Worker-driven monitoring – Redefining supply chain monitoring to improve labour rights in global supply chains

Abstract

In this article, we advance on ongoing debates about how global supply chain monitoring can work to improve conditions for workers in those supply chains. We propose an approach to worker-driven monitoring and introduce a set of core elements that should form the basis of this new approach to monitoring labour conditions in global supply chains. We argue that supply chain workers’ participation in labour monitoring systems cannot be confined to the moment of factory inspection. Instead it must be understood as part of a larger process which incorporates workers from the point of designing the systems that will be monitored through to the point of remediation. We conclude that such a systematic and comprehensive approach is necessary to ensure that monitoring processes are meaningful and place the rights of workers at their core.

Keywords

Business and human rights Corporate Social Responsibility global supply chains labour rights supply chain monitoring worker-driven monitoring

Introduction

Attempts to address human rights abuses and poor labour conditions in global supply chains (GSCs) date back several decades. The current framework on business and human rights – formulated in the United Nations (UN) Human Rights Council, 2008 and articulated in the UN Guiding Principles on Business and Human Rights (United Nations, 2011) – deals with this ongoing challenge by distributing the responsibilities between states and corporations. States have the obligation to protect the human rights of those under their jurisdiction while corporations have the responsibility to respect human rights in their commercial activities and partnerships, including through their supply chains.¹ Further attention has subsequently being paid to working conditions in a new wave of policy instruments addressing responsible business conduct and human rights due diligence in GSCs such as the OECD’s Due Diligence Guidance for Responsible Business Conduct (2018).

Despite these developments, and in the face of continued problems with conditions for workers, the monitoring of supply chains for labour rights violations has become an industry in itself (Lund-Thomsen and Lindgreen, 2014). However, the monitoring industry has been accused of ignoring conflicts of interest and using inappropriate methodologies to detect worker rights violations at the base of GSCs while not creating real improvements in the rights and conditions of those labouring within such chains (Bartley and Egels-Zandén, 2015; Ruwanpura, 2016). In order to bring about required improvements in labour conditions in GSCs, academics and labour organizations have identified workers’ voices as a crucial missing component in supply chain monitoring. However, less attention has been paid to how workers’ voices could be systematically incorporated as part of this monitoring in practice (Claeson, 2019; Hale and Opondo, 2005; Locke and Romis, 2007).

Our article is not the first that seeks to investigate the need and means for workers to be given greater influence in the monitoring of labour conditions in GSCs. For instance, Prieto-Carron (2004) argued that women workers have legitimate voices which must be heard if corporate codes of conduct are to be designed and implemented in ways that are beneficial to workers at the base of GSCs. In the last decade, a new literature has also appeared on the role of labour agency in GSCs. Here labour agency is defined as (a) workers either organizing through trade unions to collectively voice their concerns and defend their interests or as (b) workers opting in and out of different workplaces and jobs based on their personal interests and broader life circumstances (Alford et al., 2017; Coe and Hess, 2013). For example, De Neve (2014) suggests that Indian workers sometimes opt out of working for code-compliant factories in the country’s garment industry as they perceive supply chain monitoring of labour conditions as attempts at controlling their freedom and dignity at work. In short, these writings place the role of workers within GSCs at centre stage in relation to the design and implementation of labour monitoring systems in GSCs. This work does not however require systematic changes to the way that monitoring systems are designed, implemented and enforced.

This article seeks to move forward the debate on worker-centred monitoring of labour conditions in GSCs. We propose a set of core elements which we argue can form the basis of a new approach to monitoring, one that would fundamentally shift the process and outcomes in favour of supply chain workers. This model extends the idea of incorporating workers’ voices beyond a model that is worker-centred to one which is worker-driven. We argue that the participation of supply chain workers cannot be confined to the moment of factory inspection but rather has to be viewed as part of a larger process which incorporates workers from the point of designing the systems that will be monitored through to the point of remediation. This more systematic and comprehensive approach is necessary to ensure that inspections are meaningful and achieve a result that has the rights of the worker at its core.

The article is structured as follows. First, we outline the background to the rise of supply chain monitoring instruments that seek to record and rectify labour conditions and human rights violations in GSCs. In the next section, we move onto reviewing some of many shortcomings of current supply chain monitoring models. The third part of the article introduces a new model for incorporating worker voices in the design, implementation and monitoring of labour conditions in GSCs. Finally, the conclusion summarizes the main findings of the article and discusses the research and policy implications of our analysis.

Monitoring in the context of GSC governance

The operation of GSCs has frequently been linked with abuses of human rights and violations of labour standards. In the 1990s in particular, a light was shone on human rights abuses occurring in the factories and places of work forming part of the supply chain for well-known international corporations (Klein, 2000). The global apparel corporation Nike serves as a highly publicized example of a company which became publicly linked with claims and investigations revealing low wages, poor working conditions and human rights abuses (Locke, 2002). The problems identified – and reported and discussed widely both within and outside of academia – frequently occurred in the lower tiers of production; in factories located several production stages away from the lead firm responsible for the goods and from the end-user of those goods.

Adopting effective responses to these problems has been complicated by the geographical and contractual nature of modern supply chains in which there is a much discussed ‘governance gap’ (Kobrin, 2008; Ruggie, 2008). This gap arises from the fact that the relevant multinational enterprises (MNEs which here refer specifically to the lead firms in the supply chain) operate outside of traditional state-based regulatory systems. In many instances the tiers of manufacturing and assembly take place outside of the home-state of the lead firm – and thus the laws there are not applicable – but instead take place in developing or low income countries in which regulation is likely to be relatively less stringent and enforcement weak (Distelhorst et al., 2015).

During the 1990s, a plethora of corporate self-regulation instruments were developed in the framework of Corporate Social Responsibility (CSR) which had gained renewed prominence in response to changes in global business structures described above and the increased pressure on companies to assume responsibility for the impacts of their global business activities (Outhwaite and Martin-Ortega, 2016). Into this context, alternative regulatory approaches proliferated as regulatory actors including civil society, business enterprises and governments attempted to catch-up with changes to the way that companies operate and are regulated and to ensure that human rights (as well other social and environmental matters) are respected throughout the supply chain for a given product or company. In particular, the increase in CSR was expressed through participation in initiatives such as corporate codes of conduct, industry codes of conduct and multi-stakeholder codes and initiatives at domestic, industry and global levels which aimed to establish expectations, guidelines or commitments to respecting the rights of workers within GSCs (Rodríguez-Garavito, 2005).

