Corps de l’article

Introduction

As the global economy diminishes the regulatory capacity of the nation-state, transnational forms of labour regulation are created (or attempted) to fill the vacuum. This change of regulatory regimes is neither smooth nor unproblematic. Alternative forms of transnational labour regulation are yet to become universally established, but the trend is unmistakable.

Civil society institutions have played a crucial role in that process of recognition; campaigning groups, “direct action” and other citizens' movements, have generated previously unheard of awe-inspiring opportunities for social activism at both national and transnational levels (Munck and Waterman, 1999; Waterman, 1998 and 2001). International labour regulation is, for the most part, a campaign-driven domain, pursued by various groups and movements and for different purposes, not a theoretically or scholarly driven discourse. Numerous publications have examined different forms of international labour regulation, such as corporate codes of conduct, business social accountability, labelling schemes, child labour and social clauses in trade agreements. There is no common theoretical perspective, nor a clear set of distinct theoretical approaches. Most writers approach the subject from their own disciplinary, institutional, or empirical perspectives; interdisciplinary approaches are far less common. Literature has focused primarily on the merits and mechanisms of social clauses and labour regulation instruments; a small segment has concentrated on assessing and analyzing the validity of the various arguments raised (Krueger, 1996; Lee, 1997; Scherrer, 1996; Tsogas, 1999), while very little has been written on the actual or potential industrial relations and labour market effects (Tsogas, 2001).

The aim of this paper is to highlight some of the lessons learnt from existing international labour regulation schemes in the last two decades of the 20th century, and to draw some policy and research implications. We introduce an updated model of international labour regulation and create a conceptual framework for analyzing labour regulation. We provide some insights into how regulation has developed over the last decades and discuss some of the challenges it faces. Our assessment of the various regulatory regimes is based on the simple premise of whether they can provide a venue for workers' rights violations to be redressed. We conclude with some perceptions into how international labour regulation could evolve in the 21st century. We aim to provide a broad overview and an attempt at generalizing the findings and “lessons learnt” so far from an international and comparative industrial relations perspective.

Inevitably, other aspects of labour regulation, such as trade policy, labour law, and developmental or economic perspectives are beyond our scope, and so are broader contemporary IR issues, such the effects of globalization, union renewal issues, etc. Unfortunately, space limitations do not allow us to get into detailed discussions of empirical evidence and the complexities of different forms of regulation. Similarly, another limitation on our study is the focus on the formal sector of the economy, leaving aside the informal sector where huge segments of the workforce—up to 90% in some developing countries—find employment (Schlyter, 2002). International labour regulation that would incorporate the informal sector is a challenging area that demands further research.

An Analytical Framework for International Labour Regulation

There is already rich experience in introducing social clauses in international trade regimes. Accordingly, there have been a few endeavours to put forward a framework of analysis. Stone (1996), alarmed by “the challenge to domestic labour regulation of the increasingly international economic and legal order” suggested four models of transnational labour regulation “that are emerging in fact”: pre-emptive legislation; harmonization of domestic legislation; cross-border monitoring and enforcement; and extraterritorial application of domestic law (extraterritorial jurisdiction). She defines pre-emptive legislation as being “applicable to persons, business entities, and states within a transnational bloc.” Harmonization refers to transnational bodies, such as the EU, that seek to harmonize their domestic laws. Finally, cross-border monitoring and extraterritorial jurisdiction apply to legal practices in North America. She further proceeds to add two dimensions on the four models: multilateral and unilateral regulation on the one hand, and integrative and interpenetrative approaches, on the other (see table 1). Pre-emptive legislation and harmonization are seen as essentially European approaches, integrative in nature, aiming to make disparate regulatory systems congruent and consistent over time. In contrast, the North American models (cross-border enforcement and extraterritorial jurisdiction) “employ an interpenetrative approach, the temporary incursion of one legal system into the affairs of another” (Stone, 1996: 447). Secondly, the four models “can be distinguished as to whether they can be implemented unilaterally, or whether they require multilateral action for their implementation.” Pre-emptive legislation and cross-border enforcement are multilateral “in the sense that they rely for their implementation on actions by several countries jointly enforcing a particular labour standard.” In contrast, harmonization and extraterritorial jurisdiction “are unilateral in the sense that they can be implemented by unilateral action of one country” (Stone, 1996: 469).

