Towards a stronger EU approach on the trade-labor nexus? The EU-Vietnam Free Trade Agreement, social struggles and labor reforms in Vietnam

Abstract The EU-Vietnam Free Trade Agreement (EVFTA) deviates from the poor track record of ‘trade and sustainable development’ chapters in EU FTAs. Ahead of ratification, Vietnam embarked upon pathbreaking reforms, culminating in a new labor code and accession to outstanding ILO core conventions. This article assesses the role of the EVFTA in these reforms. Building on literatures on the trade-labor nexus and externalization of EU governance, we call for a more comprehensive analysis of power dynamics in partner countries and address this lacunae by embedding FTAs and labor reforms in a strategic-relational conceptualization of states. We argue that the ‘success’ of the EVFTA was the outcome of specific conjunctures of social forces in, and outside of, state institutions in the EU and Vietnam, and their mediation at the transnational level. Amid free trade skepticism in the EU, particular members of the Parliament and the Council wielded their veto powers to negotiate with Vietnam and pull the Commission into a stronger position. In Vietnam, the external pressure resonated with internal struggles and empowered reformists to drive forward labor reforms. Implementation, however, remains uncertain; and, context-dependent as it was, the EVFTA pre-ratification impact does not easily lend itself to replication in other FTAs.


Introduction
Due to widespread debates and protests related to the social impacts of trade liberalization, free trade agreements (FTAs) have increasingly included labor provisions, typically referring to the ILO core labor standards. This is not just important from a quantitative perspective, as a growing share of global trade is governed by such provisions; it is also qualitatively significant, as labor clauses in FTAs represent a rare public governance approach in a regulatory landscape that has, since the and Social Affairs (MOLISA) or other ministries, which proved challenging in the shrinking civic space before and after the VCP Congress in January 2021. To validate our arguments on the domestic aspects of the reform process, we therefore drew on informants with trusted relations to party-state insiders.
Our analysis makes three contributions to the literature on the trade-labor nexus. First, it moves beyond the study of legal texts and institutional frameworks per se, showing how the 'success' of the EVFTA was the outcome of specific conjunctures of socio-political struggles in, and outside of, state institutions in the EU and in Vietnam. In the EU, against the backdrop of growing politicization of, and mobilization against, free trade, particular members of the Parliament, and some member states, leveraged their veto threat to negotiate directly with Vietnamese decision-makers and pull the Commission into a stronger position. This external pressure resonated with struggles over labor reform in Vietnam, where it was encouraged and picked up by a minority faction of reformists, who, in the vacuum left by the US withdrawal from the Trans-Pacific Partnership (TPP), strategically used the EVFTA to boost their influence in the party-state. Hence, second, our analysis brings the partner country to center stage in analyzing the impact of labor clauses. Third, by showing how interaction between EU institutions and Vietnamese authorities was 'brokered' by actors with a 'foot in both camps'the EU's diplomatic delegation, the ILO and the Friedrich-Ebert-Stiftung (FES)our analysis stresses the crucial role of actors with 'trans-local' characteristics in establishing transnational coalitions for pro-labor reform.
Following this introduction, the next section gives an overview of different approaches to labor provisions in FTAs, introduces debates on the externalization of EU governance and lays out our conceptual approach. This is followed by an outline of the development of the EVFTA. The next two sections contain our empirical analysis. First, we present our analysis of the EU's limited efforts on TSD provisions in the negotiation phase and how and why the EU adopted a more assertive approach in the pre-ratification phase, driven by the politicization of the EU's free trade agenda and strategic action by the Parliament and specific member states. Second, after providing an outline of labor reforms in Vietnam since the 1980s, we show how the external pressure from the EU interacted with internal reform drivers, facilitated by actors with 'trans-local' characteristics. The penultimate section gives a preliminary assessment of the implementation of the reforms, and the final section concludes.

FTAs and the trade-labor nexus
After failed attempts to introduce labor provisions at the WTO level in the 1990s, labor clauses have increasingly been integrated into FTAs (ILO & IILS, 2013). In 1994, the North American Free Trade Agreement became the first FTA to regulate labor and environmental aspects. Inspired by the US, the EU included labor standards in its generalized scheme of preferences (GSP) in the mid-1990s and in its FTAs from the late-1990s. The use of labor clauses by the EU has widened and deepened, especially after the Treaty of Lisbon in 2009 elevated trade policy to an exclusive EU competence and granted the European Parliament, an institution that has pushed for the inclusion of labor and human rights concerns, a greater role in trade policymaking (Van den Putte et al., 2015). From the 2011 EU-Korea FTA onwards, labor provisions (supplemented by environmental standards) have been included in so-called TSD chapters, and this approach was aligned and made mandatory with the 2015 Trade for All strategy (Harrison et al. 2019a). TSD provisions have, thus, become an integral part of a new generation of EU FTAs and were, as of February 2022, present in 13 agreements with a total of 23 countries.
The US and EU approaches are often portrayed as opposing paradigms due to their different enforcement mechanisms. The US approachshaped by the Congressional competence over trade policyfollows a conditional strategy, where partner countries must meet certain labor standards before the agreement comes into effect (pre-ratification conditionality) and/or can be sanctioned for violations after implementation, as labor-related complaints are covered by the same dispute settlement procedure as commercial provisions (post-ratification conditionality) (ILO & IILS, 2013). The EU follows a promotional approach. It is wider in scope, as it refers to a broader set of normative instruments, including the ILO core conventions, but also more limited, as the EU has generally refrained from imposing pre-ratification conditions on TSD obligations, nor linked compliance to sanctions. Post-ratification monitoring is assigned to a set of institutional dialogue mechanisms comprising an inter-governmental committee, as well as a civil society mechanism involving 'domestic advisory groups' (DAGs) within each trading partner (representing business, trade unions and NGOs) and a joint dialogue forum to facilitate transnational cooperation. TSD provisions are exempted from the general sanction-based dispute settlement mechanism, but subject to a separate procedure consisting of government consultations and, if necessary, the establishment of a panel of experts (PoE), none of which can apply sanctions .
Scholars have, however, criticized the juxtaposition of the EU and US approaches for overshadowing important similaritiesthe focus on core labor standards, dispute settlement involving inter-governmental dialogue and PoE, civil society participation in monitoring and development programs to support implementationand common limitations ). Particularly, post-ratification conditionality has been of limited effectiveness due to excessive standards of proof (ILO & IILS, 2013;Vogt, 2015). Of the 47 complaints filed under five US FTAs by the end of 2020 (Department of Labor, 2021), the Department of Labor has formally investigated seven, with only onethe case against Guatemala under CAFTA-DRhaving proceeded to a dispute settlement panel. And in this case, the dispute was rejected as the panel judged the failure to enforce labor laws to be neither 'sustained or reoccurring' nor 'in a manner affecting trade' (Congressional Research Service, 2020).
