Implementation of Global Framework Agreements

Abstract

Over the past decade an increasing number of Global Framework Agreements (GFAs) have been signed, most of which – over 80% – have been with European-based Transnational Corporations (TNCs). But while GFAs are slowly, but continually increasing in numbers, the results of our empirical research reveal extensive deficits in their implementation. This paper begins with a review of these implementation problems and challenges. Drawing on two exemplary case studies, we introduce our multi-organizational practice perspective to illuminate how the contents of GFAs as negotiated are linked to implementation, conflict detection and resolution. We conclude by presenting a proposal for a systematic and integrated implementation procedure as a means of dealing with the complexity of the GFA process. This procedure is built on three sets of practices – information dissemination and communication; training programs; and operational measures – and encompasses both unilateral and joint policies and actions pursued by Global Union Federations and TNC management.

Full text

http://trs.sagepub.com/
of Labour and Research
Transfer: European Review
http://trs.sagepub.com/content/20/4/489
The online version of this article can be found at:
DOI: 10.1177/1024258914546270 2014 20: 489Transfer: European Review of Labour and Research
Jörg Sydow, Michael Fichter, Markus Helfen, Kadire Zeynep Sayim and Dimitris Stevis
practice perspective
Implementation of Global Framework Agreements: towards a multi-organizational
Published by:
http://www.sagepublications.com
On behalf of:
European Trade Union Institute
can be found at:Transfer: European Review of Labour and ResearchAdditional services and information for
http://trs.sagepub.com/cgi/alertsEmail Alerts:
http://trs.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
What is This?
- Nov 7, 2014Version of Record >>
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

Main Article
Implementation of Global
Framework Agreements:
towards a multi-organizational
practice perspective
Jo
¨rg Sydow
School of Business & Economics, Freie Universita
¨t Berlin, Germany
Michael Fichter
Global Labour University, Berlin, Germany
Markus Helfen
School of Business & Economics, Freie Universita
¨t Berlin, Germany
Kadire Zeynep Sayim
Hotelschool the Hague, the Hague, the Netherlands
Dimitris Stevis
Department of Political Science, Colorado State University, USA
Summary
Over the past decade Global Union Federations have signed an increasing number of Global
Framework Agreements, most of which – over 80 per cent – have been with European-based
Transnational Corporations. But while Global Framework Agreements are slowly, but continually,
increasing in numbers, the results of our empirical research reveal extensive deficits in their imple-
mentation. This article begins with a review of these implementation problems and challenges.
Drawing on two exemplary case studies, we introduce our multi-organizational practice perspec-
tive to illuminate how the contents of Global Framework Agreements as negotiated are linked to
implementation, conflict monitoring and resolution procedures. We conclude that there is a
need for a systematic and integrated implementation procedure as a means of dealing with the
complexity of the Global Framework Agreement process. Our proposal is built on three sets of
practices – information dissemination and communication; training programmes; and operational
Corresponding author:
Michael Fichter, Global Labour University, Matterhornstr. 87. 14129 Berlin, Germany.
Email: mike.fichter@global-labour-university.org
Transfer
2014, Vol. 20(4) 489–503
ªThe Author(s) 2014
Reprints and permission:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/1024258914546270
trs.sagepub.com
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

measures – and encompasses both unilateral and joint policies and actions pursued by Global
Union Federations and the management of Transnational Corporations.
Re
´sume
´
Au cours de la dernie
`re de
´cennie, les Fe
´de
´rations syndicales internationales ont signe
´un nombre
croissant d’accords-cadres internationaux, dont la plupart – plus de 80 pour cent – avec des entre-
prises transnationales base
´es en Europe. Ne
´anmoins, alors que le nombre d’accords-cadres inter-
nationaux augmentent lentement mais de manie
`re continue, les re
´sultats de notre recherche
empirique re
´ve
`lent de profondes lacunes quant a
`leur mise en œuvre. Le pre
´sent article commence
par analyser ces proble
`mes et de
´fis lie
´sa
`leur application. Sur la base de deux e
´tudes de cas exem-
plaires, nous pre
´sentons notre approche de la pratique multi-organisationnelle en vue de mettre
en lumie
`re la fac¸on dont le contenu des accords-cadres internationaux tels qu’ils sont ne
´gocie
´s est
lie
´a
`la mise en œuvre, au suivi des conflits et aux proce
´dures de re
´solution. En conclusion, nous
pouvons affirmer qu’une proce
´dure de mise en œuvre syste
´matique et inte
´gre
´eestne
´cessaire
comme moyen de traiter la complexite
´du processus d’accord-cadre international. Notre proposi-
tion repose sur trois ensembles de pratiques – diffusion d’informations et communication, pro-
grammes de formation et mesures ope
´rationnelles – et englobe tant les politiques unilate
´rales
et conjointes que les actions mene
´es par les Fe
´de
´rations syndicales internationales et le manage-
ment des entreprises transnationales.
Zusammenfassung
In den vergangenen zehn Jahren haben globale Gewerkschaftsverba
¨nde eine zunehmende Zahl von
globalen Rahmenabkommen unterzeichnet, die meisten – mehr als 80 Prozent – mit in Europa
ansa
¨ssigen transnationalen Unternehmen. Zwar steigt die Zahl dieser globalen Rahmenabkommen
langsam, aber stetig, die Ergebnisse unserer empirischen Untersuchung zeigen jedoch umfassende
Defizite bei ihrer Umsetzung. Dieser Artikel beginnt mit einer U
¨bersicht u¨ber diese Umsetzungs-
probleme. Anhand zweier exemplarischer Fallstudien entwickeln wir eine Perspektive sozialer
Praktiken, die es uns erlaubt, die Ergebnisse von Verhandlungen (die Vereinbarungsinhalte) mit den
Prozessen ihrer Umsetzung sowie der Bearbeitung und Lo
¨sung von Konflikten zu verbinden. Wir
kommen zu dem Schluss, dassein systematisches und integriertes Umsetzungsverfahren erforderlich
ist, das die Komplexita
¨t der Prozesse globaler Rahmenabkommen beru¨cksichtigt. Unser Vorschlag
baut auf drei Gruppen sozialer Praktiken auf – Information und Kommunikation, Weiterbildungs-
programme, operative Verfahren – und beinhaltet sowohl unilaterale als auch gemeinsame Strategien
und Handlungen der globalen Gewerkschaftsverba
¨nde und des Managements transnationaler
Unternehmen.
Keywords
Global Framework Agreements, social dialogue, implementation, global union federation,
transnational corporation, international labour standards
Introduction
Global Framework Agreements (GFAs) are still a unique and not widely recognized approach to
regulating global labour relations. But they are increasingly drawing the attention of academics
and practitioners as the global version of a wide range of EU-specific transnational company or
490 Transfer 20(4)
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

