Labour Standards Provisions in EU Free Trade Agreements: Reflections on the European Commission's Reform Agenda

Abstract Labour standards provisions within the Trade and Sustainable Development (TSD) chapters of EU Free Trade Agreements (FTAs) are presented as a key element of the EU's commitment to a ‘value-based trade agenda’. But criticism of TSD chapters has led the European Commission to commit to improving their implementation and enforcement, creating a critical juncture in the evolution of the EU's trade–labour linkage. This contribution synthesizes findings from academic studies that have examined the effectiveness of labour standards provisions in EU FTAs. It then considers the reform agenda as presented by the European Commission, and explains how some of the proposals could tackle failures identified. However, it also argues that there are various limitations with the Commission's current proposals, and outlines how legal obligations and institutional mechanisms created by trade agreements could better be harnessed to improve working conditions and rights at work around the world.


Introduction
Trade and Sustainable Development (TSD) chapters have been a standard component of the EU's Free Trade Agreements (FTAs) since the initial signing of the EU-South Korea FTA in 2009. TSD chapters, and the labour standards provisions contained within them, have been a critical part of the European Commission's commitment to a 'value-based trade agenda'. 1 In its strategic plan for 2016-2020, DG Trade recognizes that trade policy has come under increased public scrutiny and that part of the response to this must be to promote 'sustainable economic, social and environmental conditions' in the EU and trade partner countries, including through the use of 'strong provisions to promote the respect of labour rights'. 2 But those labour provisions have themselves come in for criticism, particularly from trade unions, civil society organizations, and the European Parliament. 3 In response, the European Commission first instigated a debate about TSD chapters by producing a 'non-paper' on the subject, and then producing a second non-paper in which it sets out 'the way forward on improving the implementation and enforcement of Trade and Sustainable Development chapters'. 4 The authors of this article came together in order to respond to the questions posed by the European Commission in its first non-paper, specifically focusing on the labour provisions within TSD chapters and how they might be reformed in the future. 5 As academics who have all conducted extensive research into the functioning of labour provisions, we felt well placed to review the existing evidence-base about how such provisions have been operating, to explore the limitations and failures of the current model, and to make suggestions in relation to the reform agenda proposed by the European Commission. In undertaking this work, we seek to engage not only with those who are interested specifically in the tradelabour linkage, but also a broader audience concerned with the social dimensions of trade policy. The EU aspires to put social values at the heart of its trade policymaking and its rhetoric suggests that labour standards provisions are critical to this endeavour. But scrutiny of the effect of the EU's approach and related policy implications has not been at the centre of academic debates in this field. Many of the most highly cited journal articles on the topic have instead tended to assess the viability or legitimacy of FTA labour provisions in hypothetical terms and have engaged predominantly with the US position. 6 By filling this lacuna, our intention 3 European Commission, Non-paper of the Commission services Trade and Sustainable Development (TSD) Chapters in EU Free Trade Agreements (FTAs), 11 July 2017, p. 2, http://trade.ec.europa.eu/doclib/ docs/2017/july/tradoc_155686.pdf (accessed 9 October 2017).
4 European Commission, supra n. 3; European Commission, Feedback and Way Forward on Improving the Implementation and Enforcement of Trade and Sustainable Development Chapters in EU Free Trade Agreements, 26 February, 2018, http://trade.ec.europa.eu/doclib/docs/2018/february/ tradoc_156618.pdf (accessed 3 March 2018). A non-paper is a contribution submitted as a basis for discussion, seeking reaction of other parties to possible solutions, without necessarily committing to a public position on the addressed matter. Non-papers are not official documents of an (EU) institution, they are also not published within the context of a formal consultation. The European Commission will also publish later this year its first annual report on the implementation of Free Trade Agreements. This will give details on the implementation of the TSD chapter in each EU FTA. is to provoke more widespread discussion within both the academic and trade policy community about the current practice and future potential of the tradelabour linkage and whether it can contribute significantly to sustainable trade policy, or to use the language of the Commission, to making 'trade for all'. 7 The remainder of this article is structured as follows. In section 2, we set out the key components of the EU's labour standards provisions as contained in the TSD chapters of EU trade agreements and compare and contrast this approach with that of the US. In section 3, we review the academic studies that have examined the effectiveness of the EU's TSD chapters, and identify nine key failures and limitations which are undermining the current approach. In section 4, we examine the European Commission's proposals for reform of the TSD chapters and identify a number of positive ideas that, if properly acted upon, could significantly strengthen the existing model. Alongside this, we identify various limitations and concerns with the current reform agenda. Finally, in section 5, we stress the importance of thinking imaginatively about how the various legal obligations and institutional mechanisms created by trade agreements can best be harnessed to further a labour standards provisions agenda. This involves situating TSD chapters in relation to other aspects of the trade agreements rather than treating them in isolation, and introducing other governance mechanisms for improving working conditions and rights at work.