Monitoring supply chains: A legacy of failure?

Monitoring supply chains – notably through ‘social auditing’ models – has now become an industry in itself, accompanied by complex procedures, accreditations and methodologies. However, an increase in the variety, complexity (and cost) of monitoring has not translated into increased effectiveness in identifying, preventing and remedying violations and inadequate working conditions in supply chains, as will be discussed in this section (Barkley and Egels-Zandén, 2015; Lund-Thomsen and Lindgreen, 2014; Locke et al., 2013).

Traditional monitoring models may seek to enforce public or private standards. They may be led or undertaken by a range of actors including MNEs (the lead firm themselves), in the case of corporate codes of conduct, by third party monitors, including a range of auditing firms, bodies and consultants and through multi-stakeholder monitoring initiatives which incorporate NGOs or ‘independent’ bodies (O’Rourke, 2003).

A range of monitoring models exist which we refer to here as ‘traditional models’. Whilst recognizing that the terminology is not always consistent in the literature, we include among these: ‘in-house’ monitoring; third party auditing; multi-stakeholder initiatives and independent monitoring. These can be distinguished from worker-driven monitoring, which we present in the ‘Conclusions’ section as an innovative next step in supply chain monitoring and reform.

Each of the traditional models presented here contains important limitations in terms of capability to identify, prevent and remedy human rights and labour rights violations and protect worker health and safety. In addition, it is important to understand that none of the models directly incorporates the structural issues – those associated with the cost and time-frame for delivery of the product to the lead firm – that have been associated with health and safety breaches, poor working conditions and human rights violations (ILO, 2017). The particular features and limitations of each of these models are explored in more detail below.

‘In-house’ monitoring involves the lead firm setting expectations or standards for its suppliers and making its own arrangements for monitoring them. This model is problematic in terms of identifying and addressing abuses or failures because both the lead firm and the supplier have a business interest in hiding violations – or at least minimizing their visibility – rather than reporting them (Locke et al., 2007). In addition, as with paid-for third party auditing (below), monitoring undertaken in-house by companies or industry associations may be perceived as unreliable: perhaps because of the conflict of interest, the methods used to undertake in-house monitoring activities have been criticized as inappropriate and inadequate with respect to workers (Outhwaite and Martin-Ortega, 2017).

Tensions or direct conflict arise between the aims of monitoring and the realistic outcomes of that process given that remedying problems found through an audit process will often have financial implications, which in turn will affect the profit margins of the brand (Outhwaite and Martin-Ortega, 2017). Structural issues such as sourcing practices, including pressures on price (which may lead to low wages and excessive voluntary overtime) and on lead-times (which may be a cause of problems such as forced overtime), have traditionally not been incorporated into the remit of these (and other) monitoring process or may act as a barrier to change even if they are incorporated (Lund-Thomsen and Lindgreen, 2014). Moreover, this process can create an internal tension for businesses since effective monitoring mechanisms imply costs to maintain as well as giving rise to further costs where remedial action is required (Claeson, 2019; Outhwaite and Martin-Ortega, 2017).

Third party auditing introduces an additional party into the monitoring process, with an external auditing firm or organization carrying out a compliance audit on behalf of a company, industry group or, potentially, a public body or other organization. This serves to create regulatory distance (at least in theory) between the lead firm and the outcome of the auditing process. Third party auditing has become a very widely used model to verify compliance, or identify non- compliance, with codes of conduct or other agreed standards in a supply chain site or factory. Although it is perhaps the most prevalent verification model, ‘social auditing’ of this kind has started to be viewed negatively, particularly in light of high-profile auditing failures (Pruett, 2005). One of the most notorious of these was the successful audit of the Rana Plaza factory in Bangladesh soon before it collapsed killing over one thousand garment factory workers as mentioned above. This incident highlighted one of the most frequently cited failings of social auditing; the ‘tick box’ and ‘snap shot’ nature of the process which provides only a limited picture of conditions at a particular factory.

The practices of auditing firms can be criticized as inadequate for identifying genuine problems – factory visits may be too infrequent and too limited in scope to provide a realistic evaluation of the situation (LeBaron and Lister, 2016). The process of auditing can itself have detrimental outcomes for workers; it has been observed that auditing can result in problems simply being shifted further along the supply chain, to the next tier that is outside the reach of the audit. For those workers who are involved in the audit, inappropriate methodologies such as interviewing workers onsite can leave them vulnerable to repercussions from management personnel and will affect audit outcomes since these workers may fear adverse consequences from speaking honestly (O’Rourke, 2003).

Similar to in-house monitoring, third party monitoring entails a perceived, if not actual, conflict of interest in that auditors report to lead firms, from whom they hope to obtain future work. Since auditors have an interest in maintaining the business of lead firms, questions have been raised as to whether the outputs of social audits can be considered reliable. Social auditing has been criticised for being a public relations vehicle for the corporation in question, rather than a monitoring exercise which has worker protection as its objective (Donaghey and Reinecke, 2017; Locke et al., 2007; O’Rourke, 2003; Short et al., 2016).

The consistently identified breaches and failings within factories that are subject to and frequently have “passed” third-party social audits have led many commentators to condemn this model as not fit for purpose (AFL-CIO, 2013; Brown, 2009). Even ‘ethical audit’ regimes have not avoided these criticisms “LeBaron and Lister’s work finds many of the same problems surrounding asymmetries of power, inappropriate working methods and structural limitations in the context of audit regimes which ultimately support ethical social and environmental efforts such as fair trade (LeBaron and Lister, 2016)”.

From the late 1990s, a number of multi-stakeholder and other collaborative voluntary supply chain initiatives emerged with a central focus on monitoring and as a direct response to the highly publicized supply chain failings of that decade (LeBaron et al., 2017). In contrast with private social auditing, these tend to be broader supply chain initiatives which incorporate monitoring as a component of their efforts to improve supply chain conditions. The Fair Labor Association (FLA) for instance adopts a model of setting voluntary standards, through a code of conduct, undertaking monitoring and inspection, and reporting on compliance to participating companies (FLA, 2019). Such initiatives have attempted to create more regulatory space between supply chain conditions and the influence of the lead firm by shifting the firm‘s dominance and involvement in setting the standards and commissioning the inspection and verification processes which report on compliance with them.

