is therefore a need for empirical research and conceptual clarity on the changing role of traditional actors, the role of emerging actors and the instruments they develop since, for instance, what classifies as an industrial relations actor at the national level may not be an actor in transnational arrangements, i.e. states (Egels-Zandén 2009).
The discourse on governance in industrial relations as a field of study is not new. In light of the common European market and the developments under the Social Chapter of the EU Treaty, scholars acknowledged the necessity to reach beyond the nation-state level which has traditionally been at the core of industrial relations analysis. Academic attention in turn resulted in a growing body of literature on the governance of European industrial relations. Leonard et al (2007), for instance, adopt a governance approach in order to discuss rather recent European industrial relations developments. The authors find the notions of ‘new governance’ (Scott and Trubec 2002) particularly useful in describing cross-sector, sector and autonomous processes of European social dialogue. The notions comprise experimentation and flexibility leading to ‘softer’ norms, devolution of responsibilities to the lowest possible level, interaction between government and private actors and finally, coordination between the different levels (ibid). The need for coordination is also emphasized by Leisink and Hyman (2005) who build on the ‘uncertain interdependencies’ between the European, national and sub-national levels and examine the open method of coordination (OMC) as a ‘pioneering’ mechanism of the European employment strategy (EES). Moreover, Marginson and Sisson (2004) not only demonstrate the strong potential of a governance framework for organizing analysis in international comparative research, but they further highlight the complexity of the EU as a multi-level system where top-down processes are only one aspect: bottom-up dynamics are equally (if not more) important.
Sisson (2007) makes a compelling case on the contribution of the concept in ‘revitalizing’ industrial relations. In his invigorating discussion on the necessity to move from ‘regulation’ to ‘governance’ of the employment relationship, the author brings to the surface the linkages of the concept of governance with the institutional tradition of industrial relations. Industrial relations are hence about ‘the institutions involved in governing the employment relationship, the people and organisations that make and administer them, and the rule making process that are involved’ (ibid: 59, italics in the original)

The author emphasizes the comprehensiveness of the definition ofin reflecting that industrial relations as a field of study attaches equal attention to private as well as public arrangements, to binding as well as non-binding regulation, to both individual and collective dimensions of the employment relationship while encompassing the different levels of where rules are made: workplace, company, sector, national and international. Most importantly, he argues that a governance perspective helps us explain the ‘causes and consequences’ of institutions. In this sense, institutions are the ‘rules of the game’ shaping the behaviour of the different actors: not only ‘what actors can do, but also their perceptions and preferences and thus what they will want to do’ (Scharpf 2000: 5). In 4 other words, a governance perspective is useful in that it puts the employment relationship into context (Sisson 2007) departing both from purely pluralist notions as well as economist arguments on institutions as external influences of otherwise rational actors.
Borrowing largely from international political economy to make our case for the global governance of the employment relationship, we reach more or less the same conclusions. Global governance is (re)institutionalized by the distributive effects of globalization on existing political actors or by the emergence of new actors. Advocates emphasize the need to depart from problems and their solutions rooted in functionalist and efficiency-based explanations, and to focus on the role of agency and choice (Kahler and Lake 2003). Globalization changes the interests of actors who in turn seek to design, build and overturn governance institutions so as to reflect their changing preferences. Whether governance will shift to the regional or global level, or to the hands of public and/or private realms is highly contingent on actors’ perceptions on which level or sphere of governance is more conducive to the realization of their interests. Equally, the effects of globalization on governance are strongly conditioned by existing (domestic) institutions. Institutions, for instance, shape globalization’s effects on political accountability, which becomes more contested as governance moves upwards (ibid).

Argument
The paper is situated within the analytical strand of the global governance literature in order to develop new insights of the IFAs as emerging instruments for the global governance of labour.  We have adopted a rather summative view of the global governance of labour, which we can briefly define as encompassing the different mechanisms and initiatives aimed at governing the relationships associated with the conduct of labour worldwide. By focusing on IFAs as only one set of mechanisms aimed at governing the conduct of labour in multinational enterprises, their operations and supply chains, we follow an issue-specific approach that allows us to examine IFAs’ causes and consequences from above and below. However, IFAs are not studied in isolation since the arrangements of the global labour governance regime interrelate and interact to varying degrees.
The analytical strand largely builds on the writings of Rosenau (1997; 2004) who defines global governance as comprising systems of rule, steering mechanisms at all levels of human activity through which authority is exercised in order to enable the governed to preserve their coherence and move towards desired goals, and where the exercise of authority has transnational implications. From the above definition, Pattberg (2006) has depicted four constituting elements that are valuable not only in describing observable governance arrangements but, as we are going to demonstrate, also in analyzing and explaining their processes and outcomes. The systemic features of global governance consist of: involvement of multiple actors; emphasis on (private) authority; multilevel nature; emphasis on non-hierarchical modes of steering. To these four premises, we add the issue of compliance. An examination of how compliance is usually (or systematically) is vital for the sustainability of arrangements.

5

Multiple actors
The term global governance seems to be rather unproblematic when employed in order to explain the political and regulatory dynamics of international regimes or intergovernmental institutions. Nevertheless, it is a contested concept when used to capture how the actions of private (profit and non-profit) actors at the global level result in institutionalized arrangements that stipulate compliance in issue-specific areas (Falkner 2003). Yet, these private arrangements resemble the governing functions of states and intergovernmental institutions and they can be seen as the functional equivalents of public forms of national and international governance (Pattberg 2005). So far the literature has largely focused on either public forms of global governance or arrangements where governance is shared between the public and private realms. The analytical perspective of global governance discusses diffuse, multi-actor arrangements, attaches equal (if not more) attention to non-state and often antagonistic actors including social movements, NGOs and MNCs and emphasizes the increasing significance of private authority in emerging governance arrangements.