Labour standards provisions in EU trade agreements
The EU has a long history of including labour standards provisions within its trade policy-making. It has included labour provisions in its unilateral Generalized System of Preferences (GSP) with developing countries since the mid-1990s, first through a sanctioning mechanism (since 1995) and then through special incentives for countries complying with the International Labour Organization (ILO) core labour standards (since 1999). In 2005, the latter was extended to become the GSP+ scheme granting additional tariff preferences to a small number of countries that ratify and implement sustainable development and good governance conventions. 8 In 1999, the EU signed an FTA with South Africa, which for the first time made reference to the ILO core labour standards, closely followed by the Cotonou Agreement in 2000, with the African, Caribbean, and Pacific group of 77 countries committing to these same standards. 9 Following the Treaty of Lisbon adopted in 2007 that accorded greater influence to the European Parliament in trade policy, the level of ambition around the trade-labour linkage has 'significantly deepened and widened'. 10 These ambitions were first acted upon within the 2008 Economic Partnership Agreement (EPA) between the EU and 15 Caribbean states (CARIFORUM), which saw the introduction of new governance procedures and reference made to more social policy norms. Since completing the negotiations for the EU-South Korea FTA in 2009, labour standards provisions have been packaged with rules around environmental protection in Trade and Sustainable Development (TSD) chapters. Such chapters have since featured in finalized agreements with Canada, Colombia/Peru (later joined by Ecuador), Central America (Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, and Panama), Georgia, Moldova, Ukraine, Singapore, and Vietnam. As of March 2018, they were also included in draft negotiating texts in the Transatlantic Trade and Investment Partnership with the US, 11 the Comprehensive Economic Partnership Agreement with Indonesia, 12 and the EU-Mercosur Association Agreement. 13 While there is some variation between the provisions in the different agreements as a result of the negotiation process with individual trading partners, the essential elements of TSD chapters are retained across all recent EU FTAs. This common formulation approach can be defined in terms of the substantive labour standards and procedural commitments relied upon, the institutional structures created, and the way that complaints are handled. 14 In terms of substantive standards, all the agreements involve the parties making commitments in relation to the ILO's eight core labour conventions included in the 1998 Declaration on Fundamental Principles and Rights at Work. 15 These conventions deal with freedom of association and collective bargaining, forced and compulsory labour, child labour, and workplace-related discrimination. Also referenced are commitments expressed in political declarations such as the 2006 Ministerial Declaration of the UN Economic and Social Council on Attainment of Full and Productive Employment and Decent Work for All. These substantive standards are accompanied by a set of procedural commitments, including those on dialogue and cooperation (via the institutional structures mentioned below); monitoring and review of the sustainability impacts of the agreement; a commitment to uphold levels of domestic protection in relation to labour standards; a commitment not to use labour standards for the purposes of disguised protectionism; and a commitment not to weaken or waive laws to encourage trade or investment.
In terms of institutional structures, all TSD chapters include the establishment of a joint committee comprised of representatives of the two parties who will oversee the implementation of the relevant chapter. They also include a civil society mechanism which brings together representatives of business, trade unions, non-governmental organizations (NGOs), andalthough not explicitly mentioned in the text of the TSD chaptersoccasionally academia into Domestic Advisory Groups (DAGs) in each of the trading partners. TSD chapters also facilitate international dialogue between these DAGs and/or other civil society actors of both the EU and its trading partner(s) in a joint Civil Society Forum (CSF).
Complaints concerning the implementation of TSD chapters are not covered by the trade agreement's general dispute settlement mechanism. TSD chapters contain their own dispute resolution mechanism consisting of government consultations and, if necessary, the establishment of a panel of experts. As such, no party can bring an action that would result in the suspension of trade preferences against the other party, that is there is no sanctioning power. 16 It is with respect to this issue of enforcement that the EU's approach is most commonly contrasted with the US model of labour provisions in its FTAs. Complaints around labour provisions in the US agreements are covered by the same chapter on dispute settlement as those concerning other provisions of the trade agreement, meaning that there are potentially significant and symbolically powerful financial or trade-based penalties if violations are found. 17  16 An exception is the EU-CARIFORUM EPA of 2008, which subjects its labour and environmental chapters to regular dispute settlement but excludes the application of trade sanctions for this purpose. This agreement includes some labour and environmental provisions in its investment chapter to which the full sanctions mechanism applies. See Bartels, pp. 301−311, supra n. 14.
also include a more open system for receiving and responding to complaints on labour issues. 18 Largely because of these differences in relation to enforcement, the ILO differentiates between the US and EU approaches to labour provisions. The US approach is termed a conditional one, which links 'compliance with labour standards to economic consequences'. 19 In contrast, the EU approach is categorized as an exclusively promotional approach since its provisions 'do not link compliance to economic consequences but provide a framework for dialogue, cooperation, and/ or monitoring'. 20 Despite these differences, the EU and US approaches also have a great deal in common. Both place great weight on the 1998 ILO Declaration on Fundamental Principles and Rights at Work and identify obligations in relation to core labour standards, 21 both seek to prevent a weakening of labour law to attract investment and increase exports (the 'race to the bottom'), both seek to involve civil society in the negotiation and monitoring of provisions, and both establish dispute settlement procedures involving inter-governmental dialogue and expert panels (although methods of enforcement do differ).