Summarizing studies on the effectiveness of TSD chapters, Harrison et al. (2019a) find no evidence of positive impacts on labor standards in the EU or partner countries and identify a series of weaknesses in practically every aspect of the dialogue-based monitoring and enforcement mechanisms. 1 The EU has in only one case, the EU-South Korea FTAand only after eight years of sustained complaints by trade unions and other civil society actorsrequested formal government consultations (Campling et al., 2021). The failure of these led to the establishment of a PoE, which in January 2021 confirmed that South Korea breached its TSD obligations and recommended that labor laws and practices are adjusted to comply with freedom of association. In April, South Korea ratified three of the four outstanding ILO core conventions, but implementation remains to be seen. Hence, even though the PoE rulings are non-binding, they did in the end, in this case, lead to the ratification of ILO conventions.
In the context of broad critique of the EU's TSD chapters, and pressure from the Parliament and civil society, in July 2017 the Commission published a 'nonpaper' to open a debate on its approach (European Commission, 2017a). After a period of consultation, a 15-point action plan towards a 'revamped ' TSD approach was presented in February 2018(European Commission, 2018. Although the overall tenets remained intact, the Commission promised to 'more assertively enforce' TSD commitments and to 'encourage early ratification of core international agreements using all available tools' (see Harrison et al. 2019b). This recognition of preratification conditionality is, likely, related to the effectiveness of the prospect of market access for gaining concessions from trading partners. As the ILO and IILS (2013, p. 42) conclude, 'the available evidence suggests that addressing labor standards at the pre-ratification stage can contribute to significant improvements, ranging from triggering legislative action to speeding up ongoing legislative proposals in the country concerned'.

Externalization of EU governance
The literature on labor provisions in EU FTAs draws on conceptualizations of the role of the EU in global governance. The concept of the 'regulatory state' (Majone, 1994) captures the ways in which the EU does not simply aim to liberalize trade and investment flows, but to pursue an externalization of its regulatory regimes, including economic and social regulation, which can either liberalize or restrict market activity (Orbie & Khorana, 2015;Smith, 2015). FTAs, in this perspective, are prominent tools for promoting 'behind the border' regulation, covering trade in goods, services and investment, public procurement, protection of intellectual property rights, trade-related 'barriers', and labor and environmental standards.
While various concepts have been used to describe the external power of the EU, and the motives for its governance externalization, two stand out: the EU as a 'market power' (Damro, 2012) and a 'normative power' (Manners, 2002). Damro (2012) perceives the EU as an internal market that is bound to externalize its (economic and social) regulatory frameworks based on material interests. In contrast, the concept of the EU as a 'normative power' focuses on the ethical and ideational dimensions of European foreign policy, asserting the normative distinctiveness of the EU as a global player (Manners, 2002). These concepts are often interpreted as 'norms versus interests'. This, however, neglects their interrelations, as market norms are neither 'objective categories' that 'can be separated from ideas, identities and ideologies' (Orbie & Khorana, 2015, p. 225), nor necessarily 'neoliberal', but often involve market interventions and social regulations (Rosamond, 2014). Rather, the EU's externalization of rules through trade policy responds both to interest-led motivations to open and regulate markets and to normative aspirations arising from the EU as a distinctive social model. But there is a clear asymmetry in how the EU pursues commercial interests through a 'hard' approach, backed by a sanction-based dispute settlement mechanism, and labor rights through a 'soft', dialogue-based approach.
While these points are useful for understanding the ways in which the EU acts in external relations, a weakness of the literature on EU external governance is that it tends to portray partner countries as mere recipients. The substance and mechanisms of governance externalization are derived exclusively from within the EU, while relations to, and dynamics in, partner countries are sidelineda weakness that is equally evident in the literature on the trade-labor nexus Orbie & Khorana, 2015). As Harrison et al. (2019b) argue, countering such 'diffusionist' approaches to external governance, it is important to interrogate the ways in which TSD chapters are translated into, and articulate with, contexts and struggles in partner countries.

State power, social struggles and geopolitical conjunctures
In refocusing on partner countries, our approach to the trade-labor nexus integrates insights from the literatures reviewed above with concepts from critical IPE, understanding FTAs as agreements between states, whose interests, priorities and strategies are formed through socio-political struggles at the national and transnational level, and within particular geopolitical and -economic conjunctures. We draw on the strategic-relational approach of Poulantzas (1978) and Jessop (1990), which conceptualizes the state as an institutional ensemble that is, at once, imbued with distinctive logics, resources and procedures and a site of contestation between socio-political forces. Existing in a dialectical relationship with society, state power is always relational and conditional, shaped by the 'action, reaction and interaction of specific social forces located both within and beyond' state institutions and the 'diverse structural and conjunctural factors that determine their relative weight' (Jessop, 1990, pp. 366, 149). As crystallizations of past struggles, state institutions are inherently biased, privileging certain actors and interests, and internally fragmented, often promoting competing 'state projects'. But (fragile) unity can be forged to the extent that sufficient support can be mobilized behind a 'hegemonic project ' (ibid. pp. 207-211).
The capacity of social groups to influence state power is not formed merely at the national but also the transnational level. Critical IPE has sought to understand the transnationalization of the state (e.g. Brand et al., 2011;Cox, 1993;Robinson, 2004) with two points being particularly important for our task: First, although nation-states remain important, they act within a multi-scalar and increasingly complex spatio-institutional configuration, as social relations have transnationalized and state apparatuses emerged beyond the national scale (Brand et al., 2011). Although a focus has been on how an emergent transnational capitalist class is able to organize and exercise power through transnational state apparatuses, pro-labor forces can alsodespite challengesuse the transnational scale for their projects (Keck & Sikkink, 1998;Merk, 2009). This is facilitated by transnational 'epistemic communities' and 'trans-local' coalitions that can connect actors and struggles across transnational space (Brand et al., 2011). Second, there are considerable power asymmetries between nation-states, linked to their positions in the global economy and within inter-state hierarchies. This is manifest in different constraints in national policymaking and unequal capacities to influence transnational state apparatuses such as international institutions or agreements (Brand et al., 2011). The focus on transnational social relations should, therefore, not overlook the importance of power relations among and within nation-states and their activebut unevenrole in transnationalization processes.
