2. Dispute Proceedings under the EU–Korea FTA

The dispute between the EU and Korea related to concerns that the EU had with (a) the inadequate protection for the right to freedom of association in Korea under the TULRAA; and (b) the inadequate efforts by Korea towards ratifying certain fundamental ILO conventions.Footnote ⁵

Specifically, with respect to the first issue, by excluding self-employed persons and dismissed employees from the definition of ‘worker’ under the TULRAA, Korea denied these categories of persons the right to establish and join trade unions, according to the EU. This, the EU alleged, was in violation of Korea's obligation under Article 13.4.3 of the EU–Korea FTAFootnote ⁶ to respect, promote, and realize, in its laws and practices, the principles concerning the fundamental rights, which include the freedom of association and the effective recognition of collective bargaining.

Turning to the second issue, the last sentence of Article 13.4.3 provides that the parties to the FTA shall make ‘continued and sustained efforts towards ratifying the fundamental ILO Conventions as well as the other Conventions that are classified as “up-to-date” by the ILO’. The EU claimed that Korea's efforts towards ratifying some of the fundamental conventions did not meet this standard set out in Article 13.4.3, considering that Korea had not ratified four of these conventions.Footnote ⁷

The EU initiated dispute settlement proceedings under the EU–Korea FTA by invoking the bespoke procedures for matters arising under Chapter 13 (Trade and Sustainable Development), which is distinct from the procedures that apply to all other disputes under the FTA.Footnote ⁸ These custom-made procedures for addressing labour and environmental issues, which are starting to become a regular feature in EU FTAs,Footnote ⁹ provide for a two-step dispute settlement mechanism consisting of consultations and panel review. Where bilateral consultations fail to resolve the underlying issues within 90 days, a Party can request a Panel of Experts to be convened. The panel is stipulated to present its report within 90 days of its composition. Article 13.15.2 brings out the non-binding nature of findings by the Panel of Experts. It provides that the parties ‘shall make their best efforts to accommodate advice or recommendations of the Panel of Experts on the implementation of this Chapter’.Footnote ¹⁰

3. Panel's Finding on Jurisdiction

Before dealing with the substantive issues raised by the EU, the Panel had to first address Korea's challenge regarding the Panel's jurisdiction. Korea raised a threshold argument that the measures fall outside the scope of the FTA and therefore the Panel had no jurisdiction as the dispute did not concern a matter that arises under the EU–Korea FTA. In raising this jurisdictional objection, Korea argued that the EU had not demonstrated that the challenged measures affected ‘trade-related aspects of labour’ within the meaning of Article 13.2.1 of the EU–Korea FTA. Article 13.2.1 provides that the Chapter on Trade and Sustainable Development applies, except as otherwise provided, to measures affecting trade-related aspects of labour. According to Korea, by virtue of this provision, the scope of the commitments under this Chapter was limited to measures that fulfilled this trade-affectation’ test and it was not the intention of the Parties to subject their entire gamut of labour laws and policies, that ‘bear no connection to trade (or investment)’ to the FTA obligations. Korea's argument was that the EU had made no effort to show that the contested measures fulfilled this requirement.Footnote ¹¹

This was indeed true. In fact, the EU had not made any attempt to rebut Korea's claims that the challenged measures affected trade between Korea and the EU. Instead, the EU had based its argument on the phrase ‘except as otherwise provided’ in Article 13.2.1. According to the EU, Article 13.4.3 constituted such an exception and therefore was not subject to the trade-affectation test.Footnote ¹²

The question before the Panel therefore was whether Article 13.4.3 imposing commitments on the parties to enforce labour commitments and to ratify fundamental ILO conventions constituted an exception to the general rule in Article 13.2.1 and therefore was not subject to the trade-affectation test. The Panel answered in the affirmative.Footnote ¹³ In doing so, the Panel placed considerable emphasis on (a) the nature of labour rights covered under Article 13.4.3 and (b) the process of ratification of ILO instruments.