Recognizing these commonalities in aims and approaches is important, since debates about labour provisions in EU FTAs have tended to be based on the assumption that any reform must proceed in either a conditional or promotional direction. As we show later, this false dichotomy is doubly problematic. First, by highlighting the major difference in terms of enforcement, academics as well as policy-makers fail to observe other relevant differences, including on pre-ratification conditionality, institutional capacities, and development financing. Second, the focus on EU−US differences has both restricted appreciation of the way that the EU might learn from the US experience, and drawn attention away from aims and instruments that neither approach currently embodies. 22 Before moving on to investigate effects, it is also important to understand what kind of labour standards the EU's approach seeks to tackle. The most common Settlement in the Trade and Sustainable Development Chapters of EU Trade Agreements, Leuven Centre for Global Governance Studies (2017), https://ghum.kuleuven.be/ggs/publications/books/final-report-9february-def.pdf, (accessed on 26 June 2018), pp. 34, 41.
18 To be considered, the allegations must raise issues relevant to the labour provisions (or specified labour provisions) in the FTA and illustrate a country's failure to comply with its obligations. If the submission meets these criteria, the US Office of Trade and Labor Affairs (OTLA) will accept the submission for review. hypothesis within the academic literature is that EU labour provisions are seeking to have some positive impact on the lives of workers (whether or not they work in internationally traded industries) in countries that are signatories to the relevant trade agreements. 23 But pronouncements by DG Trade Commissioner Malmström, as well as text in recent agreements, suggest that there may also be a second, international dimension; TSD chapters are intended to be the key provisions within EU FTAs for making global supply chains 'more responsible' and therefore are concerned with jobs in specific export-oriented industries, potentially including those outside the jurisdictions of the signatory states. 24 Under this interpretation, TSD chapters are building blocks towards improving the conditions of workers in internationally traded goods and services. A third interpretation is that labour standards provisions are included to address the social impacts of the trade agreement itself. Such a reading is supported by the fact that all TSD chapters contain an obligation to monitor the impact of the trade agreement itself on sustainable development. 25 It is also supported by arguments that the inclusion of more extensive labour standards provisions within recent trade agreements is to counterbalance the significant growth in commercially orientated 'market-creating' provisions in those same agreements. 26 Being clear about the particular pathways through which TSD chapters are intended to have positive effects is key when it comes to evaluating their impacts, as well as deciding upon the reforms to prioritise.

What are the effects of the EU's labour standards provisions?
A number of studies have examined the institutional design and implementation of the EU's TSD chapters, by investigating the functioning of those chapters both with respect to the EU and the EU's trading partners. 27 These studies have found a series of important limitations and failings with the current operation of TSD chapters in a variety of different contexts. We set out nine key findings of these studies below.
First, EU actors who are involved in the negotiation and implementation of TSD chapters view their role as very limited. 28 Opportunities to use sustainability impact assessments and 'pre-ratification conditinality'i.e, requirements for necessary changes in the domestic legislative system of a trade partner before the agreement enters into forceto make significant progress on labour standards in trade partners have been missed. 29 In comparison, the US has utilized pre-ratification leverage more successfully in relation to some of its trade partners. 30 During the negotiations, commitments to more ambitious labour provisions could also have been pursued.
Once agreements are in force, key EU interlocutors often lack detailed knowledge of relevant labour issues in trade partner countries and have not prioritized labour issues in their discussions with trade partner representatives. 31 For instance, a study on the implementation of labour standards provisions under the EU-Peru-Colombia FTA suggested that, on the one hand, commitments on labour standards, as well as civil society dialogue, are not considered a priority for the trade section at the EU Delegation, whereas the cooperation section does have more expertise in this area but is not involved in the implementation of the TSD chapter. 32 While such institutional compartmentalization is inherent to any political system, the separation between different filières is arguably stronger in the EU.
Second, government officials from trading partners with responsibility for engaging with labour issues within the institutions of the TSD chapters often do not see these issues as their responsibility and/or are not the most appropriate representatives for the task in hand. Such officials are often not based in labour ministries or other relevant governmental departments. 33 This poses problems in terms of ownership of the labour agenda by governments who have signed up to obligations contained in trade agreements and for achieving appropriate followup on relevant issues.