In this vein, we understand FTAs as state-state regulations that are shaped both by struggles and state-society relations at national levels and by asymmetries in inter-state power and the influence of transnational social forces. Determined chiefly by states and social forces in the Global North, FTAs mainly serve to integrate economic spaces for the expansion of capital and global production networks (GPN) by setting rules, regulations and enforcement mechanisms at the transnational scale (Smith, 2015). In so doing, however, they institute a set of economic and, more recently labor, governance structures that influence domestic struggles and reform processes and can be used by social forces to influence state actions .
Within this broader conceptualization of states and FTAs, we highlight five factors that shape the impact of labor clausesin our case the EU's TSD chapterson partner countries. Firstly, the approach pursued towards partner countries is influenced by the political economy contexts and state-society relations of the EU, in particular the relative influence of social forces on the key institutions involved in trade policymaking: the Commission, which negotiates FTAs; the Parliament, which approves FTAs; and member states, which, via the Council, authorize negotiations, approve, and, in the case of mixed-agreements, unanimously ratify FTAs (Damro, 2012). These dynamics affect the priorities of the EU (including demands on labor reform), the strategies it pursues (including the question of conditionality) and its commitment (how far it is willing to go in terms of resources and compromises). Whether the EU seeks to exploit its leverage in the pre-ratification phase and/or to make use of enforcement mechanisms in the post-ratification phase hinge on these questions, no matter which approach it formally follows.
Secondly, and correspondingly, the political economy contexts and state-society relations of partner countries influence the priorities, strategies and commitments they bring to the table (Table 1). The external pressure for labor reform is more likely to succeed if it resonates with internal social struggles and if the push for reform emanating from the FTA is picked up by domestic actors with at least some capacity to influence state action. For this to happen, the labor issues raised by the FTA, and the strategies pursued by EU actors pushing for labor reform, must be relevant to actors at the local level . These local actors do, however, not necessarily need to be pro-labor, but can have other interests and strategies that, nevertheless, align with the external pressure for labor reform.
A third set of factors relate to the geopolitical and -economic context and perceived interests of states therein. Presumably, the concessions and compromises that each party is ready to make in the field of labor reform and implementation depends on its wider interests in, and expected benefits from, the FTAincluding the commercial interests in gaining preferential access to foreign markets, and its role in geo-political and -economic positioning . Presiding over the largest single market in the world is, in this regard, an effective bargaining chip and a potent basis for coercion (Damro, 2012). How states navigate specific geopolitical and -economic conjunctures, however, is strongly influenced by their domestic political economies and, hence, linked to the first two points.
Fourthly, the nature of transnational linkages connecting actors within the countries in the FTA can decisively shape the outcome of labor clauses. Such transnational connections can be used to trigger a 'boomerang effect' (Keck & Sikkink, 1998), in which reformist actors with otherwise limited voice provoke the intervention of external actors to back their demands on the state. But for such a mechanism, relationships between actors pushing for labor reform in both countries are required. Diplomatic ties and links to international organizations, NGOs or other actors with 'trans-local' characteristics that have a 'foot in both camps' can facilitate such relationships and enable access to key decision-makers, help synchronize expectations and foster mutual trust and understanding.
Finally, throughout the formation and implementation of FTAs, the multi-scalar labor regimes, in which actual work takes place, are critical. Concrete worker outcomes are shaped in a complex interplay of historically constituted workplace relations, regulatory frameworks and political economies at local and national levels, as well as of power dynamics in GPNs Smith et al., 2018). Ultimately, the impact of labor clauses depends on their encounter with these multi-scalar labor regimes. Important in this regard are the broader commercial chapters and implications of FTAs, which deepen integration into GPNs and have a direct bearing on labor rights and working conditions. Summing up, we argue that the negotiation, implementation and impact of labor clauses in FTAs depend on the extent to which (1) the EU has the political will and institutional capacity to effectively impose, monitor and enforce conditions upon partner countries; (2) the external pressure resonates with, supplements and amplifies internal social struggles in partner countries; (3) the FTA is, for whatever domestic, geopolitical and -economic reasons, awarded high priority, opening a space for compromise by actors who would otherwise oppose labor reform; (4) transnational social relations and linkages facilitate cooperation between actors pushing for labor reform; and (5) the content of labor clauses is compatible with, or can change, existing labor regimes in the context of GPNs.
History and content of the EVFTA After 14 rounds of talks starting in June 2012, negotiations of the EVFTA were formally concluded in December 2015. After being split into separate trade and investment agreements related to a European Court of Justice (ECJ) ruling, which stated that the EU does not have the mandate to conclude agreements on certain investment-related issues on its own, the EVFTA was adopted by the Commission in October 2018, approved by the Council in June 2019 and, a week later, signed in Hanoi. After a green light from the Parliament in early 2020from the Committee on International Trade (INTA) in January and the Plenary in February the EVFTA was concluded by the Council in March, ratified by the National Assembly of Vietnam in June and entered into force on 1 August 2020 (Table 1).
In terms of content, the EVFTA entails an almost full and reciprocal liberalization of merchandise trade as well as provisions on non-trade barriers, competition policy and public procurement, among others (Grumiller et al., 2018). As such, it is considered 'the most ambitious and comprehensive FTA that the EU has ever concluded with a developing country' (European Commission, 2016, p. 7). With regard to the TSD chapter, as in other EU FTAs, the parties reaffirm their commitments to the ILO 1998 Declaration, including freedom of association and collective bargaining. The agreement, however, is notably vague, requiring each party to 'make continued and sustained efforts towards ratifying, to the extent it has not yet done so, the fundamental ILO conventions' (Art. 13.4.3a) and to 'effectively implement in its domestic laws and regulations and practice' the content of those conventions (13.4.4). In terms of enforcement, the TSD chapter follows the usual dialogue-based design laid out above.
The negotiations of the EVFTA took place in parallel to talks over the TPP, an FTA between 12 countries along the Pacific Rim, including the US. While the TPP was scrapped with the US withdrawal, soon after President Trump assumed office in January 2017, the groundwork of US negotiators related to the labor chapter is key to understanding the impact of the EVFTA. In this regard, the US negotiated a bilateral 'Consistency Plan', signed in February 2016, detailing the legal-institutional reforms that Vietnam needed to make prior to the entry into force of the TPP. Unlike the EU, hence, the US imposed pre-ratification conditions. Compliance with the most controversial issue, granting Vietnamese workers the right to freely form and join unions of their own choosing, was, nonetheless, given a grace period of five years (Tran et al., 2017). After the US exit, negotiations continued and in March 2018, Vietnam and ten other countries signed the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP), which retained the labor chapter, butwithout the Consistency Planplaced significantly weaker demands on Vietnam.