Private authority
Only recently has private authority started to attract academic attention and the reasons are two-fold. The term ‘authority’ has been strongly associated first, with the domestic sphere and second, with public judgment (Cutler et al 1999). Authority in the absence of government and beyond the nation state was inconceivable while, even if the discourse was about governance and not government, it was limited to public governance. In response to these narrow understandings of private authority, Hall and Biersteker (2002: 5) argue that the relationship between those who exercise and those who are subject to authority is indeed a public one ‘to the extent that claims and recognition of claims of authority involve an open, visible process’ albeit not necessarily connoting the involvement of a governmental institution. Moreover, authority is defined as ‘institutionalized expressions of power’ and -whether public or private, domestic or international or transnational- it therefore implies a sense of obligation towards the established habits, norms and rules (ibid: 4). In this sense, global governance is also concerned with the ‘regulatory mechanisms in a sphere of activity which function effectively even though they are not endowed with formal authority’ (Rosenau 1992: 5). IFAs are the joint governance output of private authority, namely MNCs and global union federations. Expressions of MNC power in establishing transnational governance arrangements are well documented in the work of Cutler et al (1999). Whether in the form of industry self-regulation, production alliances or business associations, inter-firm cooperation can translate into durable mechanisms for standard-setting and rule-making. In the case of GUFs, their authority is rooted at the national level but has taken transnational forms. GUFs have a formal internal governance system comprising a worldwide network of national affiliates (industry union federations), which is coordinated via regional structures (Croucher and Cotton 2009). Both MNCs and GUFs also cooperate with and participate in public international structures such as the ILO, UN Global Compact etc. but IFAs are so far the only institutionalized arrangement between the two actors. Taken together, MNCs and GUFs constitute a ‘sphere of authority’ (Rosenau 1997) where actors govern the relationships associated with the conduct of labour. Yet describing how the issue of private authority manifests in the case of IFAs is not sufficient. We further need to ask why/how (private) authority emerges and how it is exercised. And for that, we consider the power relations of the different actors. However, the distinction between the different actors involved in this sphere of authority is not two-fold (i.e. GUFs and MNCs) while there are instances where private authority interacts with public authority (i.e. incorporation of ILO Conventions into the texts of IFAs).

Multi-level nature
The analytical stream allows for the incorporation of the micro-level of global governance since the latter can emerge from the actions of individual actors (ibid). It therefore avoids the theoretical limitations  of a top-down approach and it departs from an exclusive focus on the agents and structures that are global in scope. Indeed, the absence of a global overarching agent has been balanced by the emergence of governances of different forms and at different levels. These governances emphasize process over institutions: governance is about the processes of integrating the different institutional levels in ways that promote the goals of the system overall (Peters and Pierre 2004). Global governance hence links the individual with the collective; and the local with the national, the international and the transnational (Whitman 2009). In other words, global governance is essentially multi-level (Marks and Hooghe 2004) where issues of power are central. Citing Whitman (2009: 66), ‘[global] governance arrangements … are both outcomes of political deliberation and contention. Indeed, contention extends to differences that arise between the many kinds and levels of governance’. In other words, in order to address the ‘[global governance] of, by and for whom’ question and hence take into account power relations (ibid: 70), we need to examine the interactions between the different levels. The range of relationships includes (i) devolution of responsibilities to lower levels (in the case of IFAs, for instance, from the CEO to subsidiary management and from GUFs to local union representatives), (ii) upward delegation of responsibilities (for example, local actors bringing issues to the attention of global actors), (iii) absence of any relationship (each level constitutes an independent and autonomous jurisdiction) and (iv) duplication of responsibilities and/or jurisdictions (duplication is usually unnecessary in which case it creates conflict).

Non-hierarchical modes of steering
The relationship between the different levels is evidently far from hierarchical. Marks and Hooghe’s (2004) writings depart from conceptualizations of multi-level governance as inevitably connoting governmental hierarchy relating to it. . Structure is largely absent from this system of governing (Peters and Pierre 2004), while the focus is on interdependencies rather than clearly delineated territories (Jessop 2004). Arrangements usually ‘overlap with existing jurisdictions in order to solve particular collective action problems’ (Marks and Hooghe 2004: 25). In the case of IFAs the above is apparent in that they make explicit references to other areas of public policy including ILO provisions, the UN Global Compact, OECD Guidelines and host-country legislation. However, there is potentially an overlap with private areas such as corporate governance, human resources and social responsibility policies. For instance, provisions on environmental matters are likely to be included in both IFAs and company CSR policies. Here we need to ask whether and how this overlap with different areas of governance translates into local practice and subsequently impacts on the IFA effectiveness. For instance, is this overlap used to reinforce and deliver the IFAs’ objectives or does it create conflict?  The absence of government-type hierarchy is subsequently associated with a high sense of autonomy between (and –we would add– within) the different groups of actors at the different levels (Rosenau 2004). The relaxation of regulatory frameworks and the overlap with different areas discussed above allow for more strategic and autonomous actor behavior (Peters and Pierre 2004). IFAs engage multiple actors in a complex process of interactive decision-making. Although their signatories usually consist of corporate management and collectivities that are transnational in scope (GUFs and sometimes EWCs or European industry federations), a significant number of these agreements are also signed by home-country unions. Apart from negotiations, the processes of implementation, monitoring and evaluation of IFAs stipulate a multi-level involvement of different actors: global, national unions and other forms of workplace representation, as well as managers with different positions within the company hierarchy. The non-hierarchical element of global governance relates to actor motivations: all the different actors seem to enjoy varying degrees of autonomy that potentially influence the institutionalization process of IFAs, their architecture and ultimately their outcome.