Table 1

Four Models of Transnational Labour Regulation

 

Multilateral regulation

Unilateral regulation

Integrative approaches

Pre-emptive legislation

Harmonization

Interpenetrative approaches

Cross-border enforcement

Extraterritorial jurisdiction

Source: Stone (1996: 470).

-> Voir la liste des tableaux

Even though, the above categorization appears to be straightforward, it nonetheless creates a rather fuzzy perspective—especially when the interrelation between the various models is extrapolated. Additionally, there is no account of a major contemporary source of international labour regulation stemming from companies that intervene and regulate employment relations in their supply chains through voluntary codes of conduct.

Tsogas (1999) introduced a framework of international labour regulation based upon the types of trade relations involved, classified into four types: unilateral, bilateral, regional, and multilateral. The unilateral approach involves trade laws of a country that contain a social clause embracing all of its trade relations. Examples of that are US laws on imports produced by child or prison labour. Additionally, corporate social responsibility (CSR), which includes various social labelling schemes and voluntary corporate or industry codes of conduct regulating labour standards in supply chains, could also be included, as policy measures taken unilaterally by a company and involving its intra-firm trade, or traded goods and services through supply chains (Mamic, 2004; Murray, 1996; NEF/CIIR, 1997; O'Rourke, 2003; Tsogas and IDS, 1998; US Department of Labor, 1996; Utting, 2002a).

The bilateral level entails social clauses in international trade agreements that regulate relations between one country and some of its trade partners. Examples of this approach include: (1) the European Communities' attempt to incorporate labour standards provisions during the Lomé II negotiations (1978–79); (2) the inclusion of labour standards in US trade legislation, such as the Caribbean Basin Initiative (CBI), the Overseas Private Investment Corporation (OPIC), the Generalized System of Preferences (GSP), the 1988 Omnibus Trade and Competitiveness Act, the 1991 Andean Trade Preference Act and the US-Jordan Free Trade Agreement; and (3) and the inclusion of social clauses in the GSP of the European Union (EU). Of these schemes, the US and EU GSPs are by far the most important (Compa and Vogt, 2001; Frundt, 1998; Tsogas, 2000).

The regional track incorporates social clauses in regional trade agreements such as the European Union and NAFTA (Compa, 1995 and 1997; Cook and Katz, 1994; Elliott, 2003; Griffin, 1997).

Finally, the multilateral track embraces efforts to incorporate a social clause in the World Trade Organization and its predecessor, the GATT—efforts which have so far proved unsuccessful (De Wet, 1995; Leary, 1997; Wachtel, 1998; Wolffgang and Feuerhake, 2002).

Here, we introduce a further dimension in the above model. In trying to make sense of the complexities and ambiguities of the types of regulation that we are seeking to present, we must account for an important distinction between “public” instruments that link states and seek to regulate their behaviour, on the one hand, and instruments that regulate the behaviour of “private” (i.e. non-governmental) parties, on the other. That differentiation can be understood and expressed in terms of a fundamental dichotomy existing between civil society and CSR-“inspired” i.e. “soft” regulation instruments, and other trade law related, i.e. “hard” approaches to international labour regulation (Kuruvilla and Verma, 2006; Sisson and Marginson, 2001).

Incorporating “Soft” and “Hard” Regulation in an Analytical Model of International Labour Regulation

For the purposes of trade-rated social clauses, “hard” approaches rely on trade laws and international treaties with legally enforceable measures, where some form of redress could be the outcome of a legal challenge or a petition and review process, initiated by an interested party (a governmental body, an association, or even an individual). The object of the hard approaches is the nation-state and the trade relations that it establishes. In contrast, “soft” approaches are voluntary, based not in legal texts but at various—usually ad hoc, company or industry specific—policy documents and implementation schemes, a “soft” alternative to the “hard,” legally abiding social clauses in trade, or even an “outsourced” or “privatized” form of labour regulation (O'Rourke, 2003; Burkett, Craig and Link, 2004). Their object is not the state but the enterprise or industry or, even more specifically, the individual decision-makers within a company. This fundamental difference makes soft regulation a truly unique, challenging, and inspiring form of regulation. Another difference between “hard” and “soft” approaches rests on who is the primary driving force behind them. “Soft” approaches have overwhelmingly been supported, and in certain cases created and administered, by civil society institutions. In contrast, social clauses in trade agreements have been established and are administered by government departments and agencies, intergovernmental organizations and other similar “formal” bureaucracies.