First, according to the Panel the ‘universality’ of the rights covered under Article 13.4.3Footnote ¹⁴ excluded the possibility that the Parties’ commitment to realize these rights domestically was confined to ‘trade-related’ aspects of labour. In other words, a commitment to eliminate ‘all forms of forced or compulsory labour’ clearly includes a commitment to eliminate such labour from all aspects and not just trade-related aspects.Footnote ¹⁵

Second, the Panel took the fact that it is ‘not legally possible for a Party to aim to ratify ILO Conventions only for a segment of their workers’ (i.e., there exists no mechanism for carve-outs for non-trade related sectors while ratifying ILO Conventions) to mean that commitments to ratify ILO conventions necessarily extended to the full scope of the relevant international instruments.Footnote ¹⁶

Based on this, the Panel concluded that Article 13.4.3, imposing commitments on Parties to realize the covered labour rights and to make continued and sustained efforts to ratify the fundamental ILO conventions, constituted an exception to the Article 13.2.1 limiting the scope of Chapter 13 to ‘trade-related aspects of labour’.Footnote ¹⁷ Accordingly, in order to demonstrate that the measures of a Party do not meet the substantive requirements under Article 13.4.3, a complainant does not need to demonstrate the threshold requirement that such measures affect trade-related aspects of labour within the meaning of Article 13.2.1.

The Panel thereafter went on to find that certain aspects of the TURLAA were inconsistent with the principle of freedom of association, as set out in the ILO Declaration, which Korea is obliged to respect, promote, and realize under Article 13.4.3. With respect to the ratification of ILO Conventions, the Panel considered that Article 13.4.3 laid down an obligation of ‘best endeavors’, i.e., one which is higher than undertaking merely minimal steps or none at all and lower than a requirement to explore and mobilize all measures available at all times.Footnote ¹⁸ The Panel considered Korea's efforts for the ratification of the conventions at issue since 2017, which indicated tangible, albeit slow, efforts in the right direction. On this basis, it found that Korea had not acted inconsistently with the last sentence of Article 13.4.3 and therefore had not failed ‘to make continued and sustained efforts’ towards ratification of the core ILO Conventions.Footnote ¹⁹ We limit our analysis below to the implications of the Panel's findings regarding jurisdiction, i.e., that the exercise of its jurisdiction under the FTA is not conditional upon a demonstration of a trade-nexus of the measures at issue.

4. Analysis of the Panel's Finding on Jurisdiction

The normative implications of the Panel's finding on jurisdiction could potentially be greater than what is immediately apparent. However, before considering these implications, it could be fruitful to consider the Panel's reasoning in arriving at this finding. Indeed, the precedent-setting or law-making effect of a judicial decision has been found to rest not so lightly on the soundness and persuasiveness of the reasoning of the decision.

The Panel's interpretive exercise here leaves much to be desired. As noted above, the Panel spent considerable ink clarifying the ‘universality’ of labour rights and the menu du jour method of ratifying ILO conventions to conclude that these rights and Conventions are not, and could not be, circumscribed to trade-related aspects. Yet, the scope of commitments made under the ILO was never at issue under this FTA challenge.

The question before the FTA Panel was a different one – whether there was anything provided in Chapter 13 of the EU–Korea FTA to suggest that measures inconsistent with Article 13.4.3 for violating ILO commitments (as ‘universal’ and cross-cutting as they may be) are covered by the FTA, even when they do not have a trade-nexus. In other words, does a failure to realize a covered (and universally applicable) labour right or to ratify a fundamental (and all-sectors covered) ILO convention fall within the scope of the FTA only when it affects ‘trade-related aspects of labour’ or otherwise? The Panel unfortunately never addressed this question. Instead, it concerned itself with the qualitative aspects of the covered labour rights and ILO Conventions.

