Trade Agreements and Sustainability: Exploring the Potential of Global Value Chain (GVC) Obligations

This article investigates the potential of global value chain (GVC)-orientated sustainability provisions in regional and bilateral trade agreements (FTAs). Such provisions impose social and environmental obligations directly onto GVCs, as opposed to creating obligations for governments. The theoretical potential of GVC provisions is examined, and the concepts of effectiveness and legitimacy are introduced as values by which to assess them. Four recent sets of provisions are then scrutinized. These are (i) palm oil sustainability standards from the Indonesia–European Free Trade Association (EFTA) Comprehensive Economic Partnerhsip Agreement (CEPA) FTA, (ii) hen welfare standards in the European Union–Mercosur Association Agreement, (iii) a stipulation of a minimum average wage for the automobile industry in the US–Mexico–Canada Agreement (USMCA), and (iv) enforcement of collective bargaining and freedom of association directly against factories, also in the USMCA. All of these provisions are found to have significant deficiencies. At the same time, it is argued that three different governance models underpin them, namely (i) third-party certification schemes, (ii) domestic regulations of one of the parties, and (iii) bespoke arrangements created for the FTA in question. The article therefore considers the potential and drawbacks of each governance model in terms of their effectiveness and legitimacy, as well as alternative and complementary commitments including unilateral measures and subject-specific trade agreements.

A number of studies have mapped out the growing number of bilateral and regional trade agreements (FTAs) in recent years, signed by a variety of different countries, which contain such provisions on sustainability issues. 1 The term 'value chain' describes the complete scope of activities required to bring a product or service from conception, through the different phases of production, its distribution to final consumers, and its final disposal. 2 The global fragmentation of tasks involved in the production of such goods and services, coordinated by transnational corporations, has transformed the world economy into GVCs. It is estimated that between half and two-thirds of international trade now takes place through GVCs. 3 In this article, a contrast is drawn between sustainability provisions that require action from governments and those that require action from transnational corporations and other actors who are directly involved in the activities of GVCs and the products and services they produce.
There is a significant literature study that has studied sustainability provisions in trade agreements. One aspect of this literature has focused on such provisions in World Trade Organization (WTO) Agreements. Most prominently, the general exception clauses in the General Agreement on Tariffs and Trade and General Agreement on Trade in Services are provisions that allow governments to deviate from their other obligations on the basis of specified objectives including those with an environmental and social remit. These clauses have given rise to a wealth of case law whereby governments have sought to justify national initiatives on, inter alia, environmental or social grounds. 4 However, WTO Agreements do not generally include sustainability provisions that impose obligations aimed at actually achieving social and environmental objectives. The attempt to include a social clause in the WTO was one of the most divisive moments in the organization's history. 5 The recently adopted Agreement on Fisheries Subsidies is unique in the WTO context, in that 'it is the first WTO Agreement to place an environmental objective at its core' . 6 As with a number of other issues that have met with resistance in the WTO, this agenda has been picked up in trading arrangements outside of the WTO system. Policymakers across the world have increasingly signed FTAs which include various types of sustainability provisions that have explicitly sought to create environmental and social obligations. Studies of such sustainability provisions have generally identified a number of key facets through which to compare and differentiate different types of provisions, including the degree to which they involve co-operative and coercive aspects (where the US and EU approaches are often contrasted 7 ); their substantive coverage (e.g. labour, environmental, human rights, and gender provisions 8 ); and the organizational structures through which sustainability obligations are considered (governmental committees and, in particular, the civil society bodies which are set up to advise them 9 ).
There is a significant literature on all of these issues. However, what has been far less scrutinized is the nature of the actors for whom sustainability obligations are created in international trade agreements. This is perhaps in part because there is great consistency in the approach, and so drawing contrasts between different models is harder; the vast majority of sustainability provisions create obligations on governments to undertake action on sustainability issues: by signing international treaties, effectively implementing those treaties, co-operating on sustainability initiatives, etc. Sustainability provisions have only rarely attempted to impose obligations directly onto those who are producing and trading goods and services: obligations which affect the 'value chains' which trade agreements govern. But examples of this latter approach have appeared in a number of recent FTAs, discussed later in this article. It is therefore timely to seriously consider the potential of sustainability provisions which create value chain obligations as a means of achieving social and environmental outcomes. 10 In fact, the failure to seriously consider GVCs in mainstream trade law scholarship on sustainability issues can be seen as an important omission. 11 This article makes an argument for why value chain obligations should be taken seriously as mechanisms for addressing sustainability concerns through FTAs, especially when compared with traditional obligations on governments. It argues that such value chain obligations are potentially a more effective and legitimate means of achieving many of the sustainability objectives which are sought through trade policy. The article then scrutinizes existing examples of value chain obligations found in current FTAs: first, palm oil sustainability standards from the Indonesia-EFTA CEPA; second, hen welfare standards on eggs in the EU-Mercosur Association Agreement (EUMAA); third, a stipulation of a minimum average wage for the automobile industry in the US-Mexico-Canada Agreement (USMCA); and fourth, the enforcement of collective bargaining and freedom of association rights directly against factories in priority sectors, also in the USMCA. It argues that all of the examples studied have serious deficiencies from an effectiveness and/or legitimacy perspective.