Third, civil society mechanisms (CSMs) institutionalized through TSD chapters are hampered by operational failings including: problems of resourcing; lack of meetings and insufficient substantive discussions where meetings do take place; difficulties with obtaining expert and representative members for CSMs; lack of awareness of the existence of CSMs among relevant domestic constituencies; lack of information-sharing on issues relevant to the work of CSMs; lack of independence from government; and inadequate inter-relationships with other bodies institutionalized within the TSD chapter. 34 The overall purposes and functions of CSMs are also not entirely clear: for instance, do they have a monitoring function in relation to the impacts of the labour (as well as wider sustainability) impacts of the trade agreement itself? 35 The expectation that CSMs will cover a wide variety of labour, social, and environmental concerns under the rubric of 'sustainable development' has also made focused discussion difficult, with some studies suggesting that meetings on these topics should be institutionally separated out. 36 Fourth, despite the focus on cooperative activities in the text of the TSD chapters, such provisions have not been systematically implemented through relevant EU instruments. 37 Further, no systematic evaluation of the cooperative activities conducted under the TSD chapters has taken place. There is also the question of policy coherence across EU institutions, for example, the EU's development cooperation arrangements do not systematically address labour issues in the EU's trade partners. 38 An additional problem is that EU aid to upper middle-income countries, such as Colombia and Peru, will phase out, which further undermines the 33  possibilities to strategically link trade and aid policies for the improvement of labour standards. 39 Fifth, the dispute resolution process appears insufficient. As mentioned above, TSD chapters are exempt from the general dispute settlement mechanism of EU FTAs and disputes are instead examined by panels of experts. This process has not yet been activated in any relevant FTA, despite, for example, serious issues being raised in relation to labour violations in South Korea. It is not simply the inadequacies of the legal process itself. The fact that the TSD chapter lacks a credible enforcement mechanism means that it is more difficult to induce compliance with obligations contained in the TSD chapter in the processes of dialogue that take place between the trade partners. 40 The absence of a credible enforcement mechanism also discourages trade unions, who often work with limited staff and resources, to mobilize around possible violations of the TSD chapters.
Sixth, the EU's common formulation approach to TSD chapters (i.e. a focus on the same labour standards pursued via the same processes of dialogue and cooperation in all FTAs) appears ill-equipped to deal with the complexity of labour issues encountered within diverse trading partner scenarios. The ILO core labour standards, which are at the heart of the EU model, are not the most pressing worker-related concerns in all trading partners. For instance, trade-related unemployment in the Caribbean and poverty wages in Moldova have arguably been bigger issues for workers in those locations. Conversely, in South Korea, where core labour standards are a concern, the government crackdown on trade unions in 2015−2016 calls into question the utility of an approach based on dialogue and cooperation. 41 Where individual 'roadmaps' have been produced in addition to the TSD chapters in EU FTAs, there has been insufficient follow-up to ensure compliance. 42 Again, this is in contrast to the US, which has devoted more resources to monitoring follow-up.
Seventh, the provisions contained in TSD chapters regarding the monitoring and assessing of the 'sustainability' impacts of the agreement itself, including on labour standards, have not been properly operationalized. 43 Furthermore, the relevant provisions are vague, leaving the parties a significant amount of leeway with regard to the modalities of monitoring, and there is little evidence that vigorous monitoring has been conducted. Also, no appropriate mechanism is in place to ensure that any identified negative effects of the FTA on labour standards are adequately remedied. As a result, the potential of TSD chapters to ensure that working conditions are not adversely affected by the FTA is significantly reduced. 44 Eighth, despite the formally reciprocal nature of the provisions, there is scant evidence that they have been operationalized in a way that considers labour issues within the EU. 45 This raises questions about whether the EU's model is actually designed to be a two-way process of dialogue, or if it rather represents a form of 'sophisticated unilateralism' wherein more powerful states negotiate provisions that reflect their own unilateral agenda, embedding them within a formally reciprocal structure. 46 Ninth, efforts to extend the reach of labour provisions beyond the trade relationship between the two trade partners and to engage with labour issues in global supply chains are limited. The FTA provisions on such links are vague. Activities have largely focused on encouraging voluntary corporate social responsibility initiatives, which are restricted in scope, vigor, and potential future impact. 47 Overall, these studies have therefore failed to find positive impacts of labour standards provisions for the situation of workers in the EU or its trade partners. Indeed, in two studies it was found that governments had actually sought to weaken labour standards protection (Peru successfully and South Korea unsuccessfully) since the trade agreements with the EU came into force. 48 Given the significant structural problems identified, the findings of these studies also raise serious questions about whether the current EU model has the potential to achieve significant changes to working conditions and rights at work in the longer term.
We should note here that there are a small number of academic studies which argue that there are positive impacts of EU trade agreements on some labour standards, in particular relating to de jure collective labour law. 49 Methodologically, these are based on quantitative analysis of collective labour law and/or Brussels- based interviews with key informants. 50 As such, they are not designed to provide a detailed understanding of how TSD chapters have been operationalized (or not) in third country contexts, and so do not, in any detail, identify strengths or deficiencies within the TSD model itself. They also struggle to engage with important questions about causality, i.e. is it the TSD chapters, the wider trade agreement, or other domestic and international factors which are causing any positive effects that occur for the protection of labour standards in trading partners?
Some of the EU's more recently negotiated agreements do contain some additional content beyond that which is analysed in existing studies. In particular, the EU-Canada Comprehensive Economic and Trade Agreement (CETA) contains additional substantive provisions (for instance on the health and safety of workers) and more detailed requirements on enforcement of labour standards at the domestic level (referencing labour inspection and the judiciary). But overall, it is not a significant departure from the existing approach, in particular concerning the implementation mechanisms which largely reproduce the EU's standard approach in terms of dialogue and cooperation. Moreover, CETA does not come with appropriate provisions to ensure that its regulatory content, e.g. in the area of investor protection or regulatory cooperation, does not adversely affect labour standards. These concerns are also not assuaged by the 'Joint Interpretative Instrument', adopted to facilitate the signing of CETA in October 2016. 51 For these reasons, CETA is likely to be largely subject to the same limitations as earlier EU trade agreements.