Compliance
The outcome of governing mechanisms is strongly associated with the notion of compliance whose significance has been emphasized in all the different global governance literatures. The outcome can be ‘intended’ or ‘unintended’ (Anderson 1994). By intended outcome we define the shared goals and policy objectives of governance. Compliance is hence about the capacity of global governance arrangements to deliver their intended outcomes. As discussed earlier, global governance is about governing relationships and hence rule-making is only a part of governance, or else what we call the governance output (i.e. IFAs). Compliance in turn is not merely about complying with the governance output but in essence with the authority itself. The ability of global unions to engage corporate management in a form of transnational social dialogue and to conclude agreements is arguably a significant achievement. Nevertheless, the authority of both GUFs and MNC headquarters should not be perceived as embedded in their role as signatories; ‘authority is relational; its existence can only be observed when it is both exercised and complied with’ (Rosenau 1999: 295). The question is therefore whether private authority in the case of IFAs is complied with and how. We can distinguish between internal and external sources of compliance. Internal sources of compliance relate to the global governance arrangement itself. First, compliance may be generated by those who govern (in our case, GUFs and headquarter management) because of their ability to ‘command and control’ (Rosenau 1997). Here the issue of autonomy comes again into play. This internal enforcement of compliance assumes that those who govern enjoy greater autonomy than those who are governed (Cutler et al 1999). However, such an explanation contradicts the underpinnings of the concept of multi-level governance discussed earlier. Actors in lower levels and in different areas

8
may be more autonomous. For example, we cannot assume that it is sufficient for headquarters to make the business-case for the adoption of IFAs in order to expect compliance by local management. If then control from the part of those who govern is weak, compliance may be generated via internal interdependencies (ibid; Rosenau 1999). Actors comply simply because they are ‘tied’ to the arrangement for practical reasons or because their needs converge with the needs of those who govern. In this sense, actors may comply with IFAs because doing otherwise would perhaps mean the termination of the arrangement and such an outcome could be undesirable. Yet, practical interdependence or the spontaneous convergence of needs are insufficient to explain why informal private governance arrangements are complied with. Despite the absence of legally binding effects, compliance may be generated by the governance output itself via its references to other areas or via its moral force. In the IFA case, for instance, the references to ILO Conventions may operate as a source of compliance even if they have not been ratified (Daugareilh 2008; Seifert 2008). External sources of compliance refer to jurisdictions that are not part of the governance arrangement such as auditing companies and tribunals. However, systematic compliance via external means might endanger the coherence and enduring architecture of arrangements.

The global union UNI has more than 3 million members in the telecommunications industry in more than 130 countries. In terms of governance, it is organized into three divisions: ‘UNI-Europa’, ‘UNI-Americas’ and ‘UNI-Apro’ (the latter represents affiliates in Asia, Pacific and Africa). IFAs or ‘global agreements’ as GUFs prefer to call them, are at the core of UNI campaigns targeting MNCs. UNI perceives IFAs as ‘global strategies’ that foster cross border cooperation between local and national trade unions and ensure that ‘multinationals come to the bargaining table in good faith in every country in which they operate’. However, in the case of Telefonica, the negotiation of an IFA was not triggered by UNI. The signature of the agreement was the culmination of union efforts from Latin America to bring their local issues to the attention of global corporate management. The text was signed in 2001 by the CEO, UNI and the telecom representatives of CC.OO and UGT.
Issues covered by the UNI-Telefonica IFA & references to ILO
1.
Free selection of employment: references to ILO Conventions 29 and 105
2.
Employment discrimination: references to ILO Conventions 100 and 111
3.
Child labour: references to ILO Conventions 138 and 182
4.
Freedom of association & the right to collective bargaining: references to ILO Conventions 87 and 98
5.
Worker representatives’ rights: references to ILO Convention 135
6.
Minimum wages: references to national legislation and collective agreements (as set out in ILO Conventions 94, 95 and 131)
7.
Working day: references to national legislation, collective agreements and industry standards (as set out in ILO Conventions 1, 47 and Recommendation 116)
8.
Health and safety: references to ILO Convention 155
9.
Working environment: commitment to harmonious relations and respect at the workplace
10.
Development and training: commitment to training especially with regard to the use of new technology
11.
Employment stability: references to national legislation and agreements
12.
Respect for the environment: general commitment to meet ‘all environmental concerns’
The Telefonica IFA: actors and processes

A

guaranteed seo

omega 3

Sell gold

hemorrhoids

Internet marketing

data recovery from crashed hard drive

Custom home builders Northern Virginia

sr22 insurance

work in haiti

emplois haiti

custom home builders Virginia

DC internet marketing

hcg diet

The affiliation of Latin American unions to UNI led to the creation of an international forum in the mid 1990s which comprised trade union representatives from the Latin American subsidiaries and also union representatives from Spain who met annually and discussed local labour and industrial relations issues. A formal invitation of the former president of Telefonica to participate in one of the meetings led to the signature of a ‘Social Protocol for International Agreements’ in 2000 which legitimized the operations of the forum (now called UNI-Telefonica trade union Alliance) and formalized the participation of corporate management. Although they were not directly involved in the negotiations, a number of trade unions from Argentina, Brazil, Chile, Mexico and Peru contributed in the formulation of the content of the 2000 Protocol by proposing (via UNI) the incorporation of specific provisions on trade union rights and health and safety. The document in turn set the grounds for the negotiation and elaboration of the IFA. The latter was signed in 2001 by UNI and both Spanish union federations. From a trade union perspective, corporate consent was largely facilitated by the public origins of the company and also by managerial prerogatives to improve corporate image and reputation. Indeed, for global headquarters, the signature of an IFA was in line with their social responsibility tradition. Headquarters also believed that it would further create the feeling amongst local management that they belonged to a ‘globality’. Management rationale for entering negotiations was therefore to ‘systematize’ social dialogue on a global scale albeit ‘to the extent possible’ and ‘respecting local traditions’.