The article also categorizes these examples into three different modes of governance, namely third-party certification schemes, domestic regulations of one of the parties, and finally bespoke arrangements created for the FTA in question. This allows for the consideration of the future potential of each governance type and the examination of the extent to which each governance type is capable of addressing the effectiveness and legitimacy issues which have undermined existing provisions. As part of this discussion, a broader range of trade-related policy measures (including unilateral trade measures and subject-specific trade agreements) are examined, and the degree to which such measures might complement or replace the sustainability agenda in FTAs with generalized coverage is considered.
Critical to the arguments made in this paper are therefore the concepts of effectiveness and legitimacy. Both of these concepts have a long history of usage in international law and . 10 This contribution builds upon an interdisciplinary project which considered the intersection between labour standards, trade governance, and GVCs to which this author was a contributor. See Smith et al., above n 7. 11 As the former Director-General of the WTO, Pascal Lamy, has stated: ' Any discussion today of international trade and investment policy that fails to acknowledge the centrality of GVCs would be considered outmoded and of questionable relevance', Pascal Lamy, 'Foreword', in Deborah Elms and Patrick Low (eds), Global Value Chains in a Changing World (Geneva: World Trade Organization, 2013) xv-xviii. international relations scholarship. 12 Effectiveness is a concept that has been defined in a variety of different ways when utilized in the context of international legal instruments, including those of international economic law. 13 In the context of the sustainability provisions which this article considers, effectiveness is defined as the degree to which those provisions induce changes in behaviour that improve the state of the underlying sustainability problems associated with those provisions. 14 Defined in this way, effectiveness is a more useful concept than compliance by which to evaluate action taken as a result of the relevant sustainability provisions. Compliance tends to focus on state action and evaluates only whether relevant actors have met the obligations contained in international obligations such as the sustainability provisions of FTAs. 15 It is not concerned with the causality of any change which has occurred. Effectiveness is a concept that is well suited to evaluating action by a wider range of actors (including corporations) and focuses on whether obligations themselves have created changes in behaviour and are therefore playing a significant role in addressing sustainability problems.
Legitimacy is defined here through two aspects-input and output legitimacy-which are widely recognized subcomponents of legitimacy. 16 Input legitimacy refers to the decisionmaking process that led to the provisions coming into force and asks how participatory and representative that decision-making process was. Output legitimacy refers to the outcomes produced by sustainability initiatives, whether they produce broader public benefits, 17 and whether the costs of producing those benefits are shared equitably, with particular account taken of the development status of cost-bearing actors. Taken together, this conceptual framework of effectiveness and legitimacy therefore allows GVC provisions to be interrogated to ascertain the degree to which they are (i) addressing key sustainability concerns and (ii) ensuring the fairness and inclusiveness of both the process by which those provisions are created and the outcomes which they give rise to.
The article proceeds as follows. Section II makes the case for value chain obligations. It argues that the imposition of sustainability obligations through trade agreements directly onto GVCs has potential attractions which may complement and, for some purposes even transcend, the more traditional sustainability obligations which are imposed on governments. Section III then examines four recent efforts to introduce value chain obligations in three different FTAs and evaluates the effectiveness and legitimacy of those provisions. Section IV considers the potential and limitations of three different types of governance models for value chain obligations in FTAs, alongside alternative and complementary trade measures. Section V briefly concludes.

II. THE DE SIRA BILIT Y OF VA LUE CH A IN OBLIG ATIONS
There are a wide variety of different sustainability concerns that trade policymakers seek to address through international trade agreements. Some of these concerns clearly need to be addressed by the parties to the agreements themselves. Most obviously, governments (often under pressure from non-governmental organizations (NGOs) and trade unions) may decide to only conclude trade agreements with other governments who demonstrably share their values across a wide range of sustainability issues from addressing climate change to ensuring that basic human rights are protected. 18 As a result, pre-ratification commitments are sometimes agreed upon to sign up to international conventions and take other action to protect and promote social and environmental outcomes. 19 Once in force, trade agreements also often contain ongoing commitments by governments. For instance, they may commit to signing additional treaties and effectively implementing those to which they are already signatories.
But there are a range of other concerns where sustainability issues are much more directly connected to the trade that occurs between the parties and which FTAs facilitate and augment. For instance, will workers' rights be abused in products produced and services whose trade is increased under the terms of the FTA? 20 Will forests be cut down to make way for soy, beef, or sugarcane which is then imported to trade partners? 21 For these latter types of issues, it is not clear that traditional sustainability obligations, imposed on national-level governments, are the most effective or legitimate policy response. There are a number of rationales why the imposition of direct obligations on the value chain may be a more effective and legitimate way of addressing these types of issues. Three such rationales are provided below.