Such initiatives were often established as, or led by, NGOs. They seek to incorporate a range of perspectives within their operational and/or governance structure but may vary in their degree of collaboration with corporations. These initiatives have not however entirely evaded the criticisms of third party monitoring extended to private sector auditing firms (see O’Rourke, 2003b). For example, the monitoring and verification operations of initiatives such as the FLA, SA8000 (2018) and WRAP (2019) (Worldwide Responsible Apparel Production) have been criticized for their lack of transparency and because they are sometimes viewed as being dominated by business interests. The FLA for instance is a multi-stakeholder initiative. The FLA's Board of Directors comprises an independent chair and proportionally equal representation from companies, NGO representatives and University representatives.

A separate tension is also present in multi-stakeholder initiatives because of the range of positions represented and the consequent need for – but often absence of – appropriate mechanisms which can be triggered when the tension between competing interests cannot be resolved (Outhwaite and Martin-Ortega, 2017). Corporate funding for monitoring activities has been identified as problematic for the same reasons (Brown, 2013).

In light of the criticisms of the multi-stakeholder models above, there has been an increased interest in ‘truly independent’ monitoring processes which seek to avoid both the conflicts of interest and the potentially detrimental and limited verification processes that have been the subject of criticism. In these initiatives, the lead brands or other private sector interests are not represented in the governance or operational structures of the initiatives. Such independent monitoring which, has been described by O’Rourke as including unions and private individuals monitoring without receiving a fee. These may be seen as ‘more accountable and trustworthy’ (Outhwaite and Martin-Ortega, 2017) contrasting with company and industry initiatives. One of the most established and positively viewed examples is the Worker Rights Consortium (WRC) which is referred to in greater detail in the section ‘Towards worker-driven monitoring’.²

The fact of independence, conversely, is not the fact of independence in itself is not a guarantee of good practices and therefore independent monitoring bodies can be open to criticism if their investigations and the results of their activities are not public or transparent (O’Rourke, 2003). Independent processes have also been subject to concerns that NGOs will be ‘co-opted’ (Outhwaite and Martin-Ortega, 2017) by corporations or about the quality of auditing (O’Rourke, 2003) or that the impacts of a confrontational model may be felt more in terms of negative publicity for the brand than on a clear means of resolving problems (Outhwaite and Martin-Ortega, 2017).

A limitation of a different nature concerns funding: independent organizations seeking to incorporate monitoring as part of an initiative to raise supply chain standards may lack the necessary funding to carry out monitoring activities, even where this is viewed as a potentially important aspect of their work (Outhwaite and Martin-Ortega, 2017). There may be a tension between the need for adequate funding to undertake investigations and the ability and willingness of members or affiliates to pay fees (or in the application of other funding structures) and this will have consequences for the scope of work that can be undertaken (Outhwaite and Martin-Ortega, 2017). A key issue for ‘truly independent’ monitoring organizations may then lie not in the concerns of conflicts and reliability but in guaranteeing and demonstrating the embodiment of good practices and in developing sustainable models that can achieve the desired outcomes.

Across the traditional monitoring models presented are problems of transparency, legitimacy, accountability and effectiveness. These models also are unable to challenge the basic terms of engagement and ways of organizing production that form the top-down pressures exerted in the chain (Faust, 2016). Ultimately, analysis of monitoring approaches cannot escape the conclusion which began this section: that despite some decades of effort there is little evidence that conditions in GSCs have in fact improved for workers and monitoring efforts are therefore not bringing about the intended change: examples can easily be found from the electronics, garment and food and fisheries (Human Rights Watch, 2017). Both the processes and the outcomes are failing to meet the needs of the workers in whose interests they were adopted (although systematic and detailed analysis of the effects of codes is scarce, as noted by Hale and Opondo (2005). At the same time however, it has been recognized that ‘despite the shortcomings and legitimate criticism of factory monitoring, for workers in factories, monitoring and auditing are part of their daily existence and will probably continue to be at least for the medium term’ (ILRF, 2016).

What is needed then is a way forward which redefines and reinvigorates the nature and role of factory monitoring in a way which enables it to truly meet the needs of supply chain workers.

Workers: The missing voice in supply chain monitoring?

To date, a definitive response as to how to make supply chain monitoring effective has been elusive. But in seeking to progress monitoring efforts, a focus has been emerging from both labour rights organizations and the academic community on the position and involvement of workers themselves in that process.

In the context of the Kenyan cut-flower industry, Hale and Opondo (2005) note that while a plethora of codes has developed, the majority of workers ‘appear[ed] to be constructed as passive subjects’ within the applicable code. Rodriguez-Garavito (2005) similarly notes that the voice of the workers, particularly in developing countries, is a missing component of the process of setting and implementing codes of conduct and their monitoring systems.³ Claeson (2019) demonstrates how social auditing in the electronics industry silences the workers’ voice, suggesting that it may in fact harm their chances for remediation. Within a wider discussion of participatory governance (see Fung and Olin Wright, 2001), the key theoretical and practical task for Rodriguez-Garavito, as regards international labour standards, is thinking of ‘ways to move from elite, firm-centric governance (as currently embodied by most monitoring systems) to an alternative approach to codes as represented by some nascent experiences in independent monitoring, worker empowerment and cross-border labor organizing…’ (2005). Locke and Romis further suggest, based on research into management and monitoring practices at two Nike plants in Mexico, that ‘interventions aimed at improving production, reorganizing work, and empowering workers on the shop floor complement codes and offer a means of further improving working conditions’ (2007). Against the numerous and ongoing failures of traditional models, and the ‘near total lack’ of worker involvement in factory-level health and safety programs, Brown (2013) notes that ‘global supply chains will become safer only with a sea-change in the international business model and the active participation of informed and empowered workers.’⁴

In parallel, efforts are being made to develop and advocate for monitoring practices in which the role and voice of the worker is central. In its activities focusing on workers in the electronics industry in Mexico, CEREAL highlighted discrepancies between the outcome of official audits of corporate codes of conduct and the personal accounts of workers, often a result of the limitations inherent in the models as described above (CEREAL, 2016). CEREAL promotes instead the notion of ‘worker-led monitoring’, as an alternative that ‘complements and contrasts the optics of audits’ and in which monitoring projects focus on

empowering workers to interview other workers in secure facilities outside the factory, preserving their anonymity – and therefore their safety from reprisals. These monitoring activities are usually presented in the context of a broader global strategy for the protection and promotion of workers’ human rights that includes training, organizing, campaigning, legal counselling and litigation. (Foust Rodriguez, 2016)

Electronics Watch is an independent monitoring organization which supports public sector organizations to work together to ensure respect for labour rights and safety standards in factories that make the ICT hardware that those organizations purchase.⁵ Electronics Watch terms its approach to monitoring as ‘worker-driven.’ Its approach articulates factory and worker access through a network of local organizations which are trusted by workers, have direct knowledge of working conditions and factory dynamics and serve as the ‘eyes and the ears on the ground’. Elsewhere, discussing as a concept ‘worker-driven social responsibility’, the WSR Network (2017) recognizes that workers have an abiding interest in ensuring that their rights are protected and are best placed to know about violations and should therefore be at ‘the head of the table’ in monitoring efforts.