“Soft” law and its use in international labour regulation have enjoyed considerable attention recently (see ILO, 2008). As Duplessis (2008) and Sisson and Marginson (2001) note, soft forms of regulation cover a wide area of actors and applications, (including even national implementation of EU policies). Here we take a more narrow view, looking at soft regulation through the lens of labour standards' implementation in conjunction to international trade relationships.

The above characteristics of our international labour regulation framework can now be summarized in table 2, named after the four levels of international trade relationships that it depicts (Unilateral, Bilateral, Regional, and Multilateral—UBRM).

Table 2

The UBRM Model of International Labour Regulation

Regulatory

Level of Trade Relations

Subjects

Objects

Unilateral

Bilateral

Regional

Multilateral

Company

Soft

CSR/CoC

Framework Agreements

 

UN Global Compact

State

Hard

USA child and prison labour legislation

GSP (USA, EU)

Lomé II

CBI

OPIC

Andean Trade

Preference Act

European Union

NAFTA/NAALC

WTO/GATT

ILO

-> Voir la liste des tableaux

In addition to the various regulation schemes identified by Tsogas (1999), we have appended here—under the multilateral level, in the soft regulation path—the United Nations' Global Compact that aims to regulate the behaviour of transnational corporations through the multilateral role of the UN (Bendell, 2004; Utting, 2002b). Similarly, framework agreements are included here as a bilateral and a soft form of regulation, negotiated between a multinational company and an international sectoral trade union federation (Riisgaard, 2005; Schömann et al., 2008).

Advantages and Policy iImplications of the UBRM Model of Regulation

In contrast to Stone's four models, the UBRM model adopts an open-ended, global trade perspective, rather than a more restrictive domestic labour law approach. In that respect, it encompasses a different philosophy on global labour issues. First, by setting international trade as its base reference point, the UBRM model links up with an age-old debate of international industrial relations: how labour standards can be regulated, safeguarded, and improved in the context of increased trade and mobility of capital. Indeed, linking forms of labour regulation with trade has a very long history in IR: since the early 19th century all discussions on international labour standards (or comparisons of national labour standards) have been conducted in the context and effects of (expanding) international trade. Secondly, its open-ended form suggests a positive, forward-looking perspective; international labour regulation is not a challenge to established domestic norms, but rather an opportunity to be fully exploited. Finally, the UBRM model provides the flexibility to account for both actual and emerging schemes to be accounted for, but can also envisage new forms of regulation. Could there, for example, be a regional form of soft regulation? We will return to the issue of how the UBRM model can stipulate new forms of regulation and venues for advocacy and public policy intervention and also aid social activism, later.

Finally, the UBRM model can also be seen as a regulatory web of the various regimes of international labour regulation that a transnational corporation could potentially encounter, now or in the future, in its operations across the globe. Which one of these schemes would be relevant for a particular company depends upon the industry and countries within which it operates, the trade agreements that link those countries with the rest of the world, the nature of its supply chains, and, of course, the labour and social responsibility policies that the company itself applies. In most cases, there could be multiple levels of labour regulation schemes affecting the same company.

However, whether that threat of multiple exposures to regulatory regimes can be an effective measure or is just wishful thinking for those of us who aspire to a well-functioning international regulatory setting of labour standards, is a matter of further detailed research. The crucial issue for those who advocate workers rights, and a meaningful protection against the excesses of corporate power, is not whether a corporation has a code of conduct in place or if it comes under the scope of NAALC or a GSP programme—however important these might be—but whether these schemes can provide a venue for worker rights violations to be redressed. This paper critically engages with that issue in the pages that follow.