Economic interests and social biases in the negotiation phase
As argued by van Apeldoorn (2009), the EU has since the 1980s evolved into a form of 'embedded neoliberalism', in which a neoliberal project, premised on internal and external market liberalization and pushed by an increasingly transnational capitalist class, has subsumed elements of two contending visions: neomercantilism and social democracy. As a 'hegemonic project', embedded neoliberalism 'seeks to advance neoliberalism through a strategy of incorporating [ … ] rival projects' for the purpose of ensuring popular support and legitimacy (van Apeldoorn, 2009, p. 22). These contradictions became visible during negotiations of the EVFTA, where the Commission's position was shaped by two opposing coalitions: exporters and import-dependent interests were pushing for market opening, supported by most member states, while import-competing industries, critical civil society organizations and trade unions and a minority of member states (including France, Benelux and Scandinavian states) were more skeptical towards opening and advocated for the inclusion of clauses on human and labor rights (Sicurelli, 2015).
That DG Trade was more receptive to business interests is indicated in its meeting activities. Via an access to information request, we have obtained a list of all meetings at which DG Trade discussed the EVFTA with external stakeholders. Of 144 meetings held between October 2012 and October 2019, 80% were with business actors, only 10% with trade unions and other civil society organizations engaged in labor and social issues, and the remaining 10% included environmental NGOs, government, academia and others. In the negotiation phase, this bias was even greater, with business representing 92% of all meetings and labor/social NGOs representing just 2%. The data shows that the sectors with the greatest access to DG Trade were textile, apparel and footwear (22%), wines and spirits (10%), pharma (8%) and automobiles (7%), all sectors with strong export or import interests in the EVFTA. Sicurelli (2015, p. 238) emphasizes the reluctance of the Commission to include labor standards in the negotiations, downplaying the TSD chapter while being preoccupied with the commercial aspects. The sentiment, according to interviews with diplomats and NGO representatives, was to leave the labor question to US negotiators in the context of TPP. 2 Some of those informants observed that the lack of political will was also reflected in the composition of the EU negotiating team, which included trade economists and lawyers, but no specialists on neither labor rights, specifically, nor sustainability more generally. With little impact in the negotiation phase, the Parliament, however, pushed for binding labor standardsin April 2013, through a parliamentary resolution on the human rights situation in Vietnam (Sicurelli, 2015) and in April 2014, through the adoption of a resolution, demanding a binding and enforceable TSD chapter (European Parliament, 2014). Testifying to the Commission's priorities, it never conducted a human rights impact assessment for the EVFTA but referred to the impact assessment for a previously negotiated (but never concluded) ASEAN FTA, which only superficially touched upon human and labor rightsan omission that the European Ombudsman judged a 'maladministration' (European Ombudsman, 2015).

Contestation and politization of EU trade policy
Over the past decade, however, EU trade policy has become increasingly contested. This trend reached a climax soon after the conclusion of the EVFTA, during negotiations of the Transatlantic Trade and Investment Partnership (TTIP), a far-reaching but deferred FTA between the EU and the US, and the Comprehensive Economic and Trade Agreement (CETA), a similar agreement with Canada. Accused of undermining labor rights, consumer protection and environmental standards, and allowing big business to sue governments through investor-state dispute settlement mechanisms, the agreements provoked widespread opposition from civil society, particularly trade unions and NGOs. A shift in public opinion is discernible in Eurobarometer data, according to which the share of Europeans with negative views on free trade grew from 17% in 2009 to 24% in 2016a tendency that was more pronounced in core member states such as Germany, where it more than doubled (11%-24%), and France, where it nearly doubled (20%-36%). 3 Coupled with the Brexit vote and the election of Trump as US President, both in 2016, and the qualification of national-conservative Le Pen for the second round of the French presidential election in 2017, these developments fueled fears, also within the Commission, of a backlash against the EU's free trade agenda. 4 The contestation of trade policy was channeled through the two democratic pillars of the EU, the Parliament and the Council. The Parliament has, since becoming a co-legislator on trade policy, demonstrated a willingness to use the threat of a vetofrom the burial of the Anti-Counterfeiting Trade Agreement in 2012 ( Van den Putte et al., 2015) to the push for an amended position on investor protection in TTIP andCETA in 2015/16 (Roederer-Rynning, 2017). This threat gained credibility with a weakening of the political majority behind free trade. With the three major groups on the liberal-conservative wing of the Parliamentthe EPP, ECR and ALDE (now Renew)favoring free trade, and both extremes of the spectrum habitually voting against it (GUE/NGL and Greens/EFA on the left; EFDD and ENF on the right), EU FTAs largely depend on the support of the social-democratic group (S&D). However, TTIP and CETA exposed deep divisions within S&D. In the final vote on the latter in February 2017 (passing it by a comfortable majority), 66 of the 175 MEPs from S&D voted nothe lowest level of cohesion among all political groups, and within S&D in any vote on international trade. 5 Similar observations have been made regarding member states in the Council. Although the widening of exclusive EU competence in trade policy formally reduced their influence, ambiguities in the definition of competences enabled some member states to carve out a more assertive role. Several parliaments argued that 'deep and comprehensive' FTAs required ratification by all member states. In October 2016, after the Commission had agreed to present CETA as a 'mixed agreement', the risk of a member state block became concrete, when Belgiumthat needed the consent of all six sub-national parliaments -(temporarily) derailed the deal due to opposition from the parliament of Wallonia (Roederer-Rynning & Kallestrup, 2017). To gain clarity, the Commission in July 2015 asked the ECJ to give its opinion on the allocation of competences, using the EU-Singapore FTA as a test case. The Court stated in May 2017 that while the EU does have exclusive competence in several areas, othersnotably portfolio investment and investorstate dispute settlementfall under shared competences with member states (European Commission, 2017a).