These perspectives articulate a vital concern in the evolution of GSC monitoring: that the worker’s voice must be heard and that worker participation is essential. These efforts reframe monitoring as worker-centred; they attempt to centralize the experience and perspective of the worker and call for worker empowerment. Reframing monitoring in this way marks an important shift in its central aims and outcomes to make it better aligned with the needs of workers in GSCs. This is a compelling but as yet ill-defined vision. At the same time, in part because this vision has not yet been subject to rigorous debate and development, it runs the risk of offering change which is too limited to make the necessary difference to working conditions. In the next section we explore why this is and argue that to truly tackle the problems with traditional approaches to monitoring new approaches to monitoring can go further still.

Monitoring models that are not only worker-centred but worker-driven offer possibilities to overcome some of the conflicts that arise in other approaches and to ensure that the voice and the active involvement of the worker is embedded throughout the process in a way that fundamentally shifts what the monitoring is trying to achieve and can indeed achieve.

Towards worker-driven monitoring

In this section, we draw on the small number of, mostly very recent initiatives to influence and change conditions within GSCs and which respond to the identified problems with traditional monitoring models. We examine key features of these initiatives, in conjunction with the criticisms and perspectives summarized in earlier sections to establish a worker-driven approach to monitoring GSCs. We present this in part (a) as a proposed set of core elements for ‘worker-driven monitoring’. A summary of these core elements is presented in Table 1. In part (b) we raise outstanding questions about the roles of and relationships between stakeholders in worker-driven monitoring that we suggest will need to be subject to further interrogation before worker-driven monitoring can be confident of overcoming – and not replicating – the problems seen in traditional monitoring models.

Table 1.

Summary of core elements for worker-driven monitoring.

Elements	Characteristics
Objectives	The overriding objective of the monitoring system is to improve conditions of workers in the supply chain.
Governance	With reference to the participation of firms or companies, governance structures avoid a real or perceived conflict of interest.Worker representation has meaningful implications and is not tokenistic.
Design	Workers are included in the design of initiatives to improve supply chain conditions. Specifically this includes determining the object of inspection.
Training and education	Training and education programmes are not mere opportunities for dissemination of policy but empowers workers and equips them with the knowledge and capacity to fully participate in the monitoring process
Inspection	Inspections are process rather than outcome driven.Inspection methods surpass the minimum needed to ensure that workers are not disadvantaged by engaging with them but actively engage or represent workers within the inspection process.
Complaints, remediation and dispute resolution	Violations can be identified by workers, rather than only through external auditing, while maintaining the safety and security of the worker. Remediation mechanisms ensure that workers are compensated where appropriate, that violations are remedied and that in achieving this, the employment position of the worker is not jeopardized in the short- or long-term.Conflicts between different interested parties can be mediated through effective dispute mechanisms for example through recourse to binding arbitration.

(a) Core elements for a new approach

Objectives

The rationale for worker-driven monitoring is that traditional models have not achieved their aims: poor working conditions and violations of workers’ rights are still commonplace in GSCs despite over two decades of effort. The presence of workers in the monitoring process has the potential to ‘interrupt’ the conflict of interest that occurs in in-house and third party monitoring, to address concerns about regulatory capture and to ensure that the practices of monitoring and remediation are not detrimental to workers and that they lead to reliable outcomes and conclusions. Only a system which enables a true assessment of likely risks and the real problems that occur can be relied upon to remedy violations, when they do occur, in a way that reflects the needs of the worker and not only the demands, capacity or willingness of the brands and suppliers.

On this basis, the objective of any initiative based on worker-driven monitoring must be to ensure that workers are meaningfully incorporated into the monitoring processes and that the outcome of monitoring processes both (i) reflect the lived reality of working conditions in the given setting and (ii) lead to necessary changes in conditions and do not have unwanted consequences in terms of the rights and conditions of other workers (for example in further tiers of production). In sum, the overriding objective of monitoring in the context of worker-driven monitoring is to improve conditions of workers in the supply chain and this is not to be subjugated by any other interests.

(ii) Governance

Another important part of worker-driven monitoring relates to the governance structures of monitoring organizations that determine which interests are represented at the highest level of the organization. Based on the discussion in section two, two issues are seen as important here: the representation or participation of workers and the representation or participation of the brands. Neither Electronics Watch nor the WRC include corporate representatives in their governance structures, addressing directly the conflicts of interest that can arise otherwise. The governance structure of WRC comprises representatives of the University Caucus, independent labour rights experts representing the WRC Advisory Council, and representatives of United Students Against Sweatshops. Electronics Watch is an affiliate organization governed by a Board of Trustees comprising labour rights and civil society representatives and academics. These organizations, since they are responsible directly for factory inspections avoid a real or perceived conflict of interest since they do not provide an opportunity for their actions to be formally influenced by corporate interests through governance structures. This may be important for avoiding conflicts of interest and so also the trust of the workers from whom information is collected and whose interests are ultimately represented.

An alternative approach is pursued in the Bangladesh Accord, which, as has been pointed out, incorporates worker representation in the form of trade union representatives, with its Steering Committee comprising equal representation from trade union and company representatives (at paragraph 4). In this way, the Accord seeks to avoid tokenism in the representation of workers while reflecting its core aims of ensuring that the applicable corporate brands play an active role in the initiative. As part of a broad sectoral initiative to demand active and systemic change from industry, the inclusion of the brands is accepted in that fora as strategically and logistically justified.