Selective Assessment of “Hard” Regulation: Have Social Clauses in Trade Agreements Worked?

The idea of a social clause in a trade agreement has a long history, predating the establishment of the ILO in 1919. The issue has been debated exhaustingly by economists and trade theorists (Bhagwati, 1994; Fields, 1990; Piore, 1990; Sengenberger, 1994), international trade and human rights lawyers (Alston, 1993; Compa, 1993), government policymakers (Marshall, 1987, 1988, 1989, 1990), international trade union bodies (ICFTU, 1998; IMF, 1988) and by a plethora of NGOs and activists.

The Unilateral Level

Social clauses in US trade legislation date back to the late 19th century. The US Tariff Act of 1890 banned imports of goods manufactured by prison labour (Zimmerman, 1992). Section 307 of the Smoot-Hawley Tariff Act of 1930, extended further the provisions of the Tariff Act to products mined or produced by prison, forced or indentured labour under penal sanctions. Section 307 remains in force. The main weakness of unilaterally enacted social clauses has been the requirement that the task of enforcement falls on the Customs Service. In practice, this has been impossible, since Customs could only rarely rely on credible knowledge of prison labour cases.

In the mid-90s, huge controversy was raised over attempts by US Senator Tom Harkin in 1989, 1990, 1993, 1995 and 1997, to introduce a bill (“The Child Labor Deterrence Act,” or commonly known as the “Harkin bill”) that would prohibit the importation of any product made, whole or in part, by children under the age of 15. The 1997 bill directed the US Secretary of Labor to make a list of foreign industries and their host countries that use child labour in the production of goods exported to the United States. Once an industry was identified, the Secretary of the Treasury was to prohibit the importation of such products. However, products from such an industry would be allowed into the US, if a US importer signed a certificate of origin stating they took reasonable steps to ensure the imported products were not made with child labour. Any company or person who violated this law would have faced civil and criminal penalties.

Hertel (2006) detailed the backlash that the Harkin bill created in Bangladesh and the controversial results it had over the fate of allegedly thousands of children who lost their jobs.

The Bilateral Level

Since 1983, the US Congress has passed several legislative items that contain worker rights clauses in bilateral US trade relationships: the Caribbean Basin Initiative (CBI) of 1983, the 1984 amendments to the Generalized System of Preferences (GSP), the 1985 amendments to the Overseas Private Investment Corporation (OPIC), the Omnibus Trade and Competitiveness Act of 1988 (OTCA). Of these, most important in terms of effectiveness and history has been the GSP programme. The law details a petition and review process, where an interested party can petition the US Trade Representative asking for a review of workers rights violations in a beneficiary country (Tsogas, 2000).

The EU has operated a GSP since 1971. In 1994–95 provisions for granting preferential treatment to countries that observe certain labour standards and withdrawal of GSP privileges from those that do not, were adopted. The scope of the European GSP is twofold: GSP tariff benefits can be withdrawn in cases of forced or prison labour, but further tariff concessions can be granted for products produced in observance with international labour standards, including freedom of association and protection of the right to organize and bargain collectively, and minimum age of employment (for a full analysis see Tsogas, 2000).

Labour rights conditionality under GSPs has provided substantial practical experience on the pitfalls and accomplishments of linking trade with labour standards, and can contribute valuable lessons. As there has been demonstrated elsewhere (i.e. Tsogas, 2000; Frundt, 1998), foreign policy considerations and economic nationalism have dominated the selection (or non-selection) of countries for review. Further, the openness of a particular country to outside scrutiny, its political relationship with the US/EU, and the links that any local trade unions or human rights organizations may have with international bodies can also be decisive. The very few countries that were suspended from the US GSP were done so either because their governments were disliked by an administration (Nicaragua, Syria) or because their trade with the US or EU was totally minimal (Romania, Paraguay, Burma, Central African Republic, Liberia, Sudan, Mauritania).

Inclusion of a social clause in the EU GSP has been very recent. “Reward” of good behaviour was only introduced in May 1998. Therefore, the effects and experiences that such incorporation would have brought are very limited, in comparison with those from the US. The degree to which the scheme offers real economic benefits also remains to be explored.