In this context of growing public dissatisfaction and social mobilization, crystallized through the Parliament and the Council, the Commission took steps to improve the legitimacy of FTAs (Orbie et al., 2016). The decision to make TSD chapters mandatory in October 2015 and initiate a public debate on the approach in July 2017, which led to the 15-point action plan towards a more 'assertive' TSD chapter, should be seen in this light. 6 The Parliament and few member states pick up pre-ratification pressure While the EU's turn to a conditional strategy after the EVFTA had been concluded was informed by these broader developments, it can only be understood as the outcome of strategic action by specific members of the Parliament, supported by a few member states, in the vacuum left by the US withdrawal from the TPPand the role of reformists in Vietnam as discussed in the next section. The momentum of the TPP was halted when Trump pulled out in January 2017 (Tran et al., 2017). As director of the ILO in Vietnam, Chang-hee Lee, recalled, 'when the US withdrew from TPP in early 2017, most people in Hanoi thought the reform agenda had died, together with the incentive of US market access'. 7 In this situation, the European Parliament stepped in, pushing the boundaries of its institutional mandate by leveraging the threat of non-consent to engage in direct negotiations with Vietnamese decision-makers and pull the Commission into a stronger bargaining position. What enabled the Parliament to play this expanded role was the uncertainty of majorities behind the EVFTA. This was not just related to the politicization of EU trade policy, discussed above, but also to the fact that a deal with communist Vietnam raised questions from an unusually broad mix of MEPs, including conservatives. 8 Figure 1 provides a chronological overview of key events.
Although the sub-committee on human rights (DROI) was also engaged, it was the involvement of the chairman of INTA, Bernd Lange (S&D), presiding over the first bottleneck in the Parliament's two-step consent procedure, that most decisively drew the EU towards pre-ratification conditionality. In a letter to the Vietnamese government in May 2017, Lange stressed the relevance of the ongoing labor code revision for the EVFTA. In September 2017, he headed a mission to Vietnam. In meetings with the Prime Minister, MOLISA, the National Assembly and VGCL, among others, Lange stressed that Vietnam's ratification of the outstanding ILO core conventions was a precondition for getting the EVFTA through the Parliament. Speaking to the press in Hanoi, he told reporters that human and labor rights are 'really at the heart of the discussion' and 'unless satisfactory solutions are found, the agreement will be in troubled water' (BBC, 2017). From then on, this was the message that Lange consistently communicated to Vietnamese decisionmakers, making the autumn 2017 a critical moment in the EU's shift to a conditional approach. 9 The parliamentary pressure on Vietnam and the Commission continued throughout 2018. Important actions included an INTA meeting in February, where MEPsin the presence of the Vietnamese ambassadorvoiced concerns over the labor rights situation 10 ; a letter to the Vietnamese Prime Minister in June, in which Lange requested a roadmap for the ratification of the outstanding ILO conventions; and a working trip to Hanoi in July, where he reiterated that concrete movement on the labor reform was needed for him 'to sell the deal' (VietnamPlus, 2018). The risk of a parliamentary veto was exposed at a public hearing in INTA in October, attended by EU Trade Commissioner Malmstr€ om and the Vietnamese chief negotiator, where several MEPs from S&D (and even the center-right EPP) joined the call for pre-ratification conditions on labor standards. 11 A few weeks later, the Parliament passed a resolution that urged Vietnam to join all relevant UN human rights treaties and ILO conventions (European Parliament, 2018).
Furthermore, the second veto mechanism in EU trade policymaking was also activated. In the Council, the influence of member states was particularly strong given the 'window of leverage' that emerged when the ECJ in May 2017 clarified the status of the EVFTA as a 'mixed agreement' that needed unanimous ratification by all member states (Russell, 2018). This window closed in June 2018, when the Commission convinced Vietnam to split the EVFTA, despite its promise not to use the 'possibility to split trade agreements to avoid submitting them to national ratification' (European Commission, 2017b). Some member statesmost strongly Spain and Belgiumused this leverage to push for concessions on labor reform. 12 While the opposition of Spain most likely reflected the influence of import-competing interest groups in the textile and apparel industry (Sicurelli, 2015), the Belgian position was aimed at preventing having to block another agreement. Hence, the Walloon parliament stepped up its scrutiny of FTAs in the pipeline, including three hearings on the EVFTA (VietnamPlus, 2017), and the Walloon Prime Minister presented a 'Namur Declaration', calling for the obligatory ratification of key human and labor rights instruments in EU FTAs. 13 Faced with such internal disagreement, according to a diplomat, the Belgian federal government put a reserve on the agreement in the Council, related to the three ILO conventions, and in bilateral relations with Vietnam, diplomats actively pointed to CETA and their pressure in the Council. 14 As the informant remembered, 'we put it in all our bilateral meetings, and it was one of the biggest speaking points for our ministers: "we want you to ratify the three conventions"'.
With the Parliament 'holding a gun to Malmstr€ om's head over Vietnam', as a journalist put it, and tensions in the Council, the Commission responded. In January 2018, Malmstr€ om sent a letter to Hanoi, requesting 'concrete progress' on 'freedom of association and the right to collective bargaining' and warning that the Commission would 'pay close attention to the reform of the labor code' (Politico, 2018). Rather than resisting the pull, the Commission 'hopped on board', as described by an advisor in the European Parliament. 15 The Commission needed to show results, generally in the area of FTAs but especially regarding its increasingly criticized TSD approach. Particularly for staff in the TSD unit of DG Trade, therefore, the EVFTA became a critical case for proving that a 'revamped' TSD approach, as proposed in the 15-point plan, could deliver results, thereby pre-empting demands for a more sanction-based model. In this context, the activation of veto powers in the policymaking apparatus was convenient, as it bolstered their leverage within the Commission and towards the Vietnamese government. 16 Once Vietnam agreed to the split, turning the EVFTA into an 'EU-only' deal that could be voted through the Council by a qualified majority, the Commission in October 2018 submitted the text to the Council. In June 2019, despite a critical letter from 14 MEPs, the Council endorsed the EVFTA, and five days later, Malmstr€ om and the Vietnamese Minister for Industry and Trade signed the deal (Reuters, 2019). With the agreement through the first veto mechanism, civil society and MEPs made a last attempt to gain further concessions. In early November 2019, 15 NGOs urged the Parliament to postpone its consent, the 'last, powerful opportunity to leverage these deals' (Human Rights Watch, 2019). And in mid-December, Lange and other INTA members sent a letter to the Vietnamese Prime Minister, asking for further commitments, including a roadmap for the implementation of the TSD commitments (S&D, 2020). Only after Vietnam had passed a new labor code (November 2019) and submitted a written and binding roadmap (January 2020), as a reaction to the final push by the Parliament, did INTA, in January 2020, pass the EVFTA, paving the way for consent in the Plenary vote the following month. Testifying to the lead role of the Parliament in the EU's turn to de facto pre-ratification conditionality, it was Lange and the S&D shadow rapporteur on the EVFTA Jude Kirton-Darlingnot the Commissionthat negotiated the content of the implementation roadmap with the Vietnamese government. 17 In addition to these concessions from Vietnam, the MEPs obtained a commitment from the Commission to closely monitor implementation of the TSD chapter and promote continued reforms (S&D, 2020). 18

Vietnam: external pressure as an amplifier of internal struggles
The responsive-repressive party-state and labor reform While authoritarian states are often interpreted in monolithic terms, the Vietnamese party-state has long been an arena of negotiation and intermediation between competing interests and factions (Vuving, 2017) and proven responsive to grassroots pressure, in particular if emanating from workers or peasants, the two primary constituencies of, and sources of political legitimacy for, the VCP (Kerkvliet, 2010). The VCP, however, is intolerant when its supremacy is challenged, and demands for fundamental political reformsincluding independent unionshave been suppressed. To capture this, Kerkvliet (2010) argues that Vietnam can best be understood as a 'responsive-repressive party-state'.