Negotiations were rather smooth without any disagreements on the content of the IFA since most of the issues discussed were already addressed by the preceding Social Protocol and some by the company’s code of ethics. The agreement is called ‘UNI-Telefonica Code of Conduct’ and the name reflects the embryonic stage of IFAs and the widespread trend of corporate codes of conduct at the time of the negotiations. The document covers twelve areas: free selection of employment, child labour, discrimination, freedom of association and collective bargaining, employee representatives’ rights, health and safety, working environment, minimum salaries, working time, training and development, employment stability and respect for the environment. The core ILO conventions are referenced by their numbers whereas the text addresses issues beyond the basic labour rights including minimum wages, working day, training and employment stability. Yet, the latter make references to national legislation in order to ensure that ‘minimum conditions will apply’ (headquarter management) and hence their language is very generic. Implementation
Dissemination is stipulated as a bilateral initiative. The text dictates that “both parties shall be responsible for the administration and implementation of this agreement [and] to that end they will engage in an ongoing dialogue and they will meet regularly”. From the management side, company intranet was used as the main (if not the only) means for 11
dissemination at lower levels. Although the text makes it clear that the company “will provide information concerning [the] agreement to all companies of the Group”, communication of the agreement to the unionized subsidiaries was a priority for headquarter management. As they emphasized, local managers in non-unionized operations ‘would not really understand’ the spirit of the IFA and they acknowledged that they ‘have done very little’ in that respect. The above relates to corporate perceptions on the IFA scope and coverage. They argue that the agreement is applicable only to unionized subsidiaries where trade unions are affiliated to UNI. This view is largely shared by home-country unions. The latter argue that the IFA is ‘a commitment between two parties’ and when one of the parties is missing any implementation claims are difficult to be met. Interestingly, UNI-Americas has adopted a similar attitude. In the absence of trade unions, it is doubtful whether local managers will proceed with implementation. Yet, the GUF’s role would be to ‘use’ the IFA in order to facilitate organization of non-unionized subsidiaries while bringing existing unions under its umbrella.

In Peru, the IFA was disseminated only to fixed-telephony operations where more than 50% of the workforce is unionized. For fixed-telephony management, diffusion of the agreement to lower managerial levels and to the employees within the plant was ‘unnecessary’ since the IFA content ‘is not any different from what [they] do already’. Therefore, no new instruments were elaborated in order to put it in practice while it serves only as ‘a formal confirmation of the shared values of the Group’. Yet, trade unions interpret the above as company attempts to ‘silence’ the agreement. In the rest of the companies, managers were not supplied with a copy despite their central positions in human resources and legal departments and they ignored the existence of the IFA until the time of the research. Unionization levels of each of these companies amount to about 2% of the workforce while some of them were, at the time of the research, in the process of trade union recognition for collective bargaining purposes.
In Argentina, the IFA was disseminated to fixed and mobile telephony subsidiaries but also to media and consultancy operations. According to the HR director of the Argentinean group, the document has been communicated only ‘where relevant’, that is, to the senior HR manager of each company and where unions are present. In turn, managers in fixed and mobile telephony used company intranet to communicate it to lower levels but they did not elaborate any specific instruments for implementation. In the rest of the operations (media; consultancy services), management did not take any actions to disseminate it. Consultancy operations are not unionized while in the media operations, employees are organized either into small different trade unions of journalists also representing the different TV channels (local representation) or by the Argentinean television union (national representation). Union representation in the media sector is so fragmented and decentralized that there are more than twenty different unions across the country. These unions are not affiliated to UNI. Management argued that their companies are well represented by unions in the sector and hence it was not imperative to communicate and implement the UNI agreement. In their words, the IFA ‘reaffirms what [they] do anyway on a daily basis’.

12
Findings on the UK operations are very similar. The IFA was communicated to O2 after the acquisition by Telefonica. The words of O2 employee relation managers are indicative of Telefonica’s preference for a non-interventionist approach: ‘one of the first questions they asked was whether we were comfortable with [the IFA] … Some things vary according to local practice but we were comfortable with it’. The agreement was disseminated only ‘to the people that needed to know about it’, that is, to senior HR and employee relations management. No new instruments were developed for implementation and our respondents invoked the positive climate of industrial relations within the company and the UK institutional context: ‘from an O2 UK perspective, given the mature relationship we have with our trade unions and the high level legislation around the working environment, we comply with both the spirit and the ‘letter’ of the Code. We are not infringing any of the principles -we are in full compliance’ (HR manager). Monitoring
The monitoring of the Telefonica IFA at the global level is exercised via two different channels. The first is the joint monitoring committee foreseen by the text of the agreement. The committee consists of corporate management, one representative of each Spanish union organization and a UNI official. Although the IFA stipulates annual meetings, in practice the committee meets on an ad-hoc basis and the process is rather informal. For corporate management, the meetings are essential only in cases of local conflicts. In their words, ‘when there are no conflicts it means that everything is working well´. The second channel is the UNI-Telefonica trade union Alliance. During the international annual meetings local unions inform both corporate management and the president on conflicts and disputes but, as stressed by trade unions in Peru, the above meetings have two limitations. First, their objective is mainly informative and therefore they have not led to the resolution of conflicts. Secondly, local management is absent. From their part, managers in Peru, Argentina and the UK acknowledged the added value of the forum in establishing a form of permanent communication with the very president of the group and emphasized the lack of information regarding the outcomes of the meetings. The annual reports of the meetings remain with the participatory unions and the global HR team. At the local level, there are not any IFA-specific monitoring mechanisms either from the management or from the union side.