Experience has shown that both the EU and the US use their economic and political might as a leverage to demand (at least in theory) improvement in conditions of employment from their trading partners (nowhere, though, in either scheme is there a mention of labour standards in the USA or the EU respectively). In practice, the US simply exercised its economic and political power over countries that the government dislikes and whenever it deems it appropriate to do so. Similarly, in the EU, the process of petition and review is a political matter; any eventual withdrawal of preferences is a decision to be taken in the EU Council by “qualifying majority,” where a small group of countries can block action. In the EU, so far, there has been a lack of political will even to start the review process, let alone proceed with the enforcement of the GSP statutes.

The Regional Level

The North American Agreement on Labour Cooperation (NAALC) does not establish any labour standards of its own nor does it provide for harmonization of existing labour legislation. The signatory countries (Canada, the USA and Mexico) undertake simply to comply with and to enforce their own labour laws fairly. NAALC has not sought to introduce new standards or to require any country to match standards prevailing elsewhere within the Free Trade Area. Despite this limitation, NAALC is the first trade agreement to establish contractual rights and obligations, with trade sanctions provisions in the event of violations of labour standards. NAALC has indeed created a unique basis for a form of “transnational access to justice” and accountability affording the practical opportunity for an interested party to pursue a company for its activities, not only in the country where these activities took place but in a third country which maintains a trade relationship with the former (Housman, 1994). However, under this “transnational access to justice”, labour and environmental issues are at the bottom of the hierarchy of access: in practice, they have no form of redress. In that respect, the pioneering legal concept of “transnational access to justice” that NAALC introduced has unfortunately remained a mainly theoretical possibility and has attracted very little attention in the literature. Moreover, the judicial procedures and timetable between a judgment and the application of sanctions in the event of a failure to correct the grievance are much quicker than the protracted procedures for labour issues. The NAALC provisions simply lack the “teeth” to correct violations—with seminars and consultations taking the place of trade sanctions.

Selective Assessment of “Soft” Regulation: Corporate Codes of Conduct and Social Responsibility

Social accountability has acquired an immediate significance for many managers, activists and scholars in recent years. This has often been the result of unfavourable media exposure in both domestic and foreign operations and the response to this by increasingly sensitized consumers. High profile brands have been ensnared in public controversy and obliged to take emergency remedial action in the glare of publicity. The underlying forces, which have shaped this new agenda, are unlikely to go away; in fact, they have grown stronger over the dawn years of the 21st century and have transformed, for example, into a general acceptance of the necessity for companies to produce social impact and CSR reports, a rarity just five years ago.

The impact of campaigning on companies has not been the subject of a detailed empirical research. Claims that share prices are affected by campaigns have been difficult to sustain against the general turmoil of markets (Ettenson et al., 2006; Klein, Smith and John, 2004). This does not, of course, mean that there is no impact on company performance. And crisis management to respond to campaigns can be unsettling and disruptive for senior management and the decision-making process. Unmanaged or mismanaged issues can spiral out of control, and lead to consumer boycotts or less tangible forms of brand avoidance.

Corporate codes of conduct avoid one thorny issue associated with other forms of regulation, but raise a host more. The thorny issue is that of the extra-territorial impact of legislation enacted by governments in the developed world with the aim of altering the behaviour of actors—companies and governments—in developing countries (Utting, 2007). The advantage of corporate codes in this respect is that it could be argued that they reflect: (mostly middle-class) “consumer choice”—in that companies are responding to the demands of their customers for “ethically-sourced” products, and “purchaser choice”—in that companies can source from where they like and on the conditions they can negotiate.

Unlike previous eras, during the 1990s and especially through the early years of the 21st century, there has been an explosion of voluntary codes dealing specifically with labour regulation issues (OECD, 2001; Tulder and Kolk, 2001). Corporate codes of conduct, intended to regulate the behaviour, practices and standards of the participants in supply chains, have assumed a new prominence because of the problems in developing and enforcing effective multilateral or industry-specific regulation of labour standards—be this in the WTO, ILO or via industry initiatives. For the most part, codes have so far embraced retailers, owners of high-profile brands, and in some cases both where these overlap and where brands own their own manufacturing facilities in developing countries. Industrial manufacturers from the developed countries have generally not figured in the debate as yet.