First, although traditional government-orientated sustainability obligations potentially create an important enabling or constraining environment at the domestic level, they have to go through a significant number of steps if they are to have an impact on the goods and services traded between the parties. For example, the ratification and effective implementation of international treaties will only have a positive impact on environmental or social outcomes in particular value chains if governments can be persuaded to ratify those treaties, implement them domestically, and then enforce them across all workplaces, including those in relevant export industries. Studies have found that these steps have created high barriers to change in the workplaces where goods and services are produced and traded. 22 Drilling down further into those barriers to change, sustainability commitments have sometimes been taken more seriously by governments in the pre-ratification period when the carrot of a potential trade deal can be used as an incentive to induce governments to commit to change. 23 This has led to Conventions being ratified and even in some cases measures taken to strengthen domestic protections, particularly on labour issues. 24 But once trade agreements are signed, then it has proved far harder to drive meaningful action. Government officials from trade partner governments generally see sustainability obligations as externally imposed and not a priority for implementation, while trade officials from the country pushing the sustainability agenda often have a limited conception of their role with regard to sustainability provisions and a limited understanding of domestic political dynamics in trade partners, making it difficult to effectively push for domestic change. 25 Where change has occurred, it has often taken sustained trade union and/or civil society pressure over many years in relation to the most clear and egregious issues and even then change may be restricted to legislative action at the national level. 26 It is therefore fair to say that, overall, sustainability commitments imposed on governments through trade agreements have struggled to be effective in taking the many steps needed to translate into more sustainable products and services being produced in particular GVCs. 27 GVC-orientated provisions are designed to directly target goods traded between the parties, thereby potentially avoiding many of the steps needed to make government measures effective in workplaces which are involved in external trade.
The case for such an approach is strengthened by a second rationale; that commercial obligations of trade agreements have both promoted the value chains through which global trade increasingly occurs and at times been found to create significant social and environmental issues within those value chains. In terms of promotion, more recent 'deep FTAs' have enabled trade in intermediate goods and services to expand and flourish. 28 For instance, rules of origin, which determine how much of a product's total value needs to be produced by parties in the agreement to receive preferential treatment, encourage the proliferation and reorganization of GVCs by creating new incentives for production in markets governed by the FTAs in question. 29 In terms of the environmental and social impacts of FTAs, empirical studies are beginning to uncover varied evidence of problems. For instance, on the environment, a study of FTAs affecting 189 countries between 2001 and 2012 has found 'a large and statistically significant increase in deforestation over the 3 years following the enactment of an [F]TA, which coincides with an increase in agricultural land conversion' . 30 Investigation of the impact of FTAs on workers' rights has been more piecemeal, relying on process tracing to formulate case studies of impacts in individual countries. One study (in which this author was a contributor) found that in Moldova, as tariff barriers were reduced due to a trade agreement negotiated with the EU, exports particularly to the UK and Italy increased. While this stabilized the previously volatile and predominantly female workforce in the apparel sector, pressure from lead firms in the UK and Italy, combined with weaknesses in the labour protection system in Moldova, also led to the entrenchment of poverty wages and heavy reliance on piece rate payments and other troubling overtime practices and production methods. Other case studies, presented in the same study, identified a range of other labour issues caused, at least in part, by commercial obligations of EU FTAs in the context of the Caribbean and South Korea. . 26 One notable case where such action occurred is South Korea. 10 years of trade union and civil society pressure did lead to a case being brought and partially won by the EU. South Korea subsequently ratified three ILO Core Conventions. But the impact of these legislative changes on the protection of workers' rights, generally, and on goods produced for the EU market, in particular, is unclear. Charting the increasing number of cases relating to TSD commitments, including the Korea case, see Geraldo Vidigal, 'Regional Trade Adjudication and the Rise of Sustainability Disputes: Korea -Labour Commitments and Ukraine -Wood Export Bans', 116 The American Journal of International Law 576 (2022). 27  If commercial provisions of FTAs facilitate the flourishing of value chains and cause or exacerbate, at least in some circumstances, social and environmental problems, is it not legitimate for countervailing obligations in those same FTAs to seek to address those social and environmental concerns? The case for this type of intervention is strengthened by the paucity of alternative transnational regulatory instruments which are available to effectively regulate a world of complex and often fragmented value chains. Areas of international law that could provide alternative regulatory fora, such as labour, human rights, and environment, have well-known enforcement problems, while transnational private regulation through instruments like codes of conduct and product certification are limited by the fact that participation by lead firms is voluntary and by the commercial pressures that such firms place on their suppliers which make it difficult for them to adhere to the required standards. 32 A critical strategy for addressing these deficiencies has been to seek to bolster private regulatory efforts with complementary public regulatory initiatives. 33 Trade agreements are widely recognized as one of the most important legally binding instruments that states can use in their external policy-making and so have the capacity to bring meaningful international public law obligations to bear on these critical sustainability issues. 34 A final justification for an approach that focuses directly on GVCs is found in the fact that the sustainability agenda is prone to accusations of disguised protectionism and/or neocolonialism, particularly when it is imposed by developed countries onto developing countries through trade deals. Disguised protectionism refers to the idea that states are seeking to restrict international trade to advantage domestic producers while using the justification of social and environmental concerns to give a veneer of moral credibility to those actions. Neocolonialism refers to the control of less-developed countries by developed countries, not through direct conquest and rule, but rather through indirect means, such as by using their economic leverage. Requiring developing country governments to conform to environmental and social conditions dictated by developed countries risks accusations of both disguised protectionism and neocolonialism. It has undermined the legitimacy of the trade linkage agenda for many years: from debates about the social clause and environmental conditionality in the WTO to concerns about neocolonialism in relation to the environmental conditionality sought by the EU in its trade deal currently being negotiated with Mercosur countries. 35 It is beyond the scope of this article to consider when or how it is morally or politically defensible for a developed country government to impose social or environmental obligations on a developing country government and the extent to which adjustments to those obligations should be made to reflect a country's level of social and economic development. It is also beyond the scope of this article to analyse whether concerns about disguised protectionism or neocolonialism are justified in relation to particular social and environmental conditionalities in specific trade deals. It is sufficient to reflect that obligations imposed directly onto the value chain are less directly affected by such legitimacy concerns. This is because they impose obligations on 32  individual corporate actors who directly benefit from the terms of the FTA rather than whole countries where the democratic legitimacy of the imposition of such measures is far more questionable. It should be noted that there are still very important legitimacy concerns which need to be addressed, including who bears the cost of sustainability initiatives within the value chain, which will be discussed further below. 36 But such concerns appear to be less intractable than for obligations imposed directly on governments.