Since the launch of the doi moi ('renovation') reforms and the transition to a 'socialist market economy' in the 1980s, Vietnam's economic development model has focused on international integration. Since then, labor reforms have been ongoing, fluctuating with the vicissitudes of internal factionalism, domestic struggles and geopolitical and -economic shifts. An early attempt to carve out a space of autonomy for the VGCL was halted by the events of 1989-90the Tiananmen protests in China and the fall of communism in Central and Eastern Europe, not least Poland where the first independent union was instrumental in bringing down the regime (Schweisshelm & Do, 2018). In the second half of the 2000s, a wildcat strike wave in export sectors reenergized the reform process. Bypassing the formal union, the surge in labor activism challenged the legitimacy of both the VGCL and the VCPand although strikes were mainly concerned with material conditions and forced the government to raise minimum wages, they also became a catalyst for union reform (Do, 2017). In a 2008 landmark directive, the Politburo, the highest authority in the political system, called on enterprise unions to 'become actual representatives to protect workers' legitimate rights and interests' (Politburo, 2008). In the following year, a comprehensive revision of the labor codeafter smaller amendments in 2002 and 2006was initiated.
The making of the new law, which was adopted in 2012, exposed the cleavages within the party-state. While the push for union reform came from the government (especially MOLISA) and reformists in the VCP, the National Assembly and the strike-ridden provinces of the South, it was opposed by conservatives in the Party and the VGCL. When the new law arrived, therefore, it was a compromise between reformist and conservative factions. On the one hand, it introduced quarterly labor-management meetings and for this purpose (only) permitted workers to elect representatives. On the other hand, it gave upper-level VGCL branches the right to represent workers in non-unionized enterprises in collective bargaining and disputes (Schweisshelm & Do, 2018). Although the 2012 labor code is 'frequently interpreted as the introduction of "social dialogue" into the Vietnamese context' (Tran et al., 2017, p. 407), it maintained the monopoly of the VGCL and its subordination to the VCP.

FTAs as reform catalysts and the turn to the EU as external leverage
In the latest round of reforms, the labor clauses of FTAs were instrumental in breaking internal resistance to union reform (Evans, 2020b;Tran et al., 2017). The pre-ratification pressure linked to TPP provided crucial leverage for reformists in Vietnam, as it legitimised open discussions on old taboos related to independent unions, shifted domestic discourses and incentivized policy experimentation (Evans, 2020a). In this context, the VGCL hesitantly embraced the calls for reform, shifting from an outright opponent of independent unions to more proactively engaging in the process. In November 2014, the VGCL declared that it no longer opposed the labor reforms required by the TPP; an announcement that enabled the Vietnamese negotiating team to move forward in the reform discussions with the US. 19 As highlighted by Evans (2020b), this switch was motivated by the VGCL wanting to make union reform an internally-driven process rather than externally imposed.
Three developments further helped reformists convince conservatives that concessions in the field of freedom of association were 'a necessary evil'. First, since 2008 economic growth has slowed down in Vietnam. Reviving the economy was critical, not least due to waning loyalty to socialist ideas and the shift to 'performance legitimacy', based on continuous improvements in living standards, as the key source of legitimacy for the VCP (Hiep, 2012). For this purpose, there has been growing consensus around a state project built on deepening export-orientation and GPN integration; a development strategy, in which FTAs are key components. Second, the growing assertiveness of China, including aggressions in the South China Sea, which culminated in mid-2014 in a six-week naval standoff and anti-Chinese protests across Vietnam (Thayer, 2017), revealed the need to diversify foreign relations. Third, frustrations among employers accumulated over the worker representation provisions under the 2012 labor code, which consigned non-unionized enterprises to negotiate with upper-level VGCL branches and had proven incapable of containing strikes. According to a consultant involved in the process, the idea of independent unions was warmly welcomed by the Vietnamese Chamber of Commerce and Industry (VCCI) as a way of circumventing the statebacked VGCL. 20 These factors paved the way for 'a new reform consensus', convincing many conservatives that the rewards of joining the new FTAsin commercial, geopolitical and-economic, and regime stability termsoutweighed the costs (Evans, 2020a).
The importance of the TPP in pushing ahead union reform is indicated by the decision of the Prime Minister in December 2015 to approve MOLISA's plan for accession to ILO conventions in the period 2016-20, instructing Ministries to bring laws in line with ratified conventions (Prime Minister, 2015). The strongest signal of change, however, came in November 2016 with Resolution 06 by the Politburo, setting out the priorities for the implementation of the new-generation FTAs, including a reform of the VGCL and a revision of the legal framework to allow for independent grassroots unions (Do, 2017;Politburo, 2016). This new direction was expressed in the first draft labor code released by MOLISA in December 2016. The US withdrawal from TPP, however, obstructed the reform process, as evident in the 'erosion of the right to form independent trade unions' in the second draft issued in March 2017 (Tran et al., 2017, p. 410). And in May 2017, the Standing Committee of the National Assembly entirely removed the labor code revision from the law-making agenda (Nhan Dan, 2017). As a Vietnamese labor expert remembered, 'everyone here was so depressedthere was nothing going on, after two years where we had been working like crazy, pushing and hoping and everything'. 21 In this situation, different actors in the trade union and ministries were actively seeking new external support to boost their reform efforts, and, from early 2017 onwards, turned their attention to the EU. A high-ranking EU diplomat remembered how 'we were now left alone with a lot of requests from all sides and felt we had a bigger responsibility than before'. 22 A political advisor to INTA recalled how, during a fieldtrip in January 2017, a Vietnamese official proposed the idea that the EU could take up the role of 'outside catalyst of change'. 23 And a bureaucrat from DG Trade described how, later that year, the EU got asked by some government officials to help garner domestic support. 24 The converging interests among the stakeholders in the EUespecially the INTA committeeand responsible officials on the Vietnamese side led to new momentum and progress.