The European Commission's proposals for reform
In July 2017, the European Commission published a non-paper which presented two options for reform of its TSD chapters and the labour provisions contained within them. 52 The first option involved enhancing the current processes of dialogue and cooperation contained within existing agreements and being more 'assertive' in terms of using the complaints mechanism and other forms of leverage. The second option was to create a 'model with sanctions'. This focused on the idea of importing a stronger dispute settlement system into the EU model, drawing on the example of the US (as well as the Canadian) approach for inspiration. After a period of dialogue and consultation around these reform proposals, in February 2018, the Commission published a second non-paper, which found in 50 Studies based on large-N datasets of labour legislation can make wider generalisations about de jure labour rights but they cannot speak to the de facto implementation of such rights. favour of the first option on the basis that there was 'a clear consensus that the implementation of TSD chapters should be stepped-up and improved'. At the same time, the Commission made clear its own misgivings about a sanctionsbased model and stated that its consultations had revealed 'divergent points of view' on this issue. In the absence of consensus, it was 'impossible to move to such an approach'. 53 The second non-paper therefore sets out a series of actions to 'revamp the TSD chapters'. 54 One aspect of this agenda is that the Commission promises to work more closely, and communicate better with key stakeholders, including a promise of closer partnerships between the Commission, the European Parliament, EU Member States, and relevant international organizations, including the ILO, to promote the TSD agenda in partner countries. The Commission also commits to more public transparency in relation to the activities of TSD institutions, better communication in relation to progress made on TSD commitments, and to respond to submissions from stakeholders within set time periods. Alongside this, there are a series of actions set out to improve the way that the TSD chapters function, and the outcomes they achieve in trade partner countries. It is worth noting that the proposals for reform are couched entirely in terms of action required in trade partners rather than within EU Member States, meaning that our eighth finding in section 3 above has not been addressed. Putting that issue aside, we identify five important actions proposed by the Commission and explain why they may address some of the criticisms set out in section 3 above. At the same time, we identify important further steps for their implementation and/or limitations in the way actions are currently conceived.
First, the Commission proposes to separately 'identify, consider and address priorities for each partner country' in relation to TSD issues. This involves moving from the current one-size fits all approach. It may lead to the inclusion of specific issues in FTAs beyond the standard TSD formulation, and will involve the identification of priority issues for implementation by trade partners. As highlighted above, labour standards issues are often very different across trade partners, as are the strategies needed for achieving change. This commitment is therefore extremely important. Its effectiveness depends on how it is operationalized. This must involve the development of detailed action plans that focus on the key concerns identified in each trade partnership. As labour standards tend to be sensitive for partner country governments and have not always been a priority for the EU, there is a risk that other TSD chapter issues such as environmental protection or corporate social responsibility initiatives will be prioritized over key labour standards concerns. The content of actions plans should therefore be carefully scrutinized and there must then be concerted follow-up to ensure action plans are acted upon (see fifth point below for more on this). 55 Such action plans can be developed informally in relation to existing agreements, but should be formally mandated in relation to future agreements.
Second, the European Commission proposes to 'encourage early ratification of core international agreements' (including the eight ILO Conventions underpinning the core labour standards) during the negotiation of new trade agreements. This would be coherent with the EU's approach towards GSP beneficiaries, which requires compliance with the ILO core conventions. A shift towards stronger forms of pre-ratification conditionality is important because this is when the EU is in the strongest position to press for legislative change in trade partners. 56 Strong forms of conditionality will increase the potential for relevant labour standards to be incorporated into the domestic law of trade partners. Research suggests the importance of the US's efforts with regard to pre-ratification conditionality in some of its trade negotiations; compared to the EU, the US has been more insistent that certain changes are made to labour law in prospective FTA partners prior to the agreement being signed. 57 There is evidence to suggest that these measures have supported domestic pressure for change in certain countries and are considered one of the strongest forms of leverage that trading partners can exert. 58 At the same time, while getting standards into law is necessary (so that it can be used as a key reference point to contest worker abuses), it is not sufficient to ensure positive improvements in working conditions and rights at work in practice. 59 These latter issues could be tackled during the negotiation phase by requirements for trade partners to act upon the observations made by the ILO's Committee of Experts on the Application of Conventions and Recommendations, so that conventions are not only ratified but also effectively implemented. Effective action is then also needed once trade agreements are in force.