Enforcement of compliance
Findings from Peru and Argentina and also a number of conflicts in Colombia, Mexico, Puerto Rico and also Ireland reveal tensions between the IFA and national legislation. In all six countries, the majority of (non) compliance issues are related to Article 2 of the Telefonica IFA on freedom of association and the right to collective bargaining. The cases have been in the agenda of the Alliance meetings since early-mid 2000 and in essence most of them remain unsolved. In Atento Mexico, a subsidiary of Telefonica which provides services to big companies and their clients through call center operations, there are two trade union organizations: one with sectoral representation (telecommunications and transport-SPTCTRM) and one representing a specific profession (union of telephony workers –STRM). The latter is 13 affiliated to UNI but it is not recognized by the company for collective bargaining purposes. The company has a collective agreement only with the former trade union organization in full compliance with Mexican labour law. Legislation on trade union recognition stipulates that, in companies where industry unions and unions representing employees of a specific profession co-exist, a collective agreement with the latter will be reached only when the number of their members is bigger than the number of the employees of the same profession who are affiliated to the industry union. STRM does not meet the above criteria but it also argues that the union enjoying majority rights is in reality controlled by the company and it has signed one of the so-called ‘protection contracts’. The latter are ‘fake’ collective agreements concluded without the knowledge or participation of the employees and they serve managerial prerogatives of officially complying with the law while unofficially avoiding contractual obligations. STRM accuses Atento of violating Article 2 of the IFA. Yet, for corporate management, the IFA ‘is not above national laws’ and since the actions of Atento Mexico are ‘lawful’ they cannot intervene. Finally, Article 2 of the IFA makes explicit reference to ILO convention 98 on the right to collective bargaining. However, the latter has not been ratified by Mexico.

During 2006-2008, trade unions in Atento Puerto Rico faced similar issues of recognition for collective bargaining purposes. The company opposed the conduct of union elections for collective bargaining representatives –a position permitted by national legislation. The union filed a petition to the National Labor Relations Board to direct the election process after conducting a hearing. During the period culminating to the hearing, management engaged in an anti-union campaign including 15 lay-offs of employees who were members of the union organizing committee and distribution of letters to the workforce ‘informing’ them that trade union affiliation ‘deprives them of their individual rights before their employer’. UNI intervention resulted in the reinstatement of the above employees after a series of meetings and formal email communication with corporate management in Spain. Yet, the anti-union campaign did not cease and UNI representatives traveled to Puerto Rico to meet with local management and inform them of their ‘obligations’ as ‘parties of the Global Agreement between Telefonica and UNI’. Local managers refused to assume any such ‘obligations’ and corporate management proposed to both parties (company and local union) to suspend the elections in order to examine in which way neutrality would be maintained. Again Atento went on with the campaign until the issuing of a ‘Settlement Agreement’ by the National Labour Relations Board. The agreement obliged the company to post notices in the workplace stating that they would not interfere when employees exercise their rights to form and join a trade union and to negotiate collectively. Atento finally complied with the Settlement Agreement but the union lost the elections. In Colombia, Telefonica is a minor shareholder of Telebucaramanga, a telecom company owned by public local authorities. The company refuses to recognize the USTC (the UNI affiliate in Colombia) for collective bargaining purposes. The above is largely allowed by national laws on collective bargaining rights of public sector employees. Collective bargaining was the privilege of private sector employees until 1997 when it was extended to the public sector via ratification of the ILO Convention 155. Yet, the convention was 14
not fully ratified since the Constitutional Court considered it contradictory to internal legislation. USTC therefore resorted to Telefonica in order to achieve recognition under the IFA provisions but corporate management refused to intervene due to the company’s minor participation (less that 10%).

In Peru, trade union recognition issues emerged from the privatization processes of the company. ‘Telefonica del Peru’ was created by the merger of two formerly public companies represented by rather strong unions. After the acquisition, the company was subdivided into distinct legal entities according to different operations. Further acquisitions led to the emergence of numerous subsidiaries the majority with small workforces. The above resulted in extensive lay-offs while the employees who were selected to pass to a new company were ‘obliged to voluntarily resign’ [SETP] before signing a new contract. According to Peruvian unions, in most cases the new contracts entailed inferior terms and conditions of employment. In one case in 2001, trade unions signed an agreement with management stipulating company’s commitment to ensure that the employees would maintain their trade union rights. Apart from local instruments, the text made references to the Social Protocol and the IFA albeit in a brief and largely ‘political and symbolic’ fashion: the local agreement was reached only a few months after the signature of the IFA and it was the only time the IFA was incorporated in a local document. Trade union recognition for collective bargaining purposes is impeded by national legislation, which foresees collective bargaining at plant level whereas trade unions currently seeking recognition are either ‘telephony’ unions or industry unions. Unions have engaged in legal disputes with the company and they are also considering a merger in order to be able to negotiate and bargain at group level.