Overall then, if there is a desire to utilize trade agreements as levers to create more socially and environmentally sustainable trade, then there are serious limitations to relying on provisions that place social and environmental obligations on governments. Such provisions have significant effectiveness and legitimacy deficits; they struggle to be effective in overcoming the many hurdles which exist to creating more sustainable products and services and at the same time their imposition by developed countries onto developing countries raises neocolonialist and protectionist concerns. Provisions that impose social and environmental obligations directly onto value chains appear to be legitimized by the presence of commercial obligations in FTAs which have both promoted the flourishing of those value chains and sometimes been found to create significant social and environmental problems within them. However, there is still a need to scrutinize individual value chain provisions carefully to understand how they actually perform in practice.

III. I M POSIN G OBLIG ATIONS THROUGH FTAS ON THE VA LUE CH A IN: R ECEN T POLIC Y DE VELOPM EN TS
Until recently, the vast majority of FTAs have included very few meaningful provisions directed at social and environmental issues in value chains. The EU generally includes in its Trade and Sustainable Development chapters provisions on corporate social responsibility (CSR) which suggest that the parties individually and/or collectively should take action to promote CSR initiatives. 37 Any activities that have taken place as a result of these vague co-operative commitments have been found to be 'restricted in scope, vigour, and potential future impact' . 38 More specific provisions on sectors such as forest products or marine resources have also tended to take the form of co-operative and information exchange commitments which do not impose any meaningful obligations on commercial actors operating in related value chains. 39 Promotion of corporate social responsibility is also common among FTAs negotiated by Switzerland and more recent FTAs involving the USA, Canada, and Australia. 40 But they all suffer from the same limitations in terms of their inability to provoke meaningful action to make GVCs more sustainable.
The last few years, however, has witnessed the development of sustainability provisions in FTAs which have moved beyond the voluntary promotion of corporate social responsibility by governments and imposed specific obligations on particular value chains with a view to fostering more sustainable trade between the parties. Four recent sets of sustainability provisions in three FTAs have sought to differentiate between sustainable and unsustainable production in particular value chains and reward sustainability and/or penalize unsustainable production. All of these initiatives have arisen as a result of specific controversies faced by the respective trade agreements and respond to them utilizing differentiated governance models to address social and environmental concerns. 36 It is often less powerful actors in the value chain, such as producers in developing countries, who are left to pick up the costs of social and environmental upgrading, while often not reaping the environmental benefits. See Ponte, above n 32, discussing this issue. 37 41 Palm oil production has caused a range of wellpublicized social and environmental impacts including deforestation, displaced communities, air and water pollution, widespread labour rights violations, and destruction of peat lands. 42 A Swiss referendum on the trade deal was very narrowly won (51.6%) only after sustainability provisions on palm oil were included in the FTA. 43 The Trade and Sustainable Development Chapter of the EFTA-Indonesia CEPA requires all vegetable oils and their derivatives to be traded according to the 'laws, policies and practices aiming at protecting primary forests, peatlands, and related ecosystems, halting deforestation, peat drainage and fire clearing in land preparation, reducing air and water pollution, and respecting rights of local and Indigenous communities and workers' . 44 EFTA countries also agree to provide knowledge transfer and financial support to assist Indonesia in establishing inclusive and sustainable production and value chains in the palm oil sector. However, the precise scope of such cooperation is not detailed. 45 In Switzerland, to benefit from preferential tariff treatment, importers of Indonesian palm oil and palm oil derivatives must meet these standards by proving compliance with specific sustainability standards produced by the Roundtable on Sustainable Palm Oil, the International Sustainability and Carbon Certification, or the Palm Oil Innovation Group. 46 Under this governance arrangement, the social and environmental standards which importers must meet in order to obtain preferential tariffs are set by third-party certification bodies. This is the first trade deal to link tariff preferences to a private regulatory initiative which seeks to affect social and environmental outcomes through the use of certification standards.