framework agreements (Telljohann et al., 2009; Ru
¨b et al., 2011; Fichter et al., 2011; Fichter et al.,
2012a; Fichter et al., 2012b; Leonardi, 2012). In contrast to the usually unilateral and voluntary
character of initiatives based on corporate social responsibility (CSR), GFAs are bilateral agree-
ments between Transnational Corporations (TNCs) and Global Union Federations (GUFs). As
implied by the term ‘framework’, a GFA provides a social arena for applying the core labour stan-
dards and the basic principles of labour relations it delineates. In most cases, this arena extends
beyond the formal organizational boundaries of the signatory TNC to cover all the operations of
a TNC and parts of its ‘global production network’ (Coe et al., 2008). In 2000, over a decade after
the first steps toward GFAs were taken (Gallin, 2008), only eight GFAs had been negotiated. By
the end of 2013, however, over 100 agreements had been signed, with 92 of them functional.
1
To
date, negotiating and signing of GFAs is still largely a European phenomenon, with only 17 of
them having been concluded with non-European TNCs.
While the importance of this growth in numbers should not be underestimated, putting the
agreed measures into practice, i.e. the implementation of existing GFAs, is arguably key to their
success as a transnational trade union policy. How do GFAs fare in practice? was the main research
question of a recently completed project conducted by an interdisciplinary and international team
of academics coordinated by researchers from the Freie Universita¨t Berlin. Our qualitative assess-
ment of the status of implementation is based on semi-structured interviews conducted by our
international project team for a total of 19 selected GFAs. In sum, we conducted over 150 inter-
views, 74 of them with representatives of management and labour at headquarters level and 54
at corporate production sites in India, USA, Brazil and Turkey. As such, we are able to reflect
on the multi-organizational composition of ‘capital’ and ‘labour’ perspectives, respectively.
As documented in the final research report
2
as well as in the three country reports that we have
so far published (Arruda et al., 2012; Fichter et al., 2013; Fichter and Stevis, 2013), along with
several other articles in academic journals (Fichter et al., 2011; Fichter et al., 2012b; Fichter
et al., 2012a; Helfen and Sydow, 2013), implementation has been limited and has proven deficits,
regardless of the particular national environment. Importantly however, in each of the countries in
which we conducted empirical research, we found clear evidence of at least partially successful
implementation within a multi-organizational environment. In two cases in Brazil, a local trade
union in a highly organized auto plant invoked the GFA signed by the parent TNC in support
of union officials at suppliers to the plant who had been fired. Beside the two unions, local man-
agement of the TNC subsidiary and the two suppliers were involved. The union officials were rein-
stated. At another Brazilian subsidiary of a major auto supplier, local management refused to
bargain collectively in good faith until the union joined forces with the responsible GUF and the
home country union and pressured the TNC’s headquarters management into honouring the GFA
at the local plant (Arruda et al., 2012). In the USA, we found a few cases in which there was a similar
pattern of transnational union networking to put pressure on headquarters management (Fichter and
Stevis, 2013). In our research in Turkey, we came across another example of such transnational
union networking. This case was marked by the fact that it involved two competing local unions, two
different GUFs, the local management of the supplier and the TNC subsidiary as well as the TNC
headquarters management (Fichter et al., 2013). Finally, in India, a concerted transnational effort
in the service sector brought together competing union federations to build a coalition for organizing
and forcing the targeted TNC to comply with national law (McCallum, 2013).
1 Non-functional are those GFAs in TNCs that have merged, been acquired, or have folded.
2 http://boeckler.de/11145.htm?projekt¼S-2008-141-2%20F
Sydow et al. 491
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