In Telefonica Argentina, trade union claims to organize mobile telephony and call center employees have been rejected on the grounds that both activities belong to the ‘commerce’ sector (i.e. retail and customer service) and not to telecommunications. The commerce sector is covered by national collective agreements that set the framework of terms and conditions of employment and leave detailed aspects to be agreed at company and/or plant level. Managers argue that employee contractual terms are agreed according to the sector agreements and hence are fair. In mobile telephony, management resorted to the ministry of labour invoking relevant legislation on company obligations to negotiate with unions that belong to the same activity and as a result, telephony unions lost the battle. In the words of union representatives, the company therefore ‘pays commerce salaries to telecoms workers’. In call centers, managers further argued that union representation was not an employee claim but more an attempt to be forced upon them by unions. However, there has been one case where the IFA had some impact on resolving a local dispute. The agreement partly contributed to the recognition of Cepetel, a small Argentinean union of engineers and technicians. Cepetel resorted to the Tribunals and used the IFA as a means to publicize the issue via UNI. It is striking that the two other unions representing Telefonica Argentina engaged in a legal battle against Cepetel in collaboration with the company because they were afraid they would lose the majority of representation. Other reasons behind this inter-union hostility comprised differences in 15 political backgrounds and orientations. Cepetel was eventually ‘legalized’ by a court decision and, in 2008, it was recognized by the company for collective bargaining purposes.
O2 Ireland does not recognize trade unions for collective bargaining purposes, while union membership is very low. The CWU represents around 3% of the staff in grievance and disciplinary proceedings. In 2007, the company went through a restructuring scheme. The plan was to outsource the company’s network technology division to external IT providers like IBM. The outsourcing meant that a number of employees would be made redundant. Some of the employees affected by the restructuring were members of the CWU and hence the union asked to be actively involved. The management agreed and signed a formal agreement with the CWU on the role of the latter in the restructuring process. After the outsourcing, the CWU asked for formal recognition but the company refused. The reasons outlined by O2 management were that membership was very low and that the rest of the employees did not express any interest to be represented by CWU. In a letter to the HR director of O2 Ireland, the CWU cited the IFA clause on trade union recognition and the right to collective bargaining as stipulated by ILO conventions 87 and 98, and emphasized that “it is now imperative that O2 Ireland honours this agreement and respects the wishes of the staff … In your reply you make it quite clear that your preference is to deal with staff ‘without the need for external involvement’ … [or] the ‘need for recourse to a third party’ … I fail to understand how Telefonica … could justify not applying the ILO Conventions to its Irish staff given that Ireland has ratified them”. The CWU argued that the Irish conflict was the main reason why union representatives went to the 2009 meeting of the Telefonica-UNI alliance. They spoke with the CEO and the global HR director who both said that they would talk to O2 Ireland management. Yet, the involvement of UNI and Telefonica headquarters did not resolve the dispute. Headquarters were aligned to the position of O2 management. In the words of the European HR director, “Telefonica was fully aligned with us. [The global HR director] wrote to UNI to reinforce that the UNI agreement guarantees rights for employees, not for trade unions and that it also guarantees neutrality by the company. So I can’t show favouritism to one union versus the other unions in Ireland unless it is because our employees have joined them … My management team in Ireland, if they did what the union said, they would actually feel it was a breach of the UNI agreement”.

Review
During the Telefonica IFA review and renewal process in 2007, there were debates on a possible reformulation of the clause on ‘neutrality’ mostly triggered by the recent dispute in Puerto Rico. UNI urged for a more prescriptive provision arguing that the problem was the lack of a clear definition of ‘neutrality’ and proposed the following wording: ‘aimed at not preventing or hindering the trade union from setting up in workplaces’. The company instead wanted to add the phrase ‘unless otherwise stated by national legislation’. In the end, the new agreement was worded according to UNI suggestions. However, for corporate management, it is very likely that similar conflicts will occur in the future and the added-value of the IFA is its potential to eventually ‘change local mentalities’. 16
Overall, Peruvian, Argentinean and UK unions evaluate the IFA as a positive instrument that could set the grounds for ‘global collective agreements’ albeit in the long-run. They argue that effective implementation is hindered by the agreement’s non-binding character, the reluctance of management to increase awareness and most importantly, the fact that they do not ‘really believe in the [IFA]: people trust more the things in which they are directly involved’ (Peruvian union official). Indeed, a number of trade unions from Argentina, Brazil, Chile, Mexico and Peru triggered the bottom-up communication with the Spanish federations and even suggested the incorporation of provisions on trade union rights and health and safety issues in the 2000 Social Protocol and subsequently in the IFA. Although a number of Latin American trade unions have even contributed in terms of content, local management has always been absent from IFA processes. Most importantly, not every union from each of the countries participating in the Alliance had been involved in the IFA formulation. The result is a tacit division between unions who support the agreement and are more pragmatic regarding its potential, and those who expect stronger commitments. From their part, local managers see the soft and generic character of the IFA as a one-way solution ‘for a tool applying to different localities’ –a view also shared by corporate management and interestingly, by home-country unions. The latter acknowledge the weaknesses of the agreement but their approach is more pragmatic.