The second example of value chain sustainability provisions comes from EUMAA. Negotiations between the EU and Mercosur started in 1999, and complex negotiating dynamics on both sides as well as various external events (e.g. the 2008 global financial crisis) have led to an extremely protracted negotiating process lasting more than 20 years. 47 In the EU, domestic agricultural lobbies have been prominent in their opposition to the deal. They have raised concerns about an influx of South American foodstuffs. These foodstuffs, it is argued, do not always meet the same environmental, health, and animal welfare standards as their European counterparts. At the same time, civil society activists have raised a variety of environmental concerns about the agreement. 48 It is in this context that the EU negotiated market access provisions which specify that, in order to benefit from the duty-free access to the EU market, Mercosur egg producers will have to certify they respect EU-equivalent rules for 'laying hen' welfare, as set out in EU Council Directive 1999/74/EC. 49 That Directive lays down minimum standards for the protection of laying hens including the requirement that they have, at a minimum, 'enriched cages' or equivalent alternatives which include a nesting area, perches, litter, and more space for the hens to move around than 'battery' cages. This is the first trade agreement to make market access conditional upon animal welfare requirements. Unlike in the EFTA-Indonesia CEPA, the governance arrangements in this scheme are not set by a third party, but by mandatory internal EU regulations which previously had applied only to domestic EU producers.
The third example concerns the USMCA which replaces the North American Free Trade Agreement (NAFTA). This was another agreement negotiated under pressure from unions and civil society groups, concerned, inter alia, about the impacts of NAFTA on the outsourcing of jobs from the USA to Mexico and the welfare of workers in factories where goods were produced for the US market in Mexico. There are various ways in which the labour provisions of USMCA are strengthened, when compared to the NAFTA model, as well as when compared to labour provisions negotiated in previous US trade agreements with other countries. Discussion of many of these changes is beyond the scope of this article. 50 From a value chain perspective, there are two important innovations. The first relates specifically to the automotive industry which is one of the most high-profile (but by no means the only) sectors where concerns have been raised about NAFTA and its impacts. Responding to this, USMCA requires that at least 40% of the content of specified vehicles must be sourced from high-wage facilities paying an average of $16 per hour, otherwise a 2.5% duty must be paid if those vehicles are then imported into one of the other USMCA countries. This is the first clause in an FTA that makes tariffs conditional on adherence to labour standards and also the first agreement that stipulates a specific wage level for workers. Unlike the palm oil and hen welfare provisions discussed above, this mechanism does not condition market access to requirements from a pre-existing scheme. Rather, the wage requirement is a new scheme whose design is bespoke and was created for the specific purpose of governing labour issues through the USMCA.
All three of the examples cited above concern trade governance initiatives that condition market access to particular value chains (palm oil, eggs, and automobiles) on respect for sustainability standards. USMCA also contains an additional bespoke innovation that is more widely applicable to a range of 'priority' GVCs-ones that manufacture goods, supply services, or involve mining. This is the Rapid Response Labor Mechanism (RRLM) which allows action to be taken when an individual factory or facility violates rights to freedom of association and collective bargaining. Where a violation is found, it allows for suspension of trade in those products from that factory or facility until the rights violation is remedied. This is the first time that an enforcement process in an FTA shifts the focus of attention from failures by government to effectively enforce labour laws and moves it to sites of production where goods are actually produced. 51 All of these provisions are therefore ground-breaking in the different ways in which they impose direct value chain obligations on those actors who seek to trade under the terms of the agreement in question, rather than on the governments of the countries who have signed the FTAs. However, they all have significant effectiveness and/or legitimacy problems. The palm oil, hen welfare, and automotive provisions all have significant limitations in terms of their likely impact on sustainable production. Import quotas of Indonesian palm oil into Switzerland mean that even if the maximum quota of 12,500 tonnes per year was reached, it would only account for 0.03% of Indonesia's palm oil exports. Overall demand for palm oil in Switzerland is stable, and most palm oil imported into the country is already certified as sustainable so there is also little prospect of sustainable palm oil replacing less sustainable alternatives. 52 Furthermore, studies have raised questions about whether certification by the Roundtable on Sustainable Palm Oil (RSPO) provides an improvement in environmental, social, and economic sustainability over non-certified estates. 53 In relation to hen welfare provisions, it has been suggested that Brazil (the main egg exporter from Mercosur) only agreed to them because officials believed that their exports would already meet the obligations contained in the EU Directive, 54 raising questions about whether these value chain commitments will actually improve sustainability in production or merely maintain the status quo. 55 Questions are also raised about the limited coverage of the provisions. They only relate to whole eggs, not egg products, while other more significant agricultural products such as beef and chicken meat (heavily imported from Mercosur countries to the EU) or milk powder (heavily exported from the EU to Mercosur countries) have no sustainability conditions attached to their trade. 56 In the context of concern about how EUMAA will increase deforestation and further entrench industrialized farming practices in the EU and South . 54 Interview with civil society activist (20 October 21) on Microsoft Teams. 55 Although even maintaining the status quo might be considered to have some value as it could prevent future deterioration of production standards. 56 One problem with utilizing sustainability standards on other products, e.g. beef, is that there are no EU wide standards (e.g. the Hilton quotas are tied to specific feeding and grading requirements).