(iii) Designing the subject of monitoring

In worker-driven monitoring, workers should be included in the design of initiatives to improve supply chain conditions so that the subsequent inspections and verification steps, as well as remediation, are meaningful in terms of worker’s needs (Prieto-Carron, 2004). Without including workers at this stage the subsequent phases, particularly inspection (the focus of most discussion of monitoring) ultimately is developed and driven by other stakeholders with, inevitably, different perspectives. As well as incorporating workers at the stages of inspection, therefore, workers should also be involved in determining what the object of inspection is. Including workers in the design phase frames the subsequent steps in monitoring compliance in terms of the perspective of workers in that supply chain. Avoiding the top-down structures common in codes of conduct and similar instruments, some of the recent and progressive efforts have incorporated workers’ voices from the outset.

The Bangladesh Accord defines itself as ‘an independent, legally binding agreement between brands and trade unions designed to work towards a safe and healthy Bangladeshi Ready-Made Garment Industry’. It is built on what Donaghey and Reinecke (2017) refer to as principles of Industrial Democracy which puts supply chain workers at the ‘centre of the design and implementation of initiatives to improve their conditions’. With respect to incorporating workers’ interests, Global Union Federations were heavily involved in the negotiation of the Accord and in the resulting governance structure (Donaghey and Reinecke, 2017) which is reflected in the composition of its Steering Committee. The Committee contains equal representation of the signatory companies and trade unions with a neutral Chair provided by the ILO, and two witness signatories, from the WRC and the Clean Clothes Campaign (see Accord on Fire and Building Safety in Bangladesh, Governance, 2019). This led to an initiative that was more reflective of workers’ interests and needs. Unions, as members of the Accord Steering Committee, have raised controversial issues, including brand responsibility for financing costly compensation and remediation (Donaghey and Reinecke, 2017).

Similarly is the Indonesian Freedom of Association Protocol, an agreement signed in 2011 by Indonesian textile, clothing and footwear unions, major supplier factories and the major sportswear brands – including Adidas, Nike and Puma. Clean Clothes Campaign has described the initiative as ‘putting workers’ agency at the centre’ of efforts to protect sportswear workers’ rights (Clean Clothes Campaign, Freedom of Association Protocol – Indonesia, 2019). The Indonesian Protocol included ‘critical input’ from Indonesian women workers and union leaders during its development and saw union groups negotiate ‘clause by clause’ with major sportswear brands and large suppliers (Gardener, 2012) Worker representatives are noted to have been ‘active decision makers and participants in the design, dissemination, implementation and governance’ of the Protocol resulting in high levels of stakeholder ownership, a Protocol which is appropriately tailored to local conditions and to the ‘active use of the Protocol as a bargaining tool within individual workplaces’ (Delaney et al., 2016).

The Fair Food Program is a partnership among farmers, farmworkers, and retail food companies. It addresses working conditions in agricultural production – it, in its own words: ensures humane wages and working conditions for the workers who pick fruits and vegetables on participating farms.⁶ The Program is described as a human rights programme and, by the same token, as a direct contrast to CSR approaches. The program labels itself as worker-driven in large part because of the role of workers (especially the Coalition of Immokalee Workers) in the development and design of the program.

(iv) Training and education

Worker-driven monitoring requires workers who are empowered and have the capacity to fully participate in the monitoring process. An ability to participate in the governance, inspection or complaints processes is limited, undermined or stifled where workers are either unaware of these processes or have insufficient information, resources or support to engage with them.

Training is a further pillar of the Bangladesh Accord. As well as providing basic safety training, the training programmes function to ‘enable workers to voice concerns and actively participate in activities to ensure their own safety’ (Section 16 of the Accord). Training is an important element of the operationalization of the Accord, in which the participation of worker representatives required in the agreement to work closely with unions to overcome resistance from factory managers, as well as the willingness to address violence and threats to union members.

The Accord employed field workers who trained and worked proactively with union leaders to inform them about their rights under the Accord to participate in inspections and obtain copies of inspection reports, made available in the local language. In cases where management refused to let unions participate in inspections, informed workers were able to get Accord case handlers to intervene and ensure their participation (Donaghey and Reinecke, 2017) The Fair Food Program adopts worker education as one of its central pillars, incorporating worker-to-worker education and materials on workers’ rights with the specified outcome of enabling them to act as ‘trained and motivated monitors on the ground’ (Fair Food Program, 2019). Workers are paid for the time they spend undertaking training, addressing a potential barrier to participation.

(v) Inspection

Worker-driven monitoring also relates to how inspections are carried out. Inspections are the stage in which violations and instances of non-compliance with given laws, standards or codes are identified. This element is in some respects the core of monitoring and notions of inspection and monitoring are sometimes used interchangeably. Inspection in worker-driven monitoring should tend towards being process – rather than outcome – driven in order to avoid the problems that persist in traditional monitoring models. Effective worker-driven monitoring should similarly seek to adopt more appropriate practice and methods than those criticized previously in this article, ensuring that workers are not disadvantaged or endangered by the chosen methods of verification and inspection. Moreover, worker-driven monitoring can and should incorporate methods that actively engage or represent workers within the inspection process.

Electronics Watch inspections are designed around the understanding that

workers are the best monitors because they are on site all day, every day. They know the process and problems of normal operations; they have ideas for resolving safety and labour rights problems; and they can verify whether corrections are implemented and actually work. (Electronics Watch, 2019)

The Electronics Watch monitoring model is based around the use of civil society organizations or labour rights groups who are local to the region of inspection to undertake factory inspections. These organizations are to maintain close and ongoing relationships with the workers in the factories of inspection so that opportunities are created for workers to report problems that recur or are long term issues; such problems, as discussed, can be overlooked or hidden in ‘snapshot’ inspection processes. The model is in the early phases of operationalization but it aims to ensure that barriers to workers reporting problems are removed and that monitoring is undertaken by actors more closely aligned with workers’ interests rather than, as in some of the traditional models, with the interests of the factory managers or corporate brands. A similar but distinctive model, relying on an overarching body with separate monitoring provision, is adopted within the Fair Food Program.

Again with respect to the practical processes of inspection, the WRC has been identified on a number of occasions as the leading example of good practice. WRC monitoring activities include conducting interviews with workers in their homes and communities (rather than in the factories where they work); interviews with plant managers, local government labour authorities, relevant labour union officials, local academic and legal experts, and other relevant parties; visual inspections, where permitted by factory managers, and review of relevant records, documents and supplemental scholarly and/or journalistic materials. WRC has on this basis been viewed by some as the ‘gold standard’ because of its approach to engaging with workers directly (Outhwaite and Martin-Ortega, 2017).