These examples are important because they all indicate that successful implementation has so far
been more the result of mobilized union pressure – resulting in positive corporate responses – rather
than the existence of clearly defined policies and procedures in the GFA. Union capacity building is a
necessary foundation for ensuring that negotiation and implementation of GFAs will be equally effec-
tive especially when confronted by management resistance. But if GFAs are to gain more legitimacy
and recognition as a means of establishing dialogue, attaining minimum standards and fostering labour
relations, then there is a need for a more systematic and robust approach to implementation that is
incorporated multi-organizationally into the overall GFA process of negotiation and implementation.
The widespread incidence of non-implementation and the cases of partial or subsidiary-specific imple-
mentation, we would argue, provide evidence that implementation needs to be treated as part of a
multi-organizational practice development process starting with the initiative to negotiate a GFA.
While details may vary due to the peculiarities of a given production networkand TNC, there are basic
elements of this implementation process that can be generally applied, i.e. a robust set of practices and
guidelines to deal effectively with the complexity in multi-organizational arrangements.
In the following sections of this article we will first present our arguments for the need for a
multi-organizational practice perspective in understanding and analysing the GFA process. Then
we use two very dissimilar case studies drawn from our empirical research to illustrate our argu-
ments. Our case presentations are structured according to a general process model that we devel-
oped from our observations and evaluations. Finally, we use this model for explaining
implementation in these two cases, supported by references to further cases, to derive our proposal
for constructing a more robust implementation practice.
The multi-organizational practice of GFA negotiation
and implementation
Our starting point is a multi-organizational practice perspective. From such a perspective, the
implementation of a GFA is the outcome of the interaction of many organizations acting ‘on
behalf’ of capital and labour; an interaction that is contested and takes place in an institutionally
complex and potentially diverse environment that structures and is structured by this interaction
(Giddens, 1984). The multi-organizational dynamics of GFAs are apparent from their negotiation
on through to their implementation.
By definition, a GFA is an agreement signed by one (or more) GUFs and the management of a
TNC. Yet a closer look at the signatures on the agreement points to the existence of a complex
organizational interaction. Our analysis of the negotiation process (Fichter et al., 2011; Helfen and
Sydow, 2013) and the work of others (e.g. Hammer, 2005; Stevis, 2009, 2010; Papadakis, 2011)
shows that the national unions in the home country of the TNC and (European) works councils are
frequently important actors on the ‘labour’ side. On the ‘capital’ side, representatives outside of
central management are usually excluded. However, as a variety of workshops organized by
national employers’ organizations and legal consultants in cooperation with the International
Organization of Employers (IOE) show, TNC negotiating policies on GFAs do not develop in a
strategic or political void.
3
3 Members of our research team participated in one such workshop organized by the Confederation of
German Employers’ Associations (BDA) and ACTEMP, the employers’ group at the ILO in 2010 in
Berlin. We were also invited to make a presentation at a seminar conducted by the international law firm
Evershed on ‘Labour Relations across Borders’. The majority of attendees at both events were corporate
managers.
492 Transfer 20(4)
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

With the commencement of the implementation phase, the potential number of organizational
actors involved is significantly increased. Beyond the signatory organizations, which remain
involved via implementation and monitoring provisions, national and local representatives of cap-
ital and labour beyond the TNC home country are recognized as agents for putting the agreement
into practice (Stevis and Boswell, 2007). It is true that the implementation of the GFA has been
first and foremost regarded as the responsibility of the TNC with the unions attempting to fulfil
a monitoring role. However, the fulfilment of central management’s commitment varies greatly
in terms of its interest and capacity. As our research has shown, there is a significant amount of
diversity in management’s implementation policies and practices, ranging from proactive efforts
to non-involvement. The interplay of such factors as the degree of organizational hierarchy and
the more or less active role of subsidiaries; quality of supplier relationships; information policies;
available resources; and managerial prerogative may all affect corporate GFA policy, and, ulti-
mately, the extent of management’s pursuit of implementation. Simply put, just as with corporate
approaches to social responsibility, such factors can conceivably generate management policies
ranging from pro forma implementation to embedded implementation, with a variety of partial
implementation policies and practices in between.
While the thrust of management policy is a crucial and possibly a dominant determinant, our
analysis of implementation as a kind of ‘micro-politics’ would be too narrow if we were to restrict
it to ‘management’ (Burns, 1961). Besides the potential multiplicity of organizational actors in the
category of ‘management’ we have a significant spectrum of organizational actors representing
labour. Based on their involvement in the negotiation phase, GUFs and other representatives of
labour such as home country trade unions and (European) works councils will want to have a say
in implementation. But since implementation is always local, i.e. at each workplace, host country
national and local unions as well as employee representative bodies, should they exist, will be
involved (Stevis and Boswell, 2007; Stevis, 2009). This range of primary actors may be extended
to include state and supra-state agency and non-governmental organizations, especially if conflicts
develop around the practice of implementation.
4
Moreover, in the local context, headquarters as
well as regional management may involve consultants and national and/or global employers’
organizations.
To summarize our argument up to this point, the GFA process represents an interaction taking
place in a complex and diverse institutional environment that, at the same time, is reproduced or
transformed by this very interaction (Giddens, 1984; Barley and Tolbert, 1997; Sydow and Winde-
ler, 1998). The result of this interaction, as much as the process itself, is contingent not only upon
this environment but also upon the power relationships among the organizations and collective
actors involved in the ‘contested fields’ (Levy, 2008; Amoore, 2002) of labour relations in TNCs.
Consequently, the scope of a GFA as well as its implementation will vary substantially, not only
from one corporation to another but also within one TNC and its global production network. For
example, in comparing TNCs, the extent to which headquarters management has been able to dom-
inate the GFA negotiations based on an existing active policy towards CSR will probably be
reflected in the GFA’s implementation, possibly even in regard to legally independent businesses
in the global production network. On the other hand, in an institutional environment with powerful
and active unions, the implementation of the agreement may be more reflective of a strong
input from the ‘labour’ side. Indeed, the complexity of the overall process, combined with its
4 The recent conflict over union recognition at the Volkswagen plant in Chattanooga USA is an extreme but
also instructive example of such an involvement. See for example Elk, 2014.
Sydow et al. 493
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