When incorporating disciplines or commitments from an agreement or Convention into another agreement, it is not altogether uncommon to incorporate only those commitments that relate to the subject matter of the latter agreement. The WTO Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement), for example, leaves out the substantive obligations on moral rights while incorporating the Berne Convention for the Protection of Literary and Artistic Works, into the TRIPS Agreement. In determining the scope of measures covered under the latter (incorporating) agreement, it would be redundant to examine whether the former agreement has a wide or universal or no-holds-barred approach. Ipso facto, this proves nothing. What is relevant is whether the entire gamut of rights set out in the earlier agreement has been incorporated into the latter agreement or whether any aspects have been left out, whether expressly or otherwise. The Panel assumes, from the universality of the referenced labour rights, that the FTA incorporating those labour rights covers all aspects of labour and not just trade-related aspects. There is nothing ‘provided in [the FTA Chapter]’ that the Panel refers to that legitimizes this assumption. Considering this elemental infirmity in the Panel's understanding of the legal question confronting it, its subsequent analysis of this question charts a deeply disconcerting path.

For instance, it is difficult to square the Panel's reasoning with the ‘Context and Objectives’ of the Chapter on Trade and Sustainable Development, provided in Article 13.1. It provides that the Parties reaffirm their commitment to ‘promoting the development of international trade in such a way as to contribute to the objective of sustainable development and will strive to ensure that this objective is integrated and reflected at every level of their trade relationship’. It further provides that it is not the intention of the Parties to harmonize the labour standards of the Parties, ‘but to strengthen trade relations and co-operation in ways that promote sustainable development …’Footnote ²⁰. Further, the scope provision (Article 13.2) provides that ‘labour standards should not be used for protectionist trade purposes’.Footnote ²¹ The message is clear. Improving labour standards in the parties is not an end-goal in itself for the FTA. Rather, the objective, as negotiated by the parties, is to ensure that trade liberalization does not result in an evisceration of labour conditions or lead to a race to the bottom in labour standards in order to obtain a competitive advantage. There are multiple places where the text of the ChapterFootnote ²² refers to phrases such as an ‘integrated’ approach, and ‘mutually reinforcing’ to make it abundantly clear that labour rights and trade liberalization are joined at the hip and the former are regulated under the FTA to the extent that they relate to trade.

The Panel finds that adhering to ILO standards does not require ‘harmonization of standards’ as it allows regulatory divergence, and cites an OECD report from 1996, as confirmed by more recent findings by the ILO, to conclude that adhering to the fundamental ILO labour standards will not erode competitive advantage of economies and therefore cannot be deemed as being enforced for protectionist purposes.Footnote ²³

This finding appears to unsettle the balance of rights and obligations struck by the FTA. Regardless of the OECD/ILO findings that ILO labour standards do not affect the competitive positioning of an economy, the text of the FTA provides that ‘labour standards should not be used for trade protectionist purposes’ (Article 13.2.2). The provision is predicated upon the very possibility that labour standards could in fact be used for protectionist purposes and therefore Parties must desist from doing so. As interpreters of the treaty, the Panel's mandate is to clarify the meaning of the treaty provisions, which in this context should have been undertaken in a manner harmonizing the text of Article 13.2.2 (non-protectionist use of labour standards) with the obligations in Article 13.4.3 (requiring adherence to ILO labour standards). Instead, the Panel appears to have relied on these studies to essentially render redundant the provisions of Article 13.2.2.

On the procedural side, there is nothing in the panel report to indicate whether these OECD and ILO findings were considered by the Panel suo moto or whether it was presented as evidence by the Parties.Footnote ²⁴ If it was the former, why did the Panel not engage meaningfully with the complete panoply of literature on the subject?Footnote ²⁵ A panel's failure to consider all the evidence and assign appropriate weight to each evidence undermines the objectivity of the panel's assessment of the matter.

5. Implications of the Panel's Finding on Jurisdiction

What could be the normative implications of a finding that measures are reviewable under an FTA's dispute settlement mechanism for their ILO-consistency, not withstanding any demonstration of trade-nexus of such measures?

In essence, it transforms the FTA into a vehicle to enforce ILO commitments and coerce countries into ratifying fundamental ILO conventions. There is no requirement to show that the underlying measures that violated the ILO commitments have a trade-nexus. All that matters is that a labour right has been violated. Any veneer that FTAs were necessarily ‘trade’ agreements, and were concerned with labour (and environment) standards only to the extent that they affected conditions of competition in the market, disappears with this line of reasoning.