America, hen welfare provisions attached only to whole egg products appears to be very limited in ambition. 57 The automotive provisions in USMCA certainly pertain to a significant industry (although other significant industries could also have been covered). Addressing the pay of workers is a controversial focus for provisions in a trade agreement, particularly as average wages in the Mexican automotive industry (averaging $7.30 in final assembly and $3.40 in supply companies 58 ) appear to be considerably lower than the wage level stipulated in USMCA (an average of $16 dollars per hour). However, the stipulation in USMCA is for an average wage of $16 per hour, rather than a minimum wage at that level. So, higher waged workers may counterbalance their lower paid counterparts and the process for calculating the average wage looks fiendishly difficult, raising questions about whether it will lead to meaningful scrutiny of wage levels in reality. 59 There are also limitations to the application of this clause with respect to US sites of production, raising legitimacy concerns about the one-sided nature of these provisions. 60 The RRLM in USMCA is very different in construction from the other provisions considered above. It does not set standards for sustainable production in value chains. Rather it utilizes existing standards (freedom of association and collective bargaining) which NAFTA has previously sought to enforce against governments and enforces them directly against factories. It has a wider sectoral coverage than the other initiatives considered here, including coverage of significant economic sectors, and so has a wider potential impact. 61 It only covers a narrow range of labour rights issues, which has been identified as a limitation of labour standards provisions in other trade agreements, because of the inability to engage with other critical labour issues affecting particular value chains such as poverty wages and forced overtime. 62 But the aspiration must be that the promotion of stronger and more independent unions will lead to the better protection of other substantive rights in the longer term.
Early invocations of the mechanism have led to the enforcement of the rights to freedom of association and collective bargaining in two automobile factories. 63 The results in these cases have been praised by US officials and trade union representatives. 64 This has led to suggestions that the model will be adopted in future US trade agreements and even by other countries seeking more effective enforcement of labour rights issues. 65 But the mechanism has also been criticized because there is no need to exhaust domestic remedies. This could lead to dangers that the transnational proceedings commenced under USMCA will conflict or interfere with domestic Mexican proceedings. 66 It has also been criticized for opening up Mexican factories to the RRLM enforcement process, while, in effect, shielding US factories from being subject to such claims. 67 At the same time, it is up to US government officials to decide when proceedings against Mexican factories are initiated, leaving scope for the exercise of political discretion and the danger that the system will not be perceived as exercising legitimate authority.
Overall then, all three examples of provisions that create new substantive standards (in relation to palm oil, eggs, and automotives) have significant limits to their effectiveness in relation to the value chains that they seek to influence. Also, each of the initiatives represents a response to a particularly high-profile controversy concerning the trade deal in question. The outcome legitimacy of such measures is therefore open to question in that they appear to seek to assuage public concern about sustainability issues, while not actually taking significant action to address those concerns. The RRLM appears to hold somewhat more promise, at least from an effectiveness perspective. It has a broader coverage and has been actively utilized already against factories which are violating covered rights. But there remain significant legitimacy questions to answer about this model. It is applied to Mexican but not US factories and triggered by US government officials rather than an independent third party. There are also questions about whether a model which focuses only on enforcement is capable of doing the capacity-building work which is needed to assist unsustainable production models to shift to more sustainable forms of production.

IV. THE FU TUR E POTEN TI A L OF VA LUE CH A IN PROVISIONS
As identified in section II, there appear to be significant theoretical benefits of including value chain sustainability provisions in FTAs. But the recent reality of such provisions, as explored in section III, does not live up to this potential. Looking beyond the flaws of these individual initiatives, this section seeks to analyse the governance strategies which underpin them and asks whether there is the potential for more effective and legitimate sustainability provisions using those governance strategies in the future. The potentials of (i) third-party certification schemes, (ii) domestic regulations of one of the parties, and (iii) bespoke arrangements created for a trade agreement are therefore considered in turn below.