The chronology of events and our interviews support the argument that the EU's newfound assertiveness was instrumental in reinvigorating the labor reforms. The turning point was the September 2017 visit by Bernd Lange. 25 Two weeks prior, MOLISA delivered an official report to the Social Affairs Committee of the National Assembly, outlining its plan for the labor code revision and a new roadmap for ratification of the three outstanding ILO conventions; a document that, according to an independent evaluation for the ILO, 'restart[ed] the process with renewed and stronger commitment' (ILO, 2019a, p. 10). In December, MOLISA formally reopened the rewriting of the labor code with the observation that the existing law was incompatible with ILO conventions (Government of Vietnam, 2017). And the following summer, the labor code revision returned on the lawmaking agenda (Nhan Dan, 2018). As the ILO director in Vietnam later stressed at a hearing in the European Parliament, the pressure from the EU 'boosted the voice of champions of reform in the system, who have always recognized the freedom of association and collective bargaining for effective labor market functioning in Vietnam. The EVFTA enabled them to gain influence inside the system and make the progress we witnessed this year, in 2019'. 26 This conclusion is supported by informants in Vietnam. One explained how the revision of the labor code restarted 'when we began to receive a lot of pressure from the European Parliament', while another described how 'the EU side pushed strongly for progress, which is why the labor code was passed last year'. 27 It should be stressed that the domestic dynamics underpinning the reform consensus persisted after the US exit from TPP. Economic recovery did not distract from the imperative of continued growth for political legitimacy. Neither did employer fatigue over the inability of the 2012 labor code to prevent strikes disappear, despite a decline in strikes. Additionally, anti-Chinese sentiments flared up again in June 2018, when a draft law on special economic zones aroused fears of dominance by Chinese investors and provoked major protests that forced a government reversal (Vuving, 2019). The urgency of a new labor code also sprang from other sources, including the need to solve longstanding financial problems of the pension system by raising the retirement age. 28 Without discrediting these domestic drivers, we argue that a union reform would not have happened when it did, and the way it did, without the external pressure from the EU. Without fundamentally changing internal dynamics, the EU's stepping-up on pre-ratification conditions tilted the balance in favor of the reformist factionand the EVFTA should, therefore, be seen more as a 'catalyst' of change than a root cause, as has also been argued for the TPP (Chan, 2020;Do, 2017).

Facilitators with 'trans-local' characteristics
The EU's stepping into the TPP vacuum was facilitated by actors with 'trans-local' characteristics, with on-the-ground operations in Vietnam and trusted relations in both the party-state and EU institutions. Watching the reform process stagnate after the removal of US leverage, these actors activated their contacts in the EU and facilitated interaction with key decision-makers in Vietnam. 29 One informant described their role as facilitators, 'translating expectations and interpreting realities', while another saw their function as enabling the parties to 'resonate and echo rather than infringe upon each other', which they did through close coordination and contact on a weeklyand, in certain periods, dailybasis. 30 First, members of the EU's own diplomatic delegation in Vietnam early on anticipated the risk that the EVFTA could be blocked by the Parliament or the Council and had, for some time, advocated a more assertive strategy. Diplomats interested in progress on labor reforms were also worried that ratification would mean an end to the EU's leverage, and that DG Trade and its negotiators, a relatively small team, would shift focus elsewhere and lose sight of implementation. 31 For these reasons, members of the EU delegation, especially the Ambassador, were instrumental in bringing parliamentary delegations to Vietnam seeking, as described by a diplomat, to 'turn the Parliament into some sort of Congress', actively using the risk of a veto as a bargaining instrument. 32 An EU diplomat explained how the delegation 'was using the European Parliament as the main argument against the Vietnamese government, to get leverage and make the Vietnamese move forward on a lot of things. We consistently pointed to the Canadian example, saying that "you shouldn't take anything for granted"'. 33 Second, FES, a German foundation close to the Social Democratic Party and associated with the German labor movementand one of the first international non-profit organizations to open in Vietnam, in 1990had for years been supporting the VGCL in redefining its role in the context of the transition to a market economy. As part of its broader work on international dialogue and development cooperation with a broad range of partners in Vietnam, FES has run various programs with VGCL since the 1990s, including training courses in labor law and legal services as well as the setup of labor service centers around Vietnam (Chan, 2020). FES also organized study trips to South Korea and the EU for representatives of MOLISA, VGCL and Vietnamese civil society organizations (Thu & Schweisshelm, 2020). According to our interviews, FES played a critical role in inviting INTA into the TPP vacuum, putting staff from Lange's office in touch with officials from VGCL and MOLISA at a conference in Hanoi in January 2017. 34 Third, the ILO, in Vietnam since 2003, had been deeply involved in reforms of labor laws and industrial relations, providing technical assistance to Vietnamese authorities on building a legal-institutional framework in line with international labor standards, with a focus on the ILO conventions 87 and 98. Also, VGCL has been working closely with the ILO, which helped organizing collective bargaining training sessions, facilitated contacts with international trade unions, organized conferences and seminars and so on (Chan, 2020). When the US pulled out, our interviews suggest, the ILO in Vietnam was also looking for actors who could replace the US as external driver and shifted focus to the EVFTA. 35 Facilitating access to top decision-makers, the ILO co-organized the high-level meeting in September 2017, where Lange met with leaders from the National Assembly, MOLISA and VGCL, among others (ILO, 2019a). In addition, throughout the preratification phase, the ILO assisted MOLISA in drafting and revising the legal documents needed to satisfy the requirements of the EU. More generally, the ILO played an important role through its status as an international organization. Evans (2020b) argues that Vietnamese leaders would not have accepted demands coming too directly and prescriptively from the EUor the USand that reference to the ILO core labor standards was critical.

Legal gaps, broader criticism and lagging implementation
The 2019 labor code, which entered into force in January 2021, brings substantial changes to the industrial relations framework of Vietnam. The law for the first time grants workers the right to form and join non-VGCL-affiliated WROs at the enterprise level, and to be represented by these in workplace dialogue and collective bargaining. It also provides for better protection against anti-union discrimination and employer interference in union activities (ILO, 2019b). However, despite these improvements, significant gaps remain. Among other things, the new WROs must be authorized by the government (that can also revoke authorization); are excluded from forming or joining federations at sectoral or regional levels; are not represented in tripartite bodies; and are funded solely via membership fees (in contrast to VGCL unions that are co-funded by a 2% union tax on employers). For these reasons, the new WROs face an uneven playing field, necessitating further amendments of the labor code ahead of the planned ratification of ILO convention 87 in 2023.