The Bangladesh Accord uses the language of ‘credible inspections’ to distinguish the process from the social audit failings that preceded it and goes further than establishing a system for independent inspection. Inspections are to be carried out by qualified personnel under the direction of an independent Safety Inspectors. Inspection reports are to be shared with worker representatives.

The Accord has been considered as notable for ‘putting democratic worker voice at the core of safety processes’ (Donaghey and Reinecke, 2017). This is achieved through the incorporation of worker presence and intervention into the inspection processes via the participation and capacity building of union leaders. This view is based on the important, close linkage within the Accord between inspection and training, where training has the specific purpose to ‘enable workers to voice concerns and actively participate in activities to ensure their own safety’ (paragraph 16). Further, suppliers are required under the Accord to establish Health and Safety Committees in their supplier factories and these are to include no less than 50% worker representation, which Evans (2015) suggests, further gives them a voice. Particularly noteworthy here also is the provision that this representation will be based on democratic election where no union is present, so that absence of a union is not a barrier to worker participation in health and safety processes. As a practical implication of these arrangements, Donaghey and Reinecke found that ‘[f]or a local union president ‘to be seen walking through the factory with those inspectors’ was considered a powerful demonstration that workers were eligible to speak on the same level with outside authorities and managers.’ All reports on inspections, as well as Corrective Action Plans for all suppliers are publicly excessive on the Accord website, which provides a new level of transparency which according to Zimmer (2016) sets a new standard of supply chain disclosure.

(vi) Complaints, remediation, and dispute resolution

Any monitoring process described as worker-driven must necessarily go beyond externally identifying specific violations of workers’ rights and labour standards and conditions to involve the worker in reporting and responding to these violations.

The Bangladesh Accord requires that a worker complaint process is established (paragraph 18) with a mechanism that ensures that workers can raise timely concerns about health and safety risks. The purpose of the complaint process is to protect the safety and confidentiality of the worker and should be aligned with the hotline which is required separately under the National Action Plan (NAP) on Fire Safety.

Remediation is a key feature of the Accord and includes worker-oriented provisions; where closures or corrective action are undertaken, signatory companies are to make ‘reasonable efforts’ to ensure that workers who lose their employment are offered safe employment with safe suppliers ‘if necessary by actively working with other suppliers to provide hiring preferences to these workers’. Workers have the right to refuse work which they reasonably believe to be unsafe without suffering negative consequences in their employment or working conditions (paragraph 15). In addition, workers are entitled to compensation beyond the requirements of domestic legislation and play a role in negotiating it.

Both brands and unions agreed that it was important to create a strong expectation of employer responsibility within the Bangladesh context and end a culture in which factory owners could ‘cut and run’ with the profits while abandoning their legal responsibilities towards workers. (Donaghey and Reinecke, 2017)

Worker-driven monitoring must also address how disputes about labour rights and code of conduct violations are resolved. While the participation of the brand and/or lead firms may be necessary to achieve the desired level of buy-in and create opportunities for change, the involvement of multiple stakeholders incorporates an ‘inherent tension’ (Outhwaite and Martin-Ortega, 2017) because of the range of interests represented. Consequently, if these differing interests are to have a genuine voice there is a need for appropriate mechanisms which are triggered when the tensions resulting from those potentially conflicting interests cannot be resolved. Without this potential ‘exit’ it is arguable that the interests that are purportedly represented are unable to exert any real influence.

The Bangladesh Accord is an example of an initiative that has built in such an exit: the mechanism for binding arbitration in the legal system of the home country can be seen as a landmark feature, and also separating it from previous agreements on internally improving industry safety in Bangladesh (Evans, 2015). Disputes are (under para 5) initially referred to the Steering Committee. The decision of the Steering Committee may be appealed to a ‘final and binding’ arbitration process which is enforceable in the home country of the signatory against whom the complaint has been initiated (the brand company’s home state). The arbitration award is enforceable under the Convention on the Recognition and Enforcement of Foreign Arbitral Awards 1958, to which all the signatory countries of the Bangladesh Agreement are party. Such a mechanism imports the ‘teeth’ that can be lacking in other processes, backing up the governance model with meaningful enforcement. In contrast, the Indonesian Protocol incorporates a two pronged dispute mechanism which is to apply at factory level and at national level. Connor et al. note that the Protocol dispute process is subject to some limitations that may deter workers from pursuing a complaint at the national level especially those surrounding the requirement for consensus based decision making (Delaney et al., 2016).

Table 1 provides an overview of what we believe should be the core elements of worker-driven monitoring.

(b) Outstanding issues: Stakeholder roles and relationships in worker-driven monitoring

Despite the positive progress offered by some recent initiatives and by the attempts to articulate a worker-driven model above, there are questions about these new approaches that will need to be further interrogated. A crucial area of concern is the roles of, and relationships between, stakeholders in worker-driven models. Specifically, there are complex interrelationships between workers, international organizations and NGOs who are the monitoring organizations, local labour organizations or monitors and the corporate brands and these relationships are seldom articulated.

With respect to the workers themselves it is not always clear what incorporation of ‘workers’ into monitoring models really means. As noted in section ‘Monitoring supply chains: A legacy of failure?’, a criticism of corporate codes of conduct was that they constructed workers as passive subjects and that monitoring structures could have the effect of silencing workers. There is a danger that whilst beginning to illuminate this issue, existing work could unwittingly perpetuate it since there is little, to date, which explores the heterogeneity of ‘the worker’ and the implications of this for monitoring processes.

This criticism can potentially also still be levelled at newer monitoring initiatives in which international nongovernmental organizations and monitoring organizations also identify workers as a passive and homogenous group or adopt processes which restrict individual workers ability to access or participate in the monitoring process. Comments from participants in our own work have implied that, for them, local labour organizations or civil society organizations and ‘workers’ were sometimes assumed to be interchangeable. While this view perhaps reflects the practical way in which supply chain monitoring takes place, it is conceptually problematic where the voice of the workers themselves is the central concern. Any organization has its own interests and each layer of representation potentially distances factory workers from involvement in the monitoring process, for instance in being able to raise concerns and seek appropriate remedies. In this regard it is important that efforts in worker-driven monitoring can articulate who workers are – and potentially that they recognize the heterogeneity within worker populations that could give rise to differing risks and monitoring needs. More fundamentally, monitoring organizations must make sure that they do not espouse commitments to worker-driven monitoring but then distance themselves from workers in practice or fail to adequately incorporate workers voices at all.