Making market access conditional on adherence to pre-existing certification processes and other schemes run by multi-stakeholder initiatives (MSIs) has some potential attractions. Such certification processes and MSIs are present across many key sectors where sustainability issues are prevalent. 68 Utilizing certification processes as the basis for sustainability provisions also avoids the accusation that trade policy is being utilized to push the internal standards of dominant trade powers. Put another way, it strengthens arguments that trade policy is being utilized for normative power rather than market power objectives. For instance, the EU's claim to be wielding normative rather than purely market power through its trade policy is partly premised on the fact that it pushes for the adoption of internationally recognized standards, giving those standards an independent legitimacy. 69 However, concerns about the effectiveness of RSPO in achieving sustainability objectives are replicated across many other certification schemes and MSIs. One recent review found that it was 'firmly embedded in the literature' that such schemes create selective or only marginal positive outcomes for final beneficiaries, while a number of studies have argued that various multi-stakeholder initiatives have created no positive outcomes at all. 70 Some studies point to weaknesses in the standards as being critical to their failings, while others point to failings in the processes by which standards are created and then monitored, calling for more inclusiveness of key stakeholders (e.g. the involvement of local NGOs and/or unions in auditing processes). 71 If certification schemes are to be utilized as a building block for creating more sustainability in trade governance, then attention needs to be paid to the weaknesses of those schemes themselves; how poorly performing schemes can be differentiated from those that are inclusive in design and are achieving significant sustainability outcomes. Consideration could also perhaps be given to how such schemes might be strengthened as a result of their inclusion in trade governance initiatives. 72 The second governance option is to use domestic regulation from one of the parties to the agreement as the basis for sustainability provisions. In reality, it is highly likely to be the domestic regulation of the party pushing for sustainability issues which is utilized, as was the case with the EU Regulation on hen welfare. But if domestic regulation of one of the parties is the basis for standards, this raises questions over whether trade agreements are the right policy instrument for addressing such sustainability issues. Should countries who want to use trade to increase sustainability in value chains instead concentrate on deploying unilateral measures to achieve their policy aims? From an effectiveness perspective, the unilateral approach has the advantage that it potentially covers all of the trade with third parties globally rather than only the trade covered by an individual agreement. In addition, it means that measures are not likely to be watered down during the negotiation process as trade partners look for concessions as part of the price for their agreement to adopt the measures in question. However, from a legitimacy perspective, it also means that measures are not negotiated between the parties but are rather imposed on trade partners. The pros and cons of the two approaches are demonstrated by measures taken in relation to palm oil. While the EFTA-Indonesia CEPA achieved consensus about relatively weak sustainability standards for small volumes of imports, the EU created great hostility and accusations of neocolonialism in palm oil-producing countries by proposing a complete ban on palm oil for use in biofuels. 73 One way of assuaging some of these neocolonialist and protectionist arguments is to target importing companies rather than particular products and to ensure that measures are based on relevant international standards. Such an approach is increasingly prevalent through various 'due diligence' initiatives which put the onus on companies to demonstrate due diligence that their supply chains are sustainable in various respects. For instance, the EU Regulation on Conflict Minerals requires companies to undertake due diligence that what they buy is sourced responsibly and does not contribute to conflict or other related illegal activities. The USA has also signed the US Uyghur Forced Labor Prevention Act which stops businesses importing goods into the USA from China's Xinjiang region, unless they can prove through due diligence processes that their products were not made with forced labour. The EU is currently discussing proposals for regulations on deforestation-free products and the prohibition of products made with forced labour from the EU market. Both of these proposed regulations will involve companies undertaking due diligence of their supply chains. These initiatives are part of a growing momentum among governments to require companies to undertake more generalized human rights due 71 Ibid. 72 One could seek to ensure, for instance, that certification processes are meeting specified criteria (e.g. meaningful complaint mechanisms and monitoring processes) before they are eligible to be used in preferential market access schemes. For an example of an attempt to create a meaningful framework for the comparative analysis of the certification process, see Bureau for the Appraisal of Social Impacts for Citizen information, 'Food Sustainable Labels and Standards: A New Analytical Framework' (September 2021), https://lebasic.com/en/food-sustainable-labels-and-standards-a-new-analytical-framework-to-make-thingsclearer/ (visited 21 May 2022). 73  diligence: from the French Duty of Vigilance Law and German Act on Corporate Due Diligence in Supply Chains to the planned due diligence initiative of the EU. 74 Due diligence obligations therefore look central to many current and upcoming efforts to make cross-border trade in a wide range of goods more sustainable. But it is vital that the methods and results of different types of due diligence obligations are carefully scrutinized to ensure that they are effective in actually transforming sustainability outcomes on the ground. 75 One concern, particularly with more generalized human rights due diligence (HRDD) processes, is that companies themselves are reporting on their own performance, with significant leeway over the modalities about how this is done and limited requirements to publish detailed results. 76 There are also legitimacy concerns about the process by which unilateral due diligence measures are enacted, and the degree to which trade partners are consulted. These are accompanied by concerns about the outcome of due diligence processes. As yet there has been little discussion of who bears the cost of addressing sustainability issues which are identified through due diligence processes. But we know that there is a general tendency for lead firms to push costs onto companies in lower tiers of the supply chain, often in developing countries, who bear the costs of social and environmental upgrading, often without reaping the economic benefits. 77 Relying on unilateral due diligence measures to do the work of ensuring GVC sustainability also breaks the link with trade governance. One of the arguments for sustainability provisions in FTAs is that trade agreements themselves have multiple and complex impacts on GVCs which should be addressed within those trade agreements. Perhaps then, unilateral measures should be seen as complements to, not replacements for, sustainability provisions in trade agreements themselves. Trade agreements could be negotiating spaces whereby trade partners are able to influence how due diligence obligations are interpreted and applied to their own legal and political context. Once we are sure that generalized HRDD processes can be relied upon to distinguish between value chains that are more or less sustainable, provisions in trade agreements could then be utilized to provide preferential market access to companies who demonstrate that due diligence had been carried out and negative social and environmental impacts had been addressed, with the costs of taking action shared fairly through the value chain.