Cutting this last cord of objective scrutiny means that we are now veritably in the terrain of cross-disciplinary enforcement of international rules. Institutional mechanisms under the FTA are assessing pure labour issues – whether national labour laws are ILO-consistent – which means that they are interpreting and applying ILO Conventions. It should not be long before FTA panels assess national laws to determine consistency with UNFCCC or other multilateral environment agreements.Footnote ²⁶

Considering the potential for such an expansive (non-trade related) interpretation of commitments under the Sustainable Development or Labour and Environment Standards chapters of FTAs, countries (particularly developing countries) could become more circumspect of negotiating disciplines in these areas. This could mean a setback to any positive advances in labour and environment standards that were being achieved through soft-law (i.e. best endeavor) mechanisms such as knowledge-sharing on best practices. Further, it lends credibility to disparate voices that it was always the intention of the industrialized countries to leverage FTAs as a backdoor to encroach into the sovereign policy space of developing countries in non-trade areas.

This is not to argue against constructive dialogue to foster reforms in Korea's relevant labour laws. Indeed, there has been a global spotlight on inadequate protection for freedom-of-association rights in Korea since the 1990s. Cognizant of these concerns, when inviting Korea to join the OECD in 1996, the OECD Council took note of Korean authorities' commitments ‘to reform existing laws on industrial relations in line with internationally-accepted standards, including those concerning basic rights such as freedom of association and collective bargaining’.Footnote ²⁷ An OECD report from the year 2000 notes that at discussions at the OECD Council and at the ILO on Korea's labour laws, attention has centered on a number of areas, including the right to trade union membership of self-employed and dismissed workers (which is the matter at issue in the extant dispute), where Korean law and practice in industrial relations was believed to be in conflict with international standards.Footnote ²⁸ Further, in a complaint brought by various Korean trade unions to the ILO alleging inter alia that the TULRAA is incompatible with freedom-of-association rights, the Committee on Freedom of Association noted in an interim report that a provision depriving workers of the right to union membership is incompatible with the principles of freedom of association since it deprives the persons concerned of joining the organization of their choice.Footnote ²⁹ It urged the Korean government to take all possible steps to find a solution to this legislative matter.

However, the issue is whether institutional mechanisms under FTAs are adept and equipped to deal with and promote a resolution of such issues regarding labour reforms. Specifically in the context of ILO rights, an ILO study (2014) notes that the ILO never intended for the 1998 Declaration to be interpreted and applied by governments outside of the ILO tripartite environment (consisting of governments, workers, and employees), and prescription of ILO standards through trade agreements is antithetical to the process through which these standards are elaborated and implemented under the ILO framework.Footnote ³⁰ Further, the principles themselves are ‘not sufficiently clear and concrete’ (given that, as the name suggests, they are in the form of principles), making them inadept for a legal evaluation against a domestic labour measure to determine whether such measures are consistent with such principles.Footnote ³¹ Panel scrutiny of ILO principles would also have a bearing on the background and qualification of the panelists and secretariat staff assisting the panelists (if any).

Additionally, are alleged ILO-rights derogations best dealt with through dispute settlement mechanisms? Claussen (2020) argues, based on the learnings from another labour standards dispute under the CAFTA-DR FTA, that an adjudicatory model does not befit resolution of questions on trade-plus issues (including labour standards).Footnote ³²

6. Trade-related Aspects of Labour/Manner affecting trade

Finally, in order to fully understand the magnitude of the implications of this Panel's finding, it appears pertinent to ask the question – by removing the threshold trade-nexus requirement, what measures could now be subject to panel review, that were hitherto outside the scope of the FTA? In other words, what are ‘trade-related aspects of labour’? If this constitutes only a small subset of labour measures, has the Panel in EU–Korea FTA really re-drawn the line by a lot?