Codes represent one of the first steps that companies can take when tackling the issue of how to respond to criticism of standards in their supply chains. Codes have also allowed companies to respond to the challenges of NGOs by initiating improvements immediately. They have the advantage—for companies—of being under corporate control and subject to internal compliance procedures. Where other parties are involved, such as trade unions via negotiation or NGOs via consultation, this control may be diluted, but the code may acquire a higher degree of external acceptance and legitimacy.

Codes not only offer a number of benefits for all actors engaged, they may also be the only form of regulation which can be realistically developed in the short-term. However, scepticism about codes on the part of NGOs and trade unions means that companies need to do more than simply issue a text without specifying effective means for implementation and monitoring, and some procedure for verifying that implementation is taking place (Barrientos and Smith, 2007).

The style and format of codes vary considerably but their main components generally are: prohibitions on child labour, either in terms of national law, their own definition or sometimes no precise definition; prohibitions on forced or indentured labour; prohibitions on discrimination based on race, religion or ethnic origin, prohibitions on certain types of disciplinary practice (physical or psychological punishment or unreasonable fines); provisions on health and safety, both at the workplace and in some cases in employee accommodation; provisions on pay, hours of work, rest breaks and time-off, and the regulation of overtime; provisions on the rights to organize and engage in collective bargaining. The precise standards required vary considerably, with some codes referring to the “core” ILO Conventions, some simply to compliance with local laws and regulations, and others making less specific provisions. Many codes contain a blanket provision requiring compliance with all applicable national laws and regulations.

The response by NGOs, governments, sectoral partnerships and individual companies to this evolving environment has given rise to a proliferation of initiatives, and a welter of public relations activity on the part of companies. The outcome has been a lack of transparency as to the institutions involved and the virtues and drawbacks of the numerous schemes currently in existence or proposed. Schemes such as the SA8000 or the AA1000 are positioning themselves as universal standards for social accountability (for a review see Tsogas, 2001: 72–76).

The New Tripartism of “Soft” Regulation: Companies, NGOs and Trade Unions

One of the most interesting characteristics of the emerging “soft” regulation is the opportunity for NGOs, companies and trade unions to form new types of tripartite cooperation. NGOs have become increasingly willing to adopt a strategy of “constructive engagement” with manufacturers and retailers as a tactical and possibly a strategic option. This has been based, however, on influence won through high-profile media exposure of leading companies and the fact that such exposure has been seen to have damaging effects on carefully and expensively crafted company and brand images. In essence, they are succeeding in bringing companies to the negotiating table, at a time when trade unions find it increasingly difficult to do so.

Many companies have also—sometimes after a sharp internal debate—decided to engage with NGOs rather than to seek to ride out campaigns or mount a legal challenge to refute NGO claims, although this has been an option which has been used where claims have not been wholly substantiated or have been misleading. The existence of government-sponsored initiatives, such as the UK's Ethical Trading Initiative (ETI), the mixed composition of the advisory board of the Council on Economic Priorities Accreditation Agency (CEPAA), and the US. Apparel Industry Partnership all testify to the fact that some NGOs and companies have found mutual advantage in working together, with trade unions, to address the issue of employment standards.

The Role of Trade Unions

Trade unions participating in joint initiatives (as opposed to negotiating an agreement) face an unchartered territory. In particular, as the historic representatives of labour, and as the institutional representative of workers within the ILO, they are concerned at being displaced by NGOs in such joint initiatives—as well as being uneasy about engaging with business on a non-negotiating basis. The international trade union movement, with its central call for the inclusion of core labour standards in trade agreements, has been somewhat sidelined by the focus on corporate initiatives. Trade union activists and representatives have tended to regard corporate efforts as, at best, inadequate to the task at hand and, at worst, tainted by the suspicion that companies are only interested in massaging their reputations. Codes of conduct have been seen as “privatizing” labour regulation—the proper subject of the state and official international bodies such as the ILO. Voluntary codes of conduct have been regarded with particular scepticism. Trade unions would prefer a system of multilateral regulation, in which trade and human and labour rights are tied together, with membership of the WTO made conditional on acceptance of basic rights (Wick, 2003). Trade unions are also concerned that issues of freedom of association may take last place in any action agenda, and possibly be dropped altogether once the more “controversial” issues, such as child labour, have been addressed.