geographical fragmentation and the involvement of a large number of autonomous and semi-
autonomous organizations characteristic of many modern TNCs (Hedlund, 1986) are consistently
strong enabling factors for varying outcomes within a single TNC and its global production
network.
Multi-organizational practices: from GFA negotiation to implementation
We noted above that the multi-organizational dynamics of GFAs become increasingly evident as
the process moves towards implementation. Given these dynamics the chances of successful
implementation depend, to a large degree, on whether the content of a GFA has been crafted with
such a dynamic in mind. A ‘strong’ GFA may go well beyond the core labour standards of the
International Labour Organization (ILO), include suppliers and possibly even subcontractors, con-
tain monitoring and sanction mechanisms, and lay out detailed procedures for conflict resolution in
the TNC and its network. ‘Weak’ GFAs, by contrast, may be confined to the core labour standards,
not include suppliers, abstain from monitoring and sanctioning, and not contain defined procedures
for conflict resolution. While such a simple bipolar typology is inadequate for making a detailed
analysis of all GFAs, it does allow for contrasting and highlighting basic differences.
Similarly, the strong-weak continuum is also applicable to implementation practices which take
place in a particular institutional context and are pursued by multi-organizational actors on the
basis of the negotiated agreement. Clearly, there is also the relationship between the agreement
(strong-weak) and its implementation (strong-weak). Weak GFAs are not likely to result in strong
implementation. And strong GFAs are not a guarantee for strong implementation in the context of
complex multi-organizational dynamics.
On this basis we distinguish four generic types of practice, ranging from ‘good’ to ‘bad’. Within
this range we also find ‘promising’ and ‘symbolic’ practices (see Figure 1).
We define implementation of an GFA as any practice that communicates, promotes, changes,
transfers or at least stabilizes a certain (inter-) organizational practice that directly or indirectly
improves or ensures minimum labour standards and collective labour relations agreed upon in the
negotiation process of an GFA. This applies not only at the headquarters level of a TNC but also to
its local subsidiaries. Although highly unlikely in the immediate future, the long-term goal is for
full coverage of suppliers throughout the TNC’s global production network.
To evaluate implementation and as a proposal for a more systematic structuring of the process,
we postulate implementation as three sets of practices. We define information and communication
practices as the intra- and inter-organizational tools used to inform the public and employees
(including managers) throughout the TNC and its global production network of the existence,
intention and content of the GFA. Similarly, we describe training practices as measures used
Implementation
Content ‘‘strong’’ ‘‘weak’’
‘‘strong’’ Good practice Symbolic practice
‘‘weak’’ Promising practice Bad practice
Figure 1. Content and implementation practices.
494 Transfer 20(4)
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

by TNCs and GUFs individually and, ideally, jointly, to further and improve the use of the GFA by
management, employees and union representatives not only at the TNC’s headquarters but also at
its subsidiaries. Possibly, such practices could include first-tier suppliers, which in turn may imple-
ment such programmes in their reactive or even strategic ‘supplier development’ policies and prac-
tices (Krause et al., 1998). In this manner, labour standards could be achieved in significant parts of
the global production network.
Lastly, operational practices are defined as the actions taken to change local conditions and
practices that provoke violations of labour standards. The development of such practices is also
a clear indication of the willingness and/or capability of management (and unions, works councils
etc.) actually to improve working conditions and labour relations throughout the TNC, its foreign
subsidiaries and possibly even its first- or second-tier suppliers. Among these operational practices,
monitoring practices rank highly. These might be distinguished between different levels of union
involvement: (1) self-monitoring by management, (2) monitoring by external agencies, (3) partic-
ipation of union representatives in the monitoring process, and (4) monitoring by employees, i.e.
union members.
Based upon our multi-organizational perspective we have constructed an implementation
scheme with three sets of practices (Table 1 below) that captures the recurrent activities of the GFA
implementation process in the TNC and its global production network and may be used for asses-
sing or comparing GFA implementation in or across individual cases. In the following presentation
of two case studies we will evaluate the GFA process along the lines of this proposal.
Exemplary evidence from case studies
The two cases we have selected for presentation in this article, MetalCorp and SecureCorp
5
, rep-
resent what might be regarded as ‘good practices’ of negotiation and implementation (see Figure 1
above) within an overall situation of rather weak implementation across all the 19 cases we have
studied in more detail. For this reason, they offer far more possibilities to evaluate instruments of
actual practice and paths only partially taken or envisioned but not started than most of our other
cases.
Information on GFA implementation was collected at the headquarters level and at the subsidiar-
ies in all four countries of our study. Both companies are global market leaders in their respective
industries with relatively large multinational production, supply and service networks respectively.
Besides operating in different industries, they differ markedly in the ways in which their GFA was
negotiated and has been implemented. As such, both companies might be said to be exceptional as
well as exemplary cases for lessons to be learned regarding implementation practices.
How negotiations affect later implementation practices
The multi-organizational dynamics of the GFA process were evident from the very beginning.
MetalCorp is one of the very few companies in which union representatives from the most impor-
tant subsidiaries (USA, Brazil) were officially involved in negotiating the agreement. This was
facilitated by a corporate world council structure for union and employee representatives backed
by a strong national union and a culture of co-determination in the home country of MetalCorp.
The responsible GUF, the International Metalworkers’ Federation, was not involved directly in the
5 The case study research was conducted on the basis of anonymity for the TNCs.
Sydow et al. 495
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