Discussion
Multi-level dynamics
The Telefonica IFA processes from its conception to its review and renewal confirm the need to depart from the traditional division between management and labour and to examine the relationships between the different levels and realms of IFA governance. We have detected top-down dynamics in the cases of dissemination of the agreement by global headquarters to the senior management of the unionized Spanish, European and Latin American operations of the group and by UNI to local trade union leaders. Downward devolution of IFA responsibilities is weak (if not absent) in the cases of practical implementation and monitoring. Awareness-raising from the part of local unions has been minimal while monitoring mechanisms are missing. Similarly, an adaptation of the text to local realities was perceived as unnecessary by senior local management who neither disseminated the IFA to middle managers and employees within subsidiaries nor elaborated specific instruments for implementation and monitoring. Although Telefonica’s structure is hierarchical, we nonetheless find that strong top-down processes are impeded by the degree of autonomy enjoyed at lower levels. Local management has made a distinct use of autonomy in the case of the IFA when compared to the social responsibility and human resources policies that are elaborated at the global group level. All the arrangements formulated at the global level allow for considerable scope for adaptation which is usually the responsibility of local managers. Yet, the latter have chosen to endorse the IFA as it is. Interestingly, the above have both generated and have been generated by a duplication and overlap of policies and practices at more than one levels. The absence of specific
17 responsibilities and clear roles from the generic IFA content has allowed for different interpretations and therefore different outcomes. Where (strong) unions are present, the standards stipulated by local and/or national collective agreements supersede the minimum requirements of the IFA text which in turn is attributed a symbolic value in the eyes of local managers. In such cases, wage issues are an important union concern but the text makes clear references only to minimum standards as these are foreseen by national legislation and agreements. In the rare occasions when pay appeared in the agenda of the Alliance meetings, global headquarters disregarded it as a local issue. Where unions are not recognized and seek recognition, precedence is given to (unfavourable) local legislation and to sector arrangements (i.e. case of Argentina). Although the document stipulates the signatories’ commitment to the freedom of association and the right to collective bargaining as conferred by the relevant ILO conventions, the latter cannot automatically meet union recognition claims (i.e. Ireland). Moreover, in cases of non-ratifications (Mexico) or ‘patchy’ ratifications (Colombia), the ILO realm serves as a barrier to effective governance.

The above have been accompanied by the emergence of a number of bottom-up dynamics. Local unions use the annual meetings of the Alliance as a means for direct upward communication by bringing local issues to the attention of global headquarters. On a more frequent basis, they engage in indirect upward communication with the global human resources team via the Spanish union federations taking advantage of their proximity to the corporate centre and their good working relationship with management. Whether direct or indirect, upward communication has had positive outcomes. Global headquarters and UNI have taken responsibility for local disputes and have jointly mediated their impact (as, for instance, in the case of Puerto Rico). We therefore see how bottom-up dynamics can lead to top-down and vice versa. The division between local unions in terms of their expectations and the reformulation of the neutrality clause are two additional IFA outcomes that have been generated from the bottom up. The first has further implications for the proactive use of the IFA at the local level while the second is intended to add clarity to the text and hence limit the scope for opting for national laws and regulations. Therefore, upward delegation also relates to duplication and overlap.

Compliance
We find that multi-level dynamics determine actors’ choices between internal and external sources of compliance. In terms of internal sources, there is little evidence that the command and control of those who govern is systematically employed to enforce compliance. The mediator role of global headquarters and the GUF is explained by the scope of autonomy between and within levels. The rather soft approach of UNI, home-country federations and of the more pragmatic local unions is justified by the impact of internal interdependencies and convergence of needs: concessions are allowed in order not to endanger the sustainability of the IFA and to pave the way for stronger commitments in the future. The lack of local management participation in the IFA processes of formulation and monitoring, the fact that only a limited number of local unions were involved in the formulation process combined with inter-union rivalries account for local actor inertia and further confirm the role of needs and interests. 18 Moreover, issues of duplication explain how the references to other areas of governance are either deficient or contradict the IFA objectives.
On the other hand, evidence for the use of external sources of compliance is rather strong. The weaknesses of internal means have prompted local unions to resort to Tribunals with favourable outcomes (i.e. Argentina, Puerto Rico). The reliance on external sources is further explained by the deficiency of the multi-level IFA architecture as this is evident in the lack of a clear delineation between levels and the persisting impact of structural imbalances (i.e. sector/industry influences, issues of corporate governance and GUF representation). The above has resulted in marked differences between the involvement and the power of local actors: greater participation does not always translate into greater influence. Yet, our cases reveal that external and internal means can have joint effects which in turn explain the success of the outcome: upward delegation of responsibilities and invoking the moral force of the IFA have been efficiently combined with the employment of legal means.