Third, the European Commission makes the important proposal to better enable 'civil society including the Social Partners to play their role in implementation'. 60 As catalogued above, a number of studies have identified serious operational issues with the functioning of Domestic Advisory Groups (DAGs) and Joint Civil Society Forums (CSFs). The Commission addresses this through a €3 million fund to better facilitate civil society activity and will also create clearer guidelines for the functioning of DAGs and CSFs. The commitment by the Commission to respond to all 'written submissions from citizens on TSD [matters] in a structured, transparent, and time-bound way' also means civil society actors have a mechanism by which they can pressure for responses to their concerns. 61 While these actions may have a positive effect on current operational deficiencies, there are also issues that are more fundamental which need to be addressed. These include: (1) recognizing that 'civil society' takes different forms in different trade partners and adapting institutions accordingly; 62 (2) ensuring that civil society actors have rights and resources that are commensurate with their roles and duties, e.g. to allow them to commission studies of issues they identify as requiring further analysis; and (3) creating greater clarity as to what the role of civil society actually is in relation to the TSD chapter and the wider trade agreement. On this last point, the Commission argues that currently DAGs and CSFs are only competent to discuss and advise on implementation of TSD chapters. 63 But in future agreements, DAGs and CSFs will be able 'to cover the implementation of the whole agreement'. 64 This clarification could produce positive results if it means that DAGs and CSFs are empowered to undertake or commission proper monitoring of the 'sustainability' impacts of the FTA as a whole, something that has been missing up to this point. 65 But this broader remit could also take the focus away from addressing labour-related concerns. For instance, in the CARIFORUM−EU EPA, where the civil society mechanism is competent to discuss all economic, social, and environmental aspects of the whole agreement, labour standards issues have never been discussed, in part because of the mechanism's very wide remit. 66 Fourth, the European Commission proposes to 'step up' the resources available for the implementation of the TSD chapter. 67 The Commission has identified how global Aid for Trade funding, and the G7's Vision Zero Fund could be utilized to develop projects to deliver on TSD chapter objectives. The Commission is already making use of its own Partnership Instrument to fund projects connected to the TSD chaptersthe 3 million euro fund to better facilitate civil society activity mentioned above, as well as a 9 million euro project on responsible supply chains. 68 Other development cooperation financial mechanisms could also be utilized in future such as the Development Cooperation Instrument, European Instrument for Democracy and Human Rights, and the European Development Fund.
These funds could be harnessed to develop further projects relating to the TSD chapters. However, two issues should be addressed. First, a distinction should be drawn between, on the one hand, developing projects which contribute to achieving the objectives of the TSD chapters (such as the supply chain project) and, on the other hand, funding necessary to ensure a proper working of the institutions set up by the TSD-chapters (such as the civil society activity project). Relying on ad hoc projects funded by existing funding instruments appears insufficient to address the latter issue. For instance, given the current weakness, even absence, of monitoring in relation to the effects of FTAs on sustainable development (see our point 7 in section 3 above), a dedicated funding stream seems necessary to effectively support monitoring activity in the future. The funds created for better facilitation of civil society activity will certainly not be sufficient to address this issue. The establishment of funds that are specific to each agreement to fulfil the key objectives of TSD chapters therefore appears important. Second, the effects of all projects developed in relation to TSD chapters need to be carefully assessed. It is unclear whether current projects make a significant contribution to the systematic implementation of the TSD chapters. A careful evaluation needs to be conducted to assess results of the projects that take place so as to better understand what can be achieved through this project-based approach in the future.
Fifth, the Commission promises a 'more assertive enforcement' of obligations under the TSD chapter, including increased monitoring of commitments, development of action plans for trade partners where concerns are identified, and triggering of dispute settlement processes where action plans are not followed. 69 Then, if expert panels make recommendations for action, follow up will be undertaken to ensure recommendations are acted upon. The Commission has already started using this more assertive approach with a number of its trade partners. Crucial to demonstrating the effectiveness of its approach will be the ability of the Commission to make significant progress on more difficult labour issues where trade partners are resistant to change. Otherwise, key stakeholders are likely to be sceptical that more concerted use of existing instruments is a sufficient step. 70 The EU model still appears to lack the concrete and tangible economic (dis)incentives to action on labour standards, which have been identified as an important element of being able to be assertive with trade partners. 71 It is notable that the Commission rejected the alternative idea for reform as presented in the non-paper (second option), which focused on the idea of importing stronger sanctions into the EU model by drawing on the example of the US as well as the Canadian approach. As discussed earlier, the US approach can be differentiated in a number of respects from that of the EU, including in terms of the sanctions available and the fact that it has a more open system for receiving and responding to complaints about the violation of labour provisions. Such complaints have been raised by transnational alliances of trade unions and labour NGOs in the US and its trading partners, and have led the US Department of Labor to formally investigate disputes in seven countries to date, in some cases resulting in government-level action plans. 72 It is therefore important to note that cases do not have to result in dispute settlement proceedings for action to take place, and that the credible threat of legal action (and ultimately sanctions) can lead to progress being made on labour issues in trade partners who would otherwise be reluctant to engage. 73 However, as with pre-ratification conditionality, the effects of these interventions should not be overstated. Agreed action plans are not always followed. 