The blurring of roles between individual workers and local organizations links with other crucial considerations for understanding and articulating the role of local organizations. Worker-driven approaches may rely on local organizations to form a bridge between individual workers and other key stakeholders, such as the monitoring organization. The Electronics Watch model, for instance, is self-defined as worker-driven and this is especially with reference to the inspection element of the process where the model relies on local civil society organizations to undertake factory inspections and then report to Electronics Watch. Where a monitoring organization is relying on local organizations there are obvious issues about transparency and accountability that will need to be articulated to avoid the problems seen with third-party auditing, namely that the local organizations have an interest in maintaining their role.

This is not to suggest that such relationships give rise to the same conflicts of interest that exist where there is a commercial benefit to the relationship: clearly they do not in the sense that they pursue different aims – although it should be recognized that civil society organizations are reliant on securing funding to maintain their own viability. Nevertheless, to engender trust and credibility appropriate reporting and transparency structures should be adopted. The relationship (or absence thereof) between local organizations and other stakeholders – such as the lead firm responsible for a brand or the factory owners and management should also be transparent.

Conclusions

Traditional approaches to supply chain monitoring have failed to detect a wide range of human and labour rights violations in GSCs and have proven unable to address the structural conditions under which these violations thrive. This article has shown that there is a movement away from such traditional approaches to monitoring GSCs and towards alternative systems of monitoring which place the worker and the workers’ voice at the centre of the process. It has also shown that this is still nascent and its main characteristics are still not defined. As we argue, the shifting of perspectives to recognition of the ‘workers voice’ as a crucial missing piece of the puzzle was a critical step in moving the approach to GSC monitoring.

A small number of initiatives have offered insights into the possibility of achieving more fundamental shifts – towards monitoring processes that are, we maintain, worker driven. In this article, we proposed a set of core elements that could form the basis of any new efforts in worker-driven monitoring and which incorporate workers throughout the process, from design through to remediation. These core elements take into account the centrality of the worker in all stages of the process, from the design, including stetting the objectives and defining the subject of monitoring and setting the governance structures and dispute resolution mechanisms, to its implementation, including inspections, training and education and complaints and remediation.

In order to develop a truly effective worker-driven model the roles and synergies between the actors involved need to be better articulated and resolved so that worker-driven monitoring does not risk replicating the mistakes or failings of earlier monitoring models. The relationships between stakeholders both at the design and implementation phases of the monitoring processes need to be carefully addressed. Furthermore, treating workers themselves as a homogeneous category ignores the complex dynamics which may impact on monitoring processes too. These are some of the further elements that require more in-depth attention from scholars and practitioners in order to articulate a worker-driven monitoring as an alternative to current mainstream practices.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors are grateful for seed funding from the University of Greenwich which facilitated this work.

Notes

References

Accord on Fire and Building Safety in Bangladesh (2019) Governance, Available at: http://bangladeshaccord.org/governance/ (accessed 1 July 2019).

AFL-CIO (2013) Responsibility Outsourced: Social Audits, Workplace Certification and Twenty Years of Failure to Protect Worker Rights. Washington, DC: AFL-CIO.

Alford

Barrientos

Visser

(2017) Multi-scalar labour agency in global production networks: Contestation and crisis in the South Africa fruit sector. Development and Change 48(4): 721–745.

Bartley

Egels-Zandén

(2015) Beyond decoupling: Unions and the leveraging of corporate social responsibility in Indonesia. Socio-Economic Review 14(2): 231–255.

Brown

(2009) Genuine worker participation – An indispensable key to effective global OHS. New Solutions: A Journal of Environmental and Occupational Health Policy 19(3): 315–333.

Brown

(2013) Fatal flaws of foreign factory audits, ISHN February 1 2013 Available at: www.ishn.com/articles/95045-fatal-flaws-of-foreign-factory-audits (accessed 1 July 2019).

Centro de Reflexión y Acción Laboral (2016) Beyond voluntary codes and audits: A challenge for the electronic Industry Seventh report on working conditions in the Mexican Electronics Industry. Good electronics. Available at: https://goodelectronics.org/wp-content/uploads/sites/3/2016/07/CEREAL-7th-Report-2016-English.pdf (accessed 1 July 2019).

Claeson

(2019) Making rights effective in public procurement supply chains: Lessons from the electronics sectors. In: Martin-Ortega

Methven O’Brien

(eds) Public Procurement and Human Rights: Opportunities, Risks and Dilemmas for the State as Buyer. Cheltenham: Edward Elgar, pp. 192–205.

Clean Clothes Campaign (2019) Freedom of Association Protocol – Indonesia. Available at https://cleanclothes.org/resources/recommended-reading/freedom-of-association-protocol-indonesia/view (accessed 1 July 2019).

10.

Coe

Hess

(2013) Global production networks, labour and development. Geoforum 44: 4–9.

11.

Delaney

Connor

Rennie

(2016) The Freedom of Association Protocol: A localised non-judicial grievance Mechanism for workers' Rights in Global Supply Chains (October 24, 2016). Non-Judicial Redress Mechanisms Report Series No. 19. Available at: SSRN: https://ssrn.com/abstract=2889112 (accessed 1 July 2019).

12.

De Neve G (2014) Fordism, flexible specialization and CSR: How Indian garment workers critique neoliberal labour regimes. Ethnography, 15(2), 184–207.

13.

Distelhorst

Locke

Pal

, et al. (2015) Production goes global, compliance stays local: Private regulation in the global electronics industry. Regulation & Governance 9(3): 224–242.

14.

Donaghey

Reinecke

(2017) When industrial democracy meets corporate social responsibility – A comparison of the Bangladesh accord and alliance as responses to the Rana Plaza Disaster. British Journal of Industrial Relations 56(1): 14–42.

15.

Electronics Watch (2019) Available at: http://electronicswatch.org/en/worker-driven-monitoring_2460012. (accessed 1 July 2019).

16.

Evans

(2015) Accord on fire and building safety in Bangladesh: An international response to Bangladesh labor conditions. North Carolina Journal of International Law and Commercial Regulation 40(2): 597–627.

17.