negotiations because of management’s unwillingness to deal with an ‘outsider’. As a result of this
organizational constellation on the labour side, the GFA has not been used systematically to orga-
nize and build unions (Helfen and Fichter, 2013). However, the GFA standards have been incor-
porated by MetalCorp into its master contractual agreement with suppliers.
The GFA that resulted from these negotiations is a case of ‘strong’ content. Although Metal-
Corp management was selective in its recognition of organizational representatives of labour and
the ILO core labour standards are not referred to by name but only in general terms, the agreement
uses appropriate language to define these standards. And it goes beyond most GFAs in its inclusion
of corporate and managerial neutrality during union organizing campaigns. Suppliers are strongly
urged to comply with the principles of the agreement and a well-defined procedure for its disse-
mination and monitoring as well as for reporting violations and for conflict resolution is included.
Table 1. Implementation practices.
Transnational Corporations (TNCs) Global Union Federations (GUFs) Joint Initiatives
Set 1. Information and
communication practices
Internet/intranet
Translations
Distribution of commentated
version to employees
Distribution to suppliers with
commentaries
Reference in social reporting
Publication of evaluation
reports and criteria
Set 1. Information and
communication practices
Internet/intranet
Translations, commentaries
to affiliates
Publications for media,
NGOs
Public presentations
Set 1. Joint I & C Practices
Press conferences
Joint web page
Roundtables
Reporting of evaluation
criteria
Commentaries
Communication channels
Set 2. Training practices
Manager training, group/
individual all levels
Special country-specific
training
Special product division
training
Subsidiary – supplier training
Set 2. Training practices
Inter- and Intra-GUF
training
Country-specific training
with affiliates
Cross-country training
workplace reps
Training TNC and supplier
unions
Set 2. Joint Training
practices
HQ management, GUF,
home
country union
Country-specific
management-union
training
Product and supplier
specific training
Set 3. Operational practices
Integration into operations,
organizational
restructuring, resources
Formalized complaint
procedures, monitoring
Hotlines
Incentive schemes
Supplier contracts
Quality-Management
Capacity building subsidiaries,
suppliers
Set 3. Operational practices
Organizational restructuring
around GFAs
Redistributing resources
Transnational union
networks
Support for local union
organizing
Monitoring and reporting
procedures
Set 3. Joint operational
practices
Joint task forces
Joint roll-out programme
Joint monitoring
Joint evaluation
Joint revision procedures
496 Transfer 20(4)
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

While MetalCorp management was favourably disposed towards negotiations over a GFA, the
situation at SecureCorp was by comparison highly contested and acrimonious. In this case, a US
affiliate of the global union, facing intransigent opposition from local management, devised and
initially led a strong campaign for union recognition at the US subsidiary. The global union (UNI
Global Union) subsequently became involved in order to campaign for a global agreement. After a
drawn-out battle, the national union at TNC headquarters came in to mediate and to overcome
strong resistance from a management team that felt it was being forced into an agreement. No
works council was involved, and UNI affiliates only played an indirect role with a range of soli-
darity actions during the campaign.
The SecureCorp GFA is also a ‘strong’ agreement. Representatives of labour and management
are fully and mutually recognized. The ILO core labour standards are included and listed by name.
Regarding union rights, the GFA is very explicit regarding freedom of association, union access to
workplaces and union recognition. While the issue of extension to suppliers is hardly mentioned,
the GFA contains detailed procedural language on implementing and monitoring.
A major difference between the two cases is that neither labour nor management at MetalCorp
regarded the GFA as being critical for improving labour standards within the company. Instead it is
the supplier level, and particularly, the suppliers of the second and the third tier which are of pri-
mary concern. That is not to say that there are no problems at the subsidiary level; however, these
are thought of as being taken care of by procedures and bodies already in place (e.g. the world
council). For SecureCorp, the situation is totally different in this regard because it is at the end
of the supply chain (by providing auxiliary services to large corporate and public customers) and
has no global structure of employee representation, so that – with the exception of some specialized
suppliers – the subsidiaries themselves are the focus of concern.
Implementation I: Information and communication practices
In both cases, corporate headquarters informed the public and disseminated the GFA through nor-
mal internal channels of organizational communication. Moreover both companies have translated
their GFA into several languages relevant to the locations of their production subsidiaries. Neither
company has published the agreement on their public website, although a reference to the agree-
ment may be found on the SecureCorp website. While standards and principles embodied in the
MetalCorp GFA may be found in its basic CSR statements, the GFA itself is not referenced. Metal-
Corp has also initiated a biannual conference programme on sustainability at its headquarters loca-
tion to which external stakeholders (academic experts, media and NGOs) are invited to discuss the
company’s performance.
However, both TNCs lack, although to differing degrees, a comprehensive communication
strategy for presenting the GFA internally or publicly as a joint labour-management policy state-
ment. At MetalCorp, local management in Turkey and the local union in Brazil took the initiative
on their own to inform employees and to communicate the GFA. In Turkey, the agreement was
translated and a printed copy distributed to all subsidiary workers. Local management also trans-
lated MetalCorp’s ‘code of conduct’, into which the content of the GFA had been incorporated,
into Turkish
6
(Fichter et al., 2013). Our research turned up nothing similar on the part of local man-
agement at MetalCorp subsidiaries in India, Brazil and the US. Two further aspects of this case are
6 ‘[MetalCorp] Davranıs¸ Kuralları’ in Turkish, of which we have a copy. This document includes the
clauses of the GFA, without naming it as such.
Sydow et al. 497
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