74 Moreover, the one case under a US FTA which has proceeded all the way to a decision by the dispute settlement panel found in favour of the respondent. The case concerned trade union rights and other labour matters pertaining to a variety of companies operating in Guatemala. The panel concluded that the US had 'proven that at eight worksites and with respect to 74 workers Guatemala failed to effectively enforce its labor laws … but not that these instances constitute a course of inaction that was in a manner affecting trade'. 75 The difficulty of meeting the standards of proof regarding this provision has led to labour advocates querying whether this creates legal hurdles that are going to be very difficult to overcome in future cases. 76 This case has already led to demands from key US stakeholders for 'beef[ed] up' provisions for enforcing labour standards in future trade agreements. 77 It is therefore questionable whether the US model should have figured so prominently in discussions about reform to the EU enforcement process, reducing the debate about economic (dis)incentives to a question of whether to use statebased sanctions. There are important lessons to learn about both the strengths and weaknesses of the US dispute settlement mechanism. But as identified in the academic literature, there are a number of models for dispute settlement which could have been drawn upon in considering options for reform, including the investment court system, the National Contact Points of the OECD Guidelines for Multinational Enterprises, certain elements of the ILO supervisory machinery, and complaint mechanisms pertaining to voluntary sustainability standards. 78 These models (as well as the US experience) demonstrate the need to move beyond a binary sanctions/non-sanctions debate and to consider a range of complex design issues in making proposals for how a more effective enforcement process could function. These include: how (and by whom) a dispute is initiated; who the complaint targets (corporations could be targeted as well as states); what types of labourrelated allegations could be the subject of a dispute; who investigates the allegations that are made; who adjudicates on any complaints that come to dispute settlement; the nature of the legal test for proving a violation has occurred; and what form of sanctions or fines are available to those who are adjudicating. 79 The opportunity for a more nuanced debate about the optimal design for a dispute settlement process now appears to have been lost. 80 What is crucial moving forward is that the Commission lives up to its commitment to 'continuously analyse the effectiveness of the implementation of the TSD chapters' and to examine whether further measures are necessary to ensure 'full and effective implementation'. 81 Previous processes of monitoring have not been adequately operationalized. 82 This time must be different if key stakeholders are to retain faith in the reform process. This will only be achieved if the Commission is able to demonstrate the ultimate objective set out in the second non-paper: 'real and lasting change on the ground, through the effective application of enhanced social and environmental standards'. 83

Going beyond the European Commission's vision of reform
There are therefore ways in which the European Commission's proposals could be built upon to create meaningful reform of the current TSD chapters. But to maximize the opportunities for EU trade agreements to positively impact upon working conditions and rights at work 'on the ground' we must go beyond the reform ideas that the European Commission has put forward. What is needed is to think imaginatively about how the various legal obligations and institutional mechanisms created by trade agreements can best be harnessed to further a labour standards agenda. Two aspects of this broader vision are set out below.
First, it is critically important to consider the impact of the obligations in the rest of the trade agreement on labour standards. 84 Some provisions may have negative effects. As identified above, it is important to operationalize existing provisions on monitoring the employment and broader social impacts of the agreements as well as to ensure that any adverse effects identified through the monitoring process are effectively addressed. At the same time, there are limits to what even a fully operationalized monitoring process can achieve in terms of identifying problems created by the trade agreement for labour standards, and consideration should be given to the exclusion or restriction of provisions that put labour standards at risk. 85 For instance, careful attention should be paid to investment protection provisions which allow international arbitratorswho may lack knowledge and understanding of labour issuesto make decisions on the basis of investment law obligations, which could have serious direct and indirect impacts on labour standards. 86 Other provisions, if included, could have positive effects. In this respect, a range of obligations in trade agreements and trade-related policies could be utilized to create strong (dis)incentives for action. For instance, this could include more relaxed rules of origin on strategic product lines for companies that demonstrate they have enhanced labour standards protection of particular types; 87 competition rules which specify that abuses of labour standards could be considered as illegal subsidies that would be potentially actionable; a negative list of prohibited labour abuses, perhaps using ILO reporting measures as a trigger, which could be assessed as 'conferring a benefit' in terms of Article 14(a) of the Agreement on Subsidies and Countervailing Measures; and provisions which specify that export credit licenses and other forms of support will only be granted to companies if they demonstrate compliance with certain labour standards. 88 These ideas are presented to make clear that future action to make trade agreements actively work in the interests of improving working conditions and rights at work need not be restricted only to the TSD chapters of EU FTAs. Clearly, they need to be carefully explored to ensure that such measures do not lead to disguised protectionism and that they are enacted in a way that respects other legal obligations (e.g. in relation to World Trade Organization Agreements). More fundamentally, they must have a positive impact on workers' lives and the working population as a whole. It is arguable that one of the explanations for why labour provisions have not been very well implemented and enforced in the past is that trade officials may perceive such provisions as impairing the competitiveness of export industries. What this points to is the need to conduct empirically grounded research on the economic effects of labour provisions in FTAs where they have been linked to improved working rights and at the same time have avoided negative economic effects. 