Fair Labor Association (2019) History. Available at: www.fairlabor.org/about-us/history (accessed 1 July 2019).

18.

Foust Rodriguez

(2016) Why worker-led monitoring is needed to challenge electronic sweatshops. The New Internationalist, 2 November. Available at: https://newint.org/blog/2016/11/02/worker-led-monitoring-in-electronic-sweatshops (accessed 1 July 2019).

19.

Fung

Olin Wright

(2001) Deepening democracy: Innovations in empowered participatory governance. Politics & Society 29(1): 5–41.

20.

Gardener

(2012) Workers’ rights and corporate accountability – The move towards practical, worker-driven change for sportswear workers in Indonesia. Gender & Development 20(1): 49–65.

21.

Hale

Opondo

(2005) Humanising the cut flower chain: Confronting the realities of flower production for workers in Kenya. Antipode 37(2): 301–323.

22.

Human Rights Watch (2017) Follow the thread: The need for supply chain transparency in the garment and footwear industry, April 2017. Available at: www.hrw.org/report/2017/04/20/follow-thread/need-supply-chain-transparency-garment-and-footwear-industry (accessed 28 August 2018).

23.

ILO, Purchasing practices and working conditions in global supply chains: Global Survey results, Inwork Issue Brief, No 10, ILO, Geneva, June 2017. Available at: www.ilo.org/wcmsp5/groups/public/—ed_protect/—protrav/—travail/documents/publication/wcms_556336.pdf (accessed 28 August 2018).

24.

International Labor Rights Forum and the Migrant Workers Rights Network (ILRF, 2016) Building a Rights Culture: How Workers Can Lead Sustainable Change in Thailand’s Seafood Processing Sector. Washington DC: ILRF.

25.

Klein

(2000) No Logo. Toronto: Random House.

26.

Kobrin

(2008) Globalization, transnational corporations and the future of global governance. In: Scherer

Palazzo

(eds) Handbook of Research on Global Corporate Citizenship. Cheltenham: Edward Elgar, pp.249–272.

27.

LeBaron

Lister

(2016) Ethical audits and the supply chains of global corporations, SPERI Global Political Economy Brief No. 1, Sheffield Political Economy Research Institute, University of Sheffield.

28.

LeBaron

Lister

Dauvergne

(2017) Governing global supply chain sustainability through the ethical audit regime. Globalizations 14(6): 958–975.

29.

Locke

(2002) The Promise and Perils of Globalization: The Case of Nike. MIT Working Paper IPC-02-007. Boston: MIT.

30.

Locke

Romis

(2007) Improving work conditions in a global supply chain. MIT Sloan Management Review 48(2).

31.

Locke

Kochan

Romis

, et al. (2007) Beyond corporate codes of conduct: Work organisation and labour standards at Nike’s suppliers. International Labour Review 146(1–2): 21–40.

32.

Locke RM, Rissing BA and Pal T (2013) Complements or substitutes? Private codes, state regulation and the enforcement of labour standards in global supply chains. British Journal of Industrial Relations 51(3): 519--552.

33.

Lund-Thomsen

Lindgreen

(2014) Corporate social responsibility in global value chains; Where are we now? Where are we going? Journal of Business Ethics 123(1): 11–22.

34.

OECD (2018) Guidance on responsible business conduct. Available at: www.oecd.org/investment/due-diligence-guidance-for-responsible-business-conduct.htm (accessed 1 July 2019).

35.

O’Rourke

(2003) Outsourcing regulation: Analysing nongovernmental systems of labor standards and monitoring. Policy Studies Journal 31(1): 1–29.

36.

Outhwaite

Martin-Ortega

(2016) Human rights in global supply chains: Corporate social responsibility and public procurement in the European Union. Human Rights and International Legal Discourse 10(1): 54–73.

37.

Outhwaite

Martin-Ortega

(2017) Monitoring human rights in global supply chains. Insights and policy recommendations for civil society, global brands and academics. BHRE Research Series. Policy Paper n.3.

38.

Prieto-Carron

(2004) Is there anyone listening? Women workers in factories in Central America, and corporate codes of conduct. Development 47(3): 101–105.

39.

Prieto-Carron

Lund-Thomsen

Chan

, et al. (2006) Critical perspectives on CSR and development: What we know, what we don't know, and what we need to know. International Affairs 82(5): 977–987.

40.

Pruett

(2005) Looking for a Quick Fix: How Weak Social Auditing Is Keeping Workers in Sweatshops. Amsterdam: Clean Clothes Campaign.

41.

Rodriguez-Garavito

(2005) Global governance and labor rights: Codes of conduct and anti-sweatshop struggles in global apparel factories in Mexico and Guatemala. Politics & Society 33(2): 203–233.

42.

Ruggie

(2008) Multinationals as global institution: Power, authority and relative autonomy. Regulation and Governance 12(3): 317–333.

43.

Ruwanpura

(2016) Garments without guilt – Uneven geographies and ethical trading – Sri Lankan labour perspectives. Journal of Economic Geography 16(2): 423–446.

44.

SA8000 (2018) Available at: www.sa-intl.org/index.cfm?fuseaction=Page.ViewPage&pageId=472 (accessed 1 July 2019).

45.

Short JL, Toffel MW and Hugill AR (2016) Monitoring global supply chains. Strategic Management Journal 37(9): 1878--1897.

46.

United Nations. (2011). Guiding principles on business and human rights: Implementing the United Nations “Protect, Respect and Remedy” framework.

47.

United Nations (UN) Human Rights Council (2008) Protect, respect and remedy : a framework for business and human rights : report of the Special Representative of the Secretary-General on the Issue of Human Rights and Transnational Corporations and Other Business Enterprises, John Ruggie, 7 April 2008, A/HRC/8/5.

48.

Worker Rights Consortium (WRC) (2019) Available at: www.workersrights.org/ (accessed 1 July 2019).

49.

WRAP (2019) Available at: www.wrapcompliance.org/en/history (accessed 1 July 2019).

50.

WSR Network (2017) What is worker-driven social responsibility (WSR)? July 2017. Available at: https://business-humanrights.org/sites/default/files/documents/What_is_WSR_0.pdf (accessed 1 July 2019).

51.

Zimmer

(2016) Corporate Responsibility in the Bangladesh Accord. Which Regulations Are Transferable to Other Supply Chains? Berlin: Friedrich Ebert Stiftung.