Finally, in the absence of pre-existing initiatives which might form the basis of sustainability provisions, do bespoke arrangements, created for the purposes of the FTA in question, have the potential to address the complexities of supply chain sustainability issues? The examples discussed above from the USMCA have significant limitations, but there are examples of much more detailed and comprehensive efforts to address sustainability issues in specific value chains through trade agreements; not in FTAs with broad coverage but in subject-specific trade agreements which focus on particular sectors. For instance, the US-Cambodia Textile Agreement (USCTA) (1999) set up a framework whereby Cambodian exporters who complied with internationally recognized core labour standards and national labour law were granted increases in market access. 78 Factory-level inspections to ensure compliance were overseen by the International Labour Organization, funded in part by the US government. USCTA was rendered obsolete in 2005 by the WTO agreement to phase out textile and clothing quotas but remains an interesting model. 79 A second more recent example is found in the Forest Law Enforcement, Governance and Trade (FLEGT) scheme, through which the EU has commenced signing Voluntary Partnership Agreements (VPAs) with participating countries. These are legally binding trade agreements between the EU and timber-producing countries outside the EU which seek to ensure that timber and timber products exported to the EU come from legal sources, as well as assisting timber-exporting countries to stop illegal logging by improving regulation and governance of the forest sector. 80 Both of these trade agreements have been found to have limitations and deficiencies. For instance, the effectiveness of USCTA's monitoring mechanism and the independence of decisions on when to grant additional market access were both questioned as was its lack of power to directly address worker complaints. 81 Meanwhile, the length and complications of the FLEGT and VPA process are such that the European Commission had even considered scaling back FLEGT and VPAs and scrapping the export licenses which countries receive to import to the EU (although pressure from NGOs and partner countries for the scheme to continue has meant these proposals have been dropped). 82 Despite these deficiencies, both models have also been praised for having a significant impact on workers' rights and forest governance issues, respectively, in trade partner countries. 83 They also both enhanced their legitimacy by engaging a range of key stakeholders, including in trade partners, in the process of creation and implementation of the measures in question.
What both these initiatives have in common is a specific focus and attention on a particular sector and the design of a sectoral trade agreement to specifically address priority sustainability issues in that sector. The negotiation of such agreements is a very different process to that of a generalized FTA which covers multiple sectors and where sustainability issues are a very small subset of negotiating efforts. Does this perhaps indicate that sustainability issues are best addressed in sectoral agreements where full attention can be paid to the particular dynamics of individual value chains rather than general FTAs where attention to such issues is inevitably going to be more marginalized within bigger negotiating processes? If generalized FTAs are to effectively and legitimately address sustainability issues, then these examples suggest negotiators of such FTAs need to dedicate the same level of time and effort to specific sustainability issues in priority sectors as negotiators of sectoral trade agreements have traditionally done.

V. CONCLUSION
This article has argued that the consideration of 'value chain' sustainability provisions in FTAs has not received the attention it deserves in the academic literature up to this point. With the recent advent of several new initiatives which seek to directly address the value chain through sustainability provisions in FTAs, it is an apposite moment to scrutinize the future potential of this approach.
From both an effectiveness and legitimacy perspective, there appear to be significant theoretical benefits of including GVC-orientated sustainability provisions in FTAs, particularly when compared and contrasted with traditional obligations imposed on governments. The former provisions have greater potential to directly target trade that occurs under the regulatory framework of the FTA, thereby addressing concerns about the sustainability impact of the FTA's commercial provisions. At the same time, the developmental critiques of trade and sustainability linkage, which occur when such linkage targets governmental behaviour, are potentially reduced. But the reality of recent value chain sustainability provisions in FTAs has largely not lived up to their promise; they have either been limited in ambition and so ineffective at having a material impact on sustainability issues in particular GVCs and/or they have been designed and applied in ways that raise fundamental questions about their legitimacy.
But it is still possible to assess the future potential of the governance models which underlie such provisions. Three models have been identified in this article. The first model involves sustainability provisions that utilize third-party certification schemes as the basis for preferential access to markets. The independence of such schemes lends them some legitimacy. But doubts exist about the inclusiveness of many of these schemes and the outcomes they achieve. High-performing certification schemes need to be effectively distinguished from their poorly performing counterparts if such schemes are to be regularly utilized in the future as the basis for GVC-orientated sustainability provisions. The second model involves utilizing domestic regulation from one of the parties. This mode of governance is becoming more important in unilateral trade measures than in FTAs. But unilateral measures such as due diligence initiatives face significant effectiveness and legitimacy concerns. The latter may at least partly be addressed by reconnecting unilateral measures to negotiated trade agreements. Finally, if bespoke provisions are the answer, then those provisions must seriously engage with the specifics of the value chain in question. This approach may be more likely in specialized sectoral trade agreements than in generalized FTAs.
There are therefore a range of possible options open to policymakers who seek to address sustainability issues with measures that directly target the value chain. But all of these options raise effectiveness and legitimacy concerns that must be overcome if future initiatives are to produce valuable results. As new initiatives emerge, they need to be carefully scrutinized to ensure that they are making a significant and equitable contribution to the underlying sustainability issues that they seek to address.