Conclusions
Overall, the IFA multi-level dynamics shape actor choices towards external sources of compliance. We find that the internal means are usually employed for limiting the extent of disputes and the external for resolving them. Issues of autonomy and duplication between and within the different institutional levels impede their effective integration in ways that promote the goals of the system overall. In other words, they impede the exercise of effective governance. We see two main implications. First, the systematic reliance on external means of enforcement might in the long run outweigh the rationale for a global tool. Second and related to the first, power imbalances when combined with structural influences might ultimately change the discourse from ‘IFA governance’ to ‘IFA participation’. It is therefore important that global labour arrangements generate clearer expectations with regard to the responsibilities of (local) actors in raising and addressing complaints. The sustainability of IFAs is also endangered by the absence of institutionalised sanction arrangements accompanied by a growing emphasis on consensus and accommodation that in turn serves as a trade off for democratic principles. Further research is hence required to examine the potential of private arrangements including IFAs to address the democratic deficit of global labour governance. REFERENCES
Anderson J. E. (1994) Public Policymaking: an introduction. Boston: Houghton Mifflin.
Burnett M. and Duvall R. (2005) Power in global governance. Cambridge: Cambridge University Press.
19
Commission on Global Governance (1995) ‘Our Global Neighborhood’. Oxford: Oxford University Press. Accessed at: www.sovereignty.nte/p/gov/chap1.htm
Croucher R. and Cotton E. (2009) Global unions, Global Business. Middlesex University Press.
Cutler C.A., Haufler V. and Porter T. (1999) Private authority and international affairs. Albany: State University of New York Press.
Daugareilh I. (2008) ‘Corporate norms on Corporate Social Responsibility and international norms’ in (ed.) Governance, International Law & Corporate Social Responsibility, ILO, p.63-78.
Dingwerth K. and Pattberg Ph. (2006) ‘Global Governance as a Perspective on World Politics’, Global Governance, 12: 185-203.
Egels-Zandén N. (2009) ‘Transnational Governance of Workers’ Rights: Outlining a Research Agenda’, Journal of Business Ethics, 89:169-188.
Falkner R. (2003) ‘Private environmental governance and international relations: exploring the links’, Global Environmental Politics 3(2).
Finkelstein L. S. (1995) ‘What is global governance’? in Global Governance 1(3).
Gereffi G. and Mayer F. W. (2004) ‘The Demand for Global Governance’. Terry Sanford Institute for Public Policy, DUKE. Working Paper Series SAN04-02.
Hall R. B. and Biersteker Th. J. (2002) ‘The emergence of private authority in the international system’ in R. B. Hall and Th. J. Biersteker (ed.) The Emergence of Private Authority in Global Governance. Cambridge, Cambridge University Press, 3-22.
Hassel A. (2008) ‘The evolution of a global labor governance regime’, Governance: An International Journal of Policy, Administration, and Institutions, 21: 231-251.
Hewson M. and Sinclair T. J. (1999) ‘The Emergence of Global Governance Theory’ in Martin Hewson and Timothy Sinclair (ed.) Approaches to Global Governance Theory. State University of New York Press, 3-22.
Kahler M. and Lake D. A. (2003) ‘Globalization and Governance’ in Miles Kahler and David A. Lake (ed.) Governance in a Global Economy. Princeton and Oxford: Princeton University Press, 1-30.
Leisink P. and Hyman R., ‘Introduction: the dual evolution of Europeanisation and varieties of governance’, European Journal of Industrial Relations, 11: 277–86.
20

Léonard E., Erne R., Marginson, P. and Smismans S. (2007) New structures, forms and processes of governance in European industrial relations. European Foundation for the Improvement of Living and Working Conditions.
Marginson, P. and Sisson, K. (2004) European integration and industrial relations. Multi-level governance in the making. Basingstoke: Palgrave.
Marks G and Hooghe L (2004) ‘Contrasting visions of multi-level governance’ in Ian Bache and Matthew Flinders (ed.) Multi-level Governance. Oxford: Oxford University Press, 15-30.
Murphy C. N. (2000) ‘Global governance: poorly done and poorly understood’ in R. Wilkinson (ed.) The global governance reader. Oxon: Routledge, 90-104.
Overbeek H. (2005) ‘Global governance, class, hegemony’ in Alice D. Ba and Matthew J. Hoffmann (ed.) Contending Perspectives on Global Governance. Oxon: Routledge, 39-56.
Pattberg Ph. (2005) ‘The institutionalization of private governance: How business and nonprofit organizations agree on transnational rules’, Governance: An International Journal of Policy, Administration, and Institutions, 18(4).
Pattberg Ph. (2006) ‘Global Governance: Reconstructing a Contested Social Science Concept’, GARNET Working Paper: No 04/06
Peters and Pierre (2004) ‘Multi-level governance and democracy: a Faustian bargain?’ in Ian Bache and Matthew Flinders (ed.) Multi-level Governance. Oxford: Oxford University Press, 75-89.
Riisgaard L. (2005) ‘International framework agreements: A new model for securing workers rights?’, Industrial Relations 44(4).
Rosenau J. N. (1992) Governance, Order, and Change in World Politics, in James N. Rosenau and Ernst-Otto Czempiel (ed.) Governance without Government. Cambridge: Cambridge University Press.
Rosenau J. N. (1997) Along the domestic-foreign frontier. Exploring governance in a turbulent world. Cambridge: Cambridge University Press.
Rosenau J. N. (1999) ‘Toward an ontology for Global Governance’ in Martin Hewson and Timothy Sinclair (ed.) Approaches to Global Governance Theory. State University of New York Press, 287-301.
21
22
Rosenau J. N. (2004) Strong Demand, Huge Supply: Governance in an Emerging Epoch, in Ian Bache and Matthew Flinders (ed.) Multi-level Governance. Oxford: Oxford University Press, 31-48.
Scharpf F. W. (2000) ‘Institutions in Comparative Policy Research’, Comparative Political Studies, 33: 762-790.
Scott J. and Trubek, D., ‘Mind the gap: law and new approaches to governance in the European Union’, European Law Journal, 8: 1–18.
Seifert A. (2008) ‘Global Employee Information and Consultation Procedures in Worldwide Operating Enterprises’, International Journal of Comparative Labour Law and Industrial Relations, 24: 327-348.
Sisson K. (2007) ‘Revitalising industrial relations: making the most of the “institutional turn”, Warwick Paper in Industrial Relations, No 85.
Stevis D. (2009) ‘International framework agreements and global social dialogue: lessons from the Daimler case’, ILO, Employment Working Paper No 46.
Whitman J. (2009) The fundamentals of global governance. Basingstoke Pallgrave Macmillan.
Wills J. (2002) ‘Bargaining for the space to organize in the global economy: a review of the Accor-IUF trade union rights agreement’, Review of International Political Economy 9(4).