But there are also some new opportunities for trade unionism. The lack of progress at multilateral level has prompted trade unions to develop alternative ways of tackling these issues, and these have been in train and gathering momentum over the past few years. Both in North America and Europe, trade unions have now begun to support and participate, through negotiations, in a new wave of activity that seeks to address the issue of labour standards at industry and corporate level through direct contact with employers and collaboration with NGOs. In some cases, this has evolved into efforts to negotiate codes of conduct with employers, even though substantial difficulties remain (Miller, 2004). Even though examples of negotiated codes are still small in number, there is already some movement in that direction. Important examples include Danone in 1988 (the very first one), the hotels chain ACCOR (the second, in 1995) and then an acceleration in the number of agreements signed per year to reach 50 by the end of 2006. Among these are the Swedish furniture company IKEA (1998); the Italian toy manufacturer Artsana (Chicco brand) in 1999, the banana company Chiquita, the German stationery producers Faber-Castell and Staedler, oil companies in Norway (Statoil), Italy (ENI) and Russia (Lukoil), the car producers Volkswagen, Daimler-Chrysler, Renault and Peugeot-Citroën, the Spanish and French electricity producers Endesa and EDF, the telecom companies in Spain (Telefonica) and Greece (OTE), and retailers in France (Carrefour) and Sweden (H&M) (Tsogas and IDS, 1998: 87–90, Schömann et al., 2008).

Indeed a negotiated approach is wholly in accord with labour movement traditions in this field, and in particular in line with the efforts by garment unions in the US in the 1920s and 1930s to establish “joint liability,” in which manufacturers were obliged to accept responsibility for conditions in their contractors (Howard, 1997).

The issue of enforcement and the legal status of agreements with trade unions is a major area still in need of clarification—although the fact that collective agreements in the UK are not legally binding simplifies this issue somewhat for British participants. In Mainland Europe, there is a strict hierarchy of agreements and a division of labour between trade unions and elected workforce representatives as to who can sign what type of agreement. Agreements may also be legally binding, depending on the status of the signatory parties. In some European jurisdictions collective agreements not only apply as between the employer and employees (as in the UK via incorporation) but also establish a contractual relationship between the signatory employer or employers' association and the signatory trade union. As a consequence, in theory, a company could be brought before a national court for breaching its contractual obligations towards a trade union on the grounds that it violated this agreement in another jurisdiction. For this reason, many agreed codes are likely to take the form of accords rather than contractual obligations; this would at least allow concentration on the substance of the issues, rather than involving lawyers in phrasing texts which might have to withstand challenge in the courts—a process which can be destructive of efforts to reach agreement. Nonetheless, most agreements are still at an early stage of implementation. As such, many have not yet encountered the problems of enforcement that have characterized voluntary codes of conduct. But, negotiated codes will also be confronted with many of the policy issues and dilemmas that have confronted those in the corporate sector who have been attempting to put voluntary codes into practice. For example, what to do about non-conformance, and what remedial programmes should be in place; what to do about employees in the informal sector whose status could be prejudiced by a concentration of work in more easily monitored centralized industrial plants.

In Europe, the institution of the European Works Council (EWC) may also offer a forum in which such issues could be aired at company level. Both employer and employee sides on EWCs are often eager to find international themes on which consensus could be achieved, and a discussion on a code of conduct or labour standards might offer such a terrain (Edwards et al., 2007). As yet, activity in this area is minimal and many employers could well be reluctant to accord a formal status to such discussions, fearing that it might set a precedent for other forms of European-level bargaining that they are eager to avoid. Employee members of EWCs would also need extensive support and advice. At present, both national and international trade unions are severely stretched, finding it difficult to service existing EWCs. They would be hard-pressed to devote resources to what might still appear to be a peripheral issue to many employee representatives. However, this might be an area in which NGOs could work to support trade union initiatives by offering expert advice, and many EWC agreements contain provisions that allow such advice to be drawn on.