interesting. First, headquarters management apparently made no effort to present the GFA as a
joint labour-management policy agreement. Rather, this corporate document was introduced to the
management in Turkey as ‘a new management style’ and ‘placed at the centre of the future man-
agement plans of the company’ (Interview MetalCorp, HR Director Turkey, 2010). Secondly, the
Turkish subsidiary is unionized, but the union is not GUF-affiliated and has stated that it had no
interest in the GFA. That stands in sharp contrast to the active engagement and the use of the GFA
by the Brazilian union representing workers at the main MetalCorp plant in that country (Arruda
et al., 2012; Fichter and Helfen, 2011).
Importantly, both companies are without an explicit bottom-up communication process in which
local actorshave to confirm their knowledge of the agreement. In the Turkish case,such an opportunity
is limited to management since the local union is neither affiliated with the relevant GUF nor does it
have any strong ties with European unions in the metal sector. Therefore the union has not been
informed bythe management or by the GUF. That was also the case atMetalCorp in India, where local
union representatives told our research team that they were unaware of the existence of the GFA. To
what extent local management at MetalCorp India was informed of the GFA remained unclear. When
the question was raised, management was at first hesitant to respond, but then claimed to be following
MetalCorp’s code of conduct, including the labour provisions referencing the GFA, to the letter.
As for a joint labour-management information and communication policy, both the MetalCorp
and the SecureCorp case match our findings for all of our cases. Typically, the first press releases
and announcements of the GFA signing were published simultaneously, but not jointly. SecureCorp,
and UNI as the signatory GUF, did differ from procedures in the other cases as they agreed upon a
series of joint local meetings for the initial target countries (South Africa, India), but not, for exam-
ple, for the USA where the campaign for recognition and for a GFA had been so acrimonious. At
MetalCorp, there is no indication that this sort of joint communicating was ever under consideration.
Implementation II: Training practices
In both cases, we found no evidence of explicit training initiatives or comprehensive programmes
jointly developed by management and the unions and targeting GFA-related activities. But there
were vestiges of unilateral training activities in the MetalCorp case: in Turkey by management and
in Brazil by the local union (see above); and in the SecureCorp case on the union side (India). In
Turkey, the HR manager of MetalCorp reported that he continuously emphasizes the significance
of the GFA document and makes it an indispensable part of all the training sessions he holds with
blue- and white-collar employees alike. In addition, subsidiary management runs training pro-
grammes for its local suppliers to raise awareness regarding the labour standards that are expected
by MetalCorp to be applied. In the case of SecureCorp, we found that UNI was working with local
unions in India to build the necessary organizational capacity. In agreement with SecureCorp head-
quarters, union representatives were then trained in collective bargaining with the intention of
developing dialogue with SecureCorp local management. However, at the time (2010), local man-
agement was proving to be uncooperative, which led to tensions between UNI and SecureCorp
headquarters as well as between UNI and local union representatives in India, who expected UNI
to be able to resolve the issue at the headquarters level.
Implementation III: Operational practices
Both MetalCorp and SecureCorp have created special bodies for CSR at their headquarters, des-
ignated as ‘CSR board’ and ‘CSR committee’, respectively. Both companies also have extensive
498 Transfer 20(4)
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

internal auditing programmes for subsidiaries. At MetalCorp, this also applies to suppliers, who are
in addition bound by MetalCorp’s purchasing contracts to respect the labour standards of the
MetalCorp GFA. In Turkey, for instance, when allegations of basic employment rights violations
by local suppliers are reported, the subsidiary HR Director has to investigate them and report the
findings to MetalCorp headquarters (Fichter et al., 2013).
A major issue in nearly all of the implementation cases we investigated was the recognition of
trade unions and the pursuit of collective bargaining. All GFAs make reference to the ILO core
labour standards which include freedom of association and the right to collective bargaining. But
in those subsidiaries without recognized union representation, we generally found that local man-
agement had taken no steps, neither strategic nor operational, to revise organizational procedures
according to these provisions of their GFA. In the US, for example, unions disputed MetalCorp’s
claim that its local management was following a clear policy of neutrality in the face of union orga-
nizing efforts (Fichter and Stevis, 2013). In Turkey, the presence of a union not affiliated to Indus-
triALL was detrimental to joint implementation at MetalCorp, leaving the initiative to
management. On the other hand, we found two related cases at suppliers in Turkey where the
MetalCorp GFA played a role because of its incorporation into the company’s master supplier con-
tract and the disregard of these provisions by these suppliers (Fichter et al., 2013).
On the other hand, SecureCorp headquarters management has recognized UNI’s commitment to
increasing unionization across countries. Together with UNI, it has defined target countries as test-
ing grounds for developing and implementing adequate organizational policies in this direction. As
reported above, this policy has been pursued in India, and to some extent in other countries as well
where UNI and its local affiliates have been proactively invoking the GFA (McCallum, 2013). The
rationale for SecureCorp management is that it can benefit from supporting UNI in its efforts to
enable local workers to organize a union. Once the unions can show that they are representative
of the workforce, they can expect to get recognition from local management to conclude collective
agreements. However, at present, in the majority of countries this is not the case.
Finally, in regard to conflict resolution and the operational handling of alleged violations of the
GFA, MetalCorp has an internal investigative unit and has established a standing committee that is
attached to the corporate global council of employee representatives. In this constellation, discus-
sions on resolving disputes and correcting violations take place between employee representatives
and management on an institutionalized basis. But the overall approach is reactive, i.e. it is based
on dealing with individually reported cases instead of proactively developing policies to prevent
violations. A more proactive approach could however be in the making in conjunction with the
inclusion of GFA labour standards in MetalCorp’s master purchasing contracts.
In contrast, SecureCorp and UNI have defined a joint implementation strategy that has focused
on two countries (India, South Africa) as testing grounds. However, this process is only just begin-
ning. Individual reports of violations are not regularly discussed within a formal labour-
management body – except for the biannual meetings of the GFA review committee – since the
focus of joint roll-out activities is primarily on finding ways further to increase local unionization.
Discussion: towards a systematic approach to GFA implementation
Let us recapitulate the lessons from our cases and identify the emergent practices and evident gaps
that have led us to our proposal of implementation guidelines. In the case of MetalCorp, informa-
tion and communication practices were short of being comprehensive and organizationally
embedded; their effectiveness was determined more strongly by local actors. By contrast, the prac-
tices of this stage at SecureCorp were more institutionalized and priority-based in line with the
Sydow et al. 499
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