There is also broader criticism of the labor reforms, viewing them as unable to or never even intended toimprove workers' power. Buckley (2021) argues that independent unions have not been demanded by striking workers, for whom wildcat activism has proven effective, and that 'rather than being a progressive step forward, freedom of association reforms are an attempt by capital to reduce labor militancy ' (2021, p. 80). If the new WROs end up curbing informal activism by coopting workers into institutionalizedbut ineffectivechannels of dispute settlement, they may, indeed, weaken workers' bargaining power. To complicate things, the EVFTA has the core objective of deepening Vietnam's integration into GPNs such as for apparel and electronics, which are widely seen to put labor standards under pressure. And several of the most pressing labor issues in these sectorsexcessive overtime, stressful work regimes and non-livable wagesare not directly addressed by the TSD chapter. But while there is little doubt that the party-state's commitment to freedom of association is reluctant and ambiguous, the reforms do represent a pathbreaking change in the context of Vietnamand for a country with a legacy of incremental reforms, it is too early to dismiss them as a mirage.
As pathbreaking as the labor reforms are on paper, it nonetheless remains uncertain how they will translate into practice. Implementation has so far been lagging: as of February 2022, MOLISA had issued only three of four implementing decrees for the new labor code, with the one regulating the new WROs still unpublished. More than two years since the passing of the new labor code, and 18 months after the entry into force of the EVFTA, therefore, it remains impossible for workers to legally organize independent WROs. 36 Despite limited scope of enforcement, the civil society mechanisms required by the EVFTA could in theory serve to hold the party-state accountable, but their implementation is wanting too. The formation of the DAG in Vietnam was delayed and subject to state interference, and when the composition of the DAG was finally announced in August 2021, it counted just three organizations, including the VCCI and the research-arm of the VGCL, both of which are subordinated to the VCP, and one 'real' NGO (Ministry of Industry and Trade, 2021). In January 2022, after pressure from the EU, it was announced that three additional NGOs had joined the Vietnamese DAG but none of them is engaged in labor issues (Ministry of Industry and Trade, 2022). Such lukewarm implementation is at odds with Vietnam's promises to the EU in the roadmap from December 2019 and does not bode well for the capacity of neither the new WROs nor DAGs in pushing for pro-labor change (Thu & Schweisshelm, 2020).
These challenges remind us that how labor clauses are implemented and play out in practice, just like their negotiation and ratification, rest on evolving conjunctures of social forces and state-society relations. Notably, the labor reforms are being implemented in the context of growing dominance of conservatives within the party-state and intensified state control over society. Since the 2016 Congress, the VCP has moved to more decisively discipline dissenting voices; a trend that was reinforced after the latest Congress in January 2021, where party loyalists consolidated their dominance (Abuza, 2021). This context underscores how important external pressure was for reformist forces in Vietnam in bringing forward labor reforms. But it also shows that with the relaxation of external pressurethe EU no longer holding a veto on the EVFTA and limited prospects for hard enforcement of the TSD chapterthe Vietnamese leadership seems to be re-evaluating the situation, embarking upon the balancing act of seeking compliance with the formal obligations under the agreement while making minimal sacrifices.
Given these developments, the implementation of the TSD provisions will critically depend on the EU, and whether local reformists can also activate EU pressure post-ratification. The European Parliament urged Vietnamese authorities to refrain from interfering in the DAG (European Parliament, 2021) and, in cooperation with the National Assembly of Vietnam, formed a Joint Parliamentary Monitoring Group, an institutional novelty tasked with monitoring the TSD chapter; and the EU DAG has expressed 'deep regret and concern' over the process (European Economic and Social Committee, 2021). But although the van der Leyen Commission has taken steps towards a stronger position on the trade-labor nexus including the appointment of a Chief Trade Enforcement Officer in July 2020, tasked with strengthening the enforcement of TSD commitments (European Commission, 2020), and the announcement of An Open, Sustainable and Assertive Trade Policy in February 2021 (European Commission, 2021)it is doubtful whether there is political will to seriously challenge the backtracking of the partystate. What is clear is that the EU institutions now find themselves in a weaker bargaining position vis-a-vis Vietnamand that the Parliament and the Council, with the loss of veto power, are less well-placed to pull the Commission into a more conditional strategy.

Conclusions
In this article, we assessed the role of the EVFTA in the pathbreaking labor reforms in Vietnam in the run-up to its ratification. We argue that the EVFTA did play a crucial role as an external reform catalyst. In the vacuum left by the US exit from TPP, specific members of the European Parliament, and some member states, used their veto powers to negotiate with Vietnamese decision-makers and pull the Commission into an unusually assertive bargaining position. The impact of the EVFTA, however, can only be understood in its interaction with longstanding struggles over labor reform within Vietnam, where external pressure from the EU was leveraged by a minority faction of reformists, who strategically used it to move forward a reform process under way for years. In this process, a key role was played by actors with 'trans-local' characteristicsthe EU's diplomatic delegation, the ILO and FESwhich connected reformists on the two sides, brokered access to decision-makers and helped building mutual trust.
Our analysis suggests that the impact of the EVFTA was contingent upon the historical conjunctures in which it occurred. Hence, the 'success' of the EVFTA, context-dependent as it was, does not easily lend itself to replication. Despite signs of a stronger approach, it is doubtful whether the EU has the political will to prioritize pre-and also post-ratification conditionality on labor standards in future FTAs. Under the influence of export-oriented and import-dependent factions of European capital, DG Trade is unlikely to allow TSD provisions to obstruct market-opening instruments. The critical question is whether the Parliament is ready to consolidate its expanded role by placing demands on partner countries as a condition for its consent to FTAs. Even in that case, however, our analysis shows that conditions in Vietnam were particularly conducive to EU intervention. In other countries, a similar complementarity between TSD provisions, EU actors' strategies and domestic labor struggles may not be there. Indeed, as Harrison and colleagues (2019, p. 273) sum up previous research, while the EU has hitherto not 'sought to 'aggressively' export labor standards through its trade agreements', 'neither have state officials in trading partners readily imported them'.
Conceptually, our article sought to bring a strategic-relational approach to the state into debates on the trade-labor nexus in FTAs. Moving beyond the analysis of legal texts and institutional frameworks per se, such a perspective contributes to understanding the impact of labor clauses in FTAs by foregrounding state-society relations and social struggles around labor reform in partner countries and their interrelations to struggles within the EU. Our conceptual framework and empirical analysis highlight five factors as critical to the impact of labor clauses: the will to impose conditionality; resonance with domestic social struggles; geopolitical and -economic priorities; bridge-building through trans-local actors; and compatibility with multi-scalar labor regimes in GPNs.