For many companies, on the other hand, working with trade unions is not something rooted in their own cultures and practices. Most individuals involved with ethical sourcing in companies have had no exposure within their own organizations to employee representatives. For them, the issue has been one of responding to perceived or actual customer concerns, to lobbying by NGOs, and to initiating action through quality mechanisms, not employee relations' structures. The institutions in which companies, NGOs, consultants and trade unions sit together have not been operating sufficiently long enough to establish whether these experiments can succeed.

Each side has still to find a comfortable and legitimate role within a new tripartitism based entirely on voluntarism and “soft” regulation of labour; one in which NGOs may have to recognize that companies may not be able to make the radical changes which they might wish for as swiftly as they would like, one in which companies recognize that NGOs should not be drawn into quiescence—however convenient that might appear—and one in which trade unions are not the de facto representatives of the workers nor rely on collective bargaining to bring about improvement of working conditions. Unless mutual learning and some accommodation take place, simply bringing together organizations with different views and agendas—and subject to different exigencies—will not resolve problems but simply internalize them for a period.

Conclusions

Our analysis has shown the limits of country, government, and hard-law based international labour regulation, but also the ambiguities and challenges of soft labour regulation.

On the one hand, rooted in traditional, state-based employment systems, 20th century labour regulation (designed by and for social actors and problems of the late 19th and early 20th centuries) has been proven ill suited to tackle labour rights violations in the global economy era. The ILO was set up in 1919 to provide a solution for a late 19th and early 20th century “social question”: to dissuade working classes in developed capitalist economies from embracing revolutionary ideas. Welfare capitalism, collective bargaining, and state-sponsored tripartism worked exceptionally well throughout the 20th century for the formal economies of the “U-labour” (unionized and unionizable workforce, see Waterman, 2001). But, in the 21st century new forms of work go far beyond both the “U” form, and the global economy makes full use of the informal sector. On a global scale, the informal economy is the norm, not the exception, nor does it rest any longer at the periphery of global economic activity. Yet, current forms of labour regulation apply only to formal economic structures, leaving aside that vast majority of working people. That made some sense in the industrial era, where the industrial worker was the locomotive of economic growth, but as the global economy now engulfs all forms of work (formal, informal, precarious) current labour regulation is left without an object. 21st century international labour regulation must incorporate the informal sector. However, that cannot be done by thinking and acting within the same reference constraints; the state does not have to be the only legitimate source of work regulation. People in developing countries, for example,—in the near total absence of regulatory enforcement mechanisms—regulate their work through the collective entities to which they belong (neighbourhood, village, tribe, social group, etc); they certainly do not wait for the government. Labour regulation in the 21st century (especially if the informal economy is to be included) has to be understood in wider terms, involving the social formations people join through and for their work. And what has hard labour regulation achieved through social clauses? Lack of openness to public scrutiny, slow processes, and ultimately complete lack of redress offered to victims of labour standards violations, have marked its more than 20 years' experience.

Even though further research could clarify this crucial matter, it seems that we have now reached—and passed—a turning point; those in civil society who are genuinely concerned about workers' rights and social development have opted to go directly after the corporate culprits and pressure them to act upon labour standards abuses that take place in their own supply chains, rather than rely on what they perceive to be impersonal bureaucracies to conduct country reviews, reports and hold summits.

That movement, problematic and scary to some as it might be, brings about a sea change on how we understand and deal with employment and its protection. Of course, that does not mean that hard-won and well-functioning mechanisms of labour regulation are to be abandoned—it is by no means an either/or situation (Arthurs, 2001). But the trend is already well established and the implications can be immense, but yet quite undetermined by academic research.

On the other hand, soft regulatory approaches and CSR have their own shortcomings. Can they ever be a substitute method for pursuing meaningfully across the globe violations of labour standards? For the time being, we can only conclude that hard law and social clauses in trade agreements have been proven to be completely unable to deliver.