jointly agreed roll-out strategy. With respect to training practices, the case of SecureCorp suggests
that they may be more effective and better managed when based on a prioritizing of the countries,
in this case India, in which the agreement is to be implemented. This would enable implementation
techniques and processes to be tested and evaluated, and would focus resources more efficiently.
Similar arguments could also be made for setting up a priority schedule based on product divisions
or individual plants. In the case of MetalCorp, there was no strategic approach to developing this
stage of implementation as the focus in this TNC was remedial. Only in the local Brazilian union,
whose leader had been a representative of labour in the negotiation of the GFA, did we find a con-
certed effort to train union officials at MetalCorp in the use of the GFA.
The MetalCorp GFA is a good example of both the deficits and the successes in realizing the
operational practices of implementation. While the organizational recognition of trade unions and
trade union rights – although explicitly delineated in the GFA – proved to be attainable only by
pressure and conflict at the local level, MetalCorp headquarters readily incorporated GFA stan-
dards into its master supplier contracts, thus recognizing its social responsibility for labour rights
beyond its own organizational boundaries. The changes in operational practices brought about by
the SecureCorp GFA are possibly even more far-reaching, inasmuch as this TNC had previously
condoned virulent anti-union management policies in its operations outside of its home country
base. The practices, especially the joint activities in rolling out the agreement, that have developed
out of the negotiated provisions, are further indications of wide-ranging and long-term organiza-
tional changes.
These cases of limited and uneven implementation, combined with lessons from our other cases,
have helped us conceive of a ‘full scale’ implementation covering the complete TNC and its key
suppliers. We would argue that for a ‘full scale’ implementation of a GFA, apart from a strong
content, all three sets of practices involving all relevant actors are necessary. The first set concerns
the information dissemination and communication practices. The second set is marked by training
practices while the third is operational, referring to the introduction of routines, rules and proce-
dures as well as related organizational and inter-organizational structures (e.g. an inter-
organizational team in charge of coordinating the monitoring process). Furthermore, in line with
the multi-organizational practice perspective, implementation should be conceived as a process of
combined and joint activity and decision-making by management and labour at all levels and
stages of the TNC and its global production network. To illustrate possible activities towards rea-
lizing these three steps of implementation, we have proposed specific policies for each set of prac-
tices as examples that reflect the particular roles of the representatives of ‘capital’ and ‘labour’,
either separately or jointly (Table 1 above). Changes internal to the organization (e.g. setting up
or increasing the size of a respective department, overcoming cross-departmental communication
barriers) may accompany respective changes in inter-organizational practices – in firms as much as
in unions.
Apart from the importance of company-specific adaptation of implementation guidelines we
need to take into account the process character of implementation and the fact that this process
spans many national borders and systems of labour relations (Bamber et al., 2011). Regarding its
process character, we argue that, in effect, implementation commences already during the initia-
tion of negotiations, even if its ultimate nature has yet to be defined (Helfen and Sydow, 2013). As
for its multinational dimension, local actors, local rules and local practices are determining factors,
representing different situations in which implementation is meant to take place. These two con-
ditions clearly interact in our case studies, as can be illustrated by the ones we discuss in this
article. To give an example, in India the system of industrial relations is highly fragmented;
national unions exist alongside plant unions that are often employer-dependent. Moreover, formal
500 Transfer 20(4)
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

employment relations are only a small portion of the labour market as a whole, which is dominated
by the informal economy. In such a situation, the implementation of a GFA requires different
instruments and a different approach from that of a country such as the US in which internal plant
Ersatz-unions are banned and the majority of employment relationships are in some way legally
defined and regulated – but nevertheless reflective of local practices (Fichter and Stevis, 2013).
Concluding remarks
The number of Global Framework Agreements is continually growing; in the past 20 years over
100 GFAs have been negotiated and signed. While this may be taken as a signal of success, as our
research has shown, GFAs are still ‘work in progress’. And the modest successes in implementa-
tion we have presented are not even guaranteed for long; they require continual input and support
to mature and institutionalize. For GFAs to be truly effective and contribute to establishing a basis
for comprehensive labour standards, mounted and concentrated efforts by both unions and employ-
ers are required. GFAs are an instrument for facilitating the establishment of working relationships
across globalized production networks towards raising standards and addressing problems that
result in violations. For employers it seems that the greatest challenge is in enforcing recognition
throughout their operations both of GFAs internally as corporate policy and of unions as legitimate
representatives of employees. Management’s failings – and unwillingness – on this account is cer-
tainly a major obstacle facing unions in strengthening GFAs. At the same time, GFAs present an
opportunity as arenas for unions to develop the organizational presence and transnational links that
are essential to effective interest representation and dialogue with management. The framework of
transnational labour relations has been erected, but since GFAs are clearly a policy instrument of
unions, its relevance and effectiveness will depend primarily on how unions are able to develop it
further. Finding the right approach to negotiations with a TNC, bringing relevant union affiliates
on board (ownership), creating viable transnational union networks to ensure implementation, and
linking GFA policy with other policies around TNCs (see Fairbrother et al., 2013) will be neces-
sary towards that end.
Acknowledgements and Funding
This work was supported by the Hans-Bo¨ ckler-Foundation [grant number S-2008-141-2 B]. Addi-
tional support for country workshops and for the publication of our research results was provided
by the Friedrich-Ebert-Foundation [no grant number].
References
Amoore L (2002) Globalisation contested. An international political economy of work. Manchester: Manche-
ster University Press.
Arruda L, Fichter M, Helfen M and Sydow J (2012) International Framework Agreements – a powerful tool
for ensuring Core Labor Standards in a globalized world? Insights from Brazil. Ana´lises e propostas Sao
Paulo / Berlin Friedrich-Ebert-Foundation.
Bamber GJ, Lansbury RD and Wailes N (2011) International and comparative employment relations. 5th ed.
London: Sage.
Barley SR and Tolbert PS (1997) Institutionalization and structuration: Studying the links between action and
institution. Organization Studies 18(1): 93–117.
Burns T (1961) Micropolitics: Mechanism of institutional change. Administrative Science Quarterly 6(3):
257–281.
Sydow et al. 501
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