89 87 An example of the EU already using relaxed rules of origin for social purposes, albeit with limited effects so far, is the scheme for Jordanian exporters employing a minimum share of Syrian refugees. See Heliodoro Temprano Arroyo, 'Encouraging the Employment of Refugees through Trade Preferences', Policy Brief of the Immigration Policy Centre, Issue 2017/35 December 2017, http://cadmus.eui.eu/bitstream/handle/1814/49584/PB_2017_35_MPC.pdf?sequence=1&isAllowed=y (accessed 26 June 2018). Given that rules of origin (RoO) are designed to support industries in FTA partner countries by blocking third countries from benefitting from a tariff advantage through simple transhipment, there is the risk that liberalized RoO would result in reduced jobs through trade diversion. However, in practice, RoO can often be highly restrictive, limiting the availability of raw materials or intermediate products available to domestic export-oriented processors, thereby undermining potential employment; as has been well documented in relation to EU preferential RoO on clothing and fish products from the African, Caribbean, and Pacific group. See for example, L. 'trade-based labour standards program in Cambodia that appears to have helped boost employment conditions without jeopardising export growth' − see p. 57. The authors emphasize the need in such schemes to combine labour-related trade measures with strong independent monitoring and domestic policies that Second, there is also a need to think seriously about the kind of labour issues the EU's approach seeks to tackle (see discussion at the end of section 1) and the mechanisms most appropriate for tackling those issues. For instance, if TSD chapters do aim to have an impact on global supply chains, then proponents need to address the supply chain dynamics affecting, among other things, prices paid and delivery times expected, which structure the kinds of working conditions and rights at work that employers are able to provide. How can trade agreements be utilized to tackle the labour abuses arising from these inter-firm power relations? Answers will involve moving beyond commitments to support corporate social responsibility and efforts at sharing best practice in the sector. One mechanism would be to 'establish roadmaps for action in key export sectors with clear monitoring processes aimed at enhancing working conditions relevant to those economic sectors', which would include closer scrutiny on the activities of EU-based firms co-ordinating such supply chains. 90 A second mechanism would be to engage further with voluntary sustainability standards or other certification initiatives which aim to foster compliance with the ILO conventions at the level of the producer and their supply chain. 91 This would build on commitments already included in EU trade agreements. 92 As yet, little concrete action has been taken to promote these voluntary instruments through TSD chapters.
Also key to effective implementation of labour provisions in leading export sectors is a widening of the scope of the provisions themselves. Research on global value chains and labour standards has shown that a focus on ILO core labour standards is important but insufficient to enhance working conditions in promote productivity and fairness, and argue that this 'increases the chances that the trade-linked strategy would work to improve labour standards while minimising risks to employment and export growth' − see p. 81. 92 One can find in almost any agreement signed after 2011 references to the importance of voluntary mechanisms to achieve the objectives of the agreement. For example, Article 271 in the EU−Colombia/Peru mentions that 'The Parties recognise that flexible, voluntary, and incentive-based mechanisms can contribute to coherence between trade practices and the objectives of sustainable development. In this regard, and in accordance with its respective laws and policies, each Party will encourage the development and use of such mechanisms.' Or Article 273c in the same agreement refers to the use of voluntary sustainability standards specifically in the forest sector. Another example is Article 9d in the EU−Vietnam agreement stipulates that 'The Parties recognize that voluntary initiatives can contribute to the achievement and maintenance of high levels of environmental and labour protection and complement domestic regulatory measures. Therefore, each Party, in accordance with its laws or policies, shall encourage the development of and participation in such initiatives, including voluntary sustainable assurance schemes such as fair and ethical trade schemes and eco-labels.' many sectors. Recourse should be also made to a broader set of labour standards − including living wage provisions, occupational health and safety, 93 and hours of work; as well as migrant workers' rightsand give particular attention to key problems such as protecting workers in the informal economy, including through social protection instruments. 94 Overall then, the European Commission's current reform agenda has been constrained by focusing on how the TSD chapters and US equivalents currently operate, and identifying incremental improvements. Rather we should think about the key labour problems that should be prioritized and how the legal obligations and institutional mechanisms within EU trade agreements can be utilized to address them.

Conclusion
In its 'Trade for All' strategy the European Commission has positioned itself as a leader in promoting a 'values-based' model of free trade, which promotes labour standards and sustainable development as it integrates economies. To deliver on this promise and convince the growing chorus of critics that a 'social dimension of globalisation' can be advanced, an effective trade-labour linkage is crucial. Labour standards provisions in TSD chapters are central to the European Commission achieving this objective. We have identified numerous studies examining the functioning of those chapters which have found serious limitations and failings with the current model. The fact that the European Commission has responded to criticism and has created a set of proposals to 'revamp' TSD chapters is therefore to be welcomed. The proposals do address a number of the limitations and failings we identify. But we have also set out important further steps for the proper implementation of these proposals as well as limitations in the way some of the proposed actions are currently conceived. Moving forward, there needs to be careful scrutiny of the way in which the reform agenda is operationalized to ensure it leads to real effects 'on the ground'. At the same time, opportunities for maximizing the potential for trade agreements to really support better working conditions and rights at work across the globe have not been taken. There remains an opportunity to be more imaginative in terms of harnessing the mechanisms available within the wider trade agreement to achieve those loftier objectives.