Coe NM, Dicken P and Hess M (2008) Global production networks: realizing the potential. Journal of
Economic Geography 8(3): 271–295.
Elk M (2014) After Historic UAW Defeat at Tennessee Volkswagen Plant, Theories Abound. In These Times,
15 February 2014. http://inthesetimes.com/working/entry/16300/after_uaw_defeat_at_volkswagen_in_
tennessee_theories_abound
Fairbrother P, Levesque C and Hennebert M-A (ed.) (2013) Transnational trade unionism. Building union
power. Routledge studies in employment and work relations in context 8. Routledge: New York.
Fichter M and Helfen M (2011) Going local with global policies: Implementing international framework
agreements in Brazil and the United States. In: Papadakis K (ed.) Shaping Global Industrial Relations.
The Impact of International Framework Agreements. Houndsmills: Palgrave Macmillan, pp.73–97.
Fichter M and Stevis D (2013) Global framework agreements in a union-hostile environment: the case of the
USA. Berlin: Friedrich-Ebert-Stiftung. Available at: http://library.fes.de/pdf-files/id/10377.pdf.
Fichter M, Helfen M and Sydow J (2011) Employment Relations in Global Production Networks – Initiating
Transfer of Practices via Union Involvement. Human Relations 64(4): 599–624.
Fichter M, Sayim KZ and Agtas O
¨B (2013) Oganization and Regulation of Employment Relations in Trans-
national Production and Supply Networks. Ensuring Core Labor Standards through International Frame-
work Agreements? Ankara: Friedrich-Ebert-Foundation. Available at: http://www.fes-tuerkei.org/media/
pdf/Partnerpublikationen/layouts%20web%20english.pdf.
Fichter M, Stevis D and Helfen M (2012a) Bargaining for corporate responsibility: The global and the local of
framework agreements in the USA. Business and Politics 14(3): 1–31.
Fichter M, Sydow J, Helfen M, Arruda L, Agtas O
¨B, Gartenberg I, McCallum JK, Sayim KZ and Stevis D
(2012b) Globalising Labour Relations. On Track with Framework Agreements? Berlin: Friedrich-
Ebert-Foundation. Available at: http://library.fes.de/pdf-files/iez/09422-20121129.pdf.
Gallin D (2008) International Framework Agreements: A Reassessment. In: Papadakis K (ed.) Cross-Border
Social Dialogue and Agreements: An Emerging Global Industrial Relations Framework? Geneva: Inter-
national Institute for Labour Studies, pp.15–41.
Giddens A (1984) The constitution of society. Cambridge: Polity Press.
Hammer N (2005) International Framework Agreements: global industrial relations between rights and bar-
gaining. Transfer 11(04): 511–530.
Hedlund G (1986) The Hypermodern MNC – A Heterarchy? Human Resource Management 25(1): 9–35.
Helfen M and Fichter M (2013) Building transnational union networks across global production networks:
Conceptualising a new arena of labour-management relations. British Journal of Industrial Relations
51(3): 553–576.
Helfen M and Sydow J (2013) Negotiating as institutional work: The case of labor standards and International
Framework Agreements. Organization Studies 34(8): 1073–1098.
Krause DR, Handfield RB and Scannel TV (1998) An empirical investigation of supplier development: reac-
tive and strategic processes. Journal of Operations Management 17(1): 39–58.
Leonardi S (ed.) (2012) European Action on Transnational Company Agreements: a stepping stone towards a
real internationalisation of industrial relations? Rome: Istituto di Ricerche Economiche e Sociali – IRES.
Levy D (2008) Political Contestation in Global Production Networks. Academy of Management Review 33(4):
943–963.
McCallum JK (2013) Global Unions, Local Power: The New Spirit of Transnational Labor Organizing.
Ithaca: ILR Press.
Papadakis K (ed.) (2011) Shaping Global Industrial Relations. The Impact of International Framework
Agreements. Geneva: International Labour Organization / Palgrave Macmillan.
Ru
¨b S, Platzer H-W and Mu
¨ller T (2011) Transnationale Unternehmensvereinbarungen. Berlin: edition
sigma.
502 Transfer 20(4)
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

Stevis D (2009) International Framework Agreement and Global Social Dialogue: Lessons from the Daimler
Case. International Labour Office, Employment Sector, Employment Working Paper No. 46.
Stevis D (2010) International Framework Agreements and Global Social Dialogue: Parameters and Prospects.
International Labour Office, Employment Sector, Employment Working Paper No. 47.
Stevis D and Boswell T (2007) International Framework Agreements: Opportunities and Challenges for Global
Unionism. In: Bronfenbrenner K (ed.) Global Unions: Challenging Global Capital through Cross-border
Campaigns. Ithaca, NY: Cornell University Press, pp.174–194.
Sydow J and Windeler A (1998) Organizing and evaluating interfirm networks – A structurationist perspective
on network management and effectiveness. Organization Science 9(3): 265–284.
Telljohann V, Da Costa I, Mu
¨ller T, Rehfeld U and Zimmer R (2009) European and international framework
agreements: Practical experiences and strategic approaches. Luxembourg: European Foundation Dublin.
Available at: http://www.eurofound.europa.eu/publications/htmlfiles/ef08102.htm (29 June 2009).
Sydow et al. 503
at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from

at Freie Universitaet Berlin on November 7, 2014trs.sagepub.comDownloaded from