According to the EU–Korea FTA Panel, there is total overlap between labour measures and measures affecting and related to trade. This Panel, as a blanket, catch-all obiter, ends its analysis on the jurisdictional question with the following: ‘[i]n the [P]anel's view, national measures implementing [labour] rights are therefore inherently related to trade’. As noted above, the Panel had already found that the measures at issue fall under the exception clause of Article 13.2.1 and need not therefore be ‘affecting trade-related aspects of labour’. However, the Panel went further to note that, in any event, all national labour measures affect trade. In other words, there is complete overlap between labour measures and measures affecting trade-related aspects of labour. Therefore, even if Korean measures were not captured by the exception clause, they would still fall within the Panel's jurisdiction and are therefore reviewable. The redundancy of this finding (given that they were not necessary to resolve the dispute between the parties) aside, this finding does not appear to be consonant with prior WTO and FTA dispute settlement practice.

As a preliminary matter, it is worth acknowledging that it is difficult to conceive of domestic measures that do not affect trade, at least potentially. However, the Panel's above conclusion was not rooted in any analysis of the measures at issue in this dispute. It was based on the Parties’ objective to create a ‘strong connection’ between protection of labour rights and trade. This arguably begs the very question it is seeking to resolve. The fact that Parties intended to create a strong labour-trade connection could mean that only those aspects of labour that impinge upon trade should be regulated by the FTA and therefore require the Panel to determine what those measures are. It does not seem to support the conclusion that all labour measures are ‘inherently’ related to trade.

Outside the context of the EU–Korea FTA Panel, adjudicatory bodies confronted with this question of whether a measure affects or is related to trade have gone about responding to it by examining the measure at issue.

In WTO dispute settlement, panels have typically recognized the wide ambit of ‘measures affecting international trade’, but only after reviewing the actual measures that were at issue in the dispute. For example, the scope provision of the WTO Sanitary and Phytosanitary Agreement (SPS Agreement) provides that the Agreement applies to ‘all sanitary and phytosanitary measures which may, directly or indirectly, affect international trade’. In determining if the challenged measure meets this requirement, panels have traditionally restrained from setting a legal standard for a ‘measure affecting international trade’. Instead, they have restricted their analysis to the confines of the specific measure at issue and gone on to find that trade limiting or prohibiting measures affect international trade.Footnote ³³ One panel has expressly noted that this requirement of ‘affecting trade’ cannot be assumed away: ‘Even if a measure falls within the definition of an SPS measure in Annex A(1) of the SPS Agreement, further to Article 1.1 of the SPS Agreement, such measure still needs to be a measure that directly or indirectly affect[s] international trade to be covered by the disciplines of the SPS Agreement.’ Such an interpretation gives due credit to the treaty drafters’ view that there indeed could be measures out there that do not affect international trade and the measures are therefore subject to review to see if they satisfy this requirement.Footnote ³⁴

Similarly, in determining whether the measures at issue are ‘trade-related investment measures’ within the meaning of Article 1 of the WTO Agreement on Trade-Related Investment Measures (‘TRIMS Agreement’), panels have based their conclusion on the impact of the specific measure at issue on foreign products, such as their impact on imports,Footnote ³⁵ sale and purchase.Footnote ³⁶

The panel in CAFTA-DR, however, has gone one step further and laid out a legal standard for measures affecting trade. The panel found that labour measures ‘affecting trade’ refer to measures that confer ‘some competitive advantage on an employer or employers engaged in trade between the Parties’.Footnote ³⁷

In order to determine if such an advantage exists, the panel assessed the following: (a) whether the enterprise (or enterprises) in question exports to the FTA Parties in competitive markets or competes with imports from FTA Parties; (b) the effects of a failure to enforce labour laws; and (c) whether these effects are sufficient to confer some competitive advantage on such an enterprise or such enterprises. The proxy for a ‘measure affecting trade’ appears to be the impact that the measure has on market competition.Footnote ³⁸ The panel went one step further and added a quantitative element to the meaning of ‘competitive advantage’ – not every failure to enforce labour laws would result in a competitive advantage. Sometimes, the benefits from such a failure could be too brief, too localized, or too small to confer a competitive advantage. Accordingly, the effects appear to meet a de minimis threshold in order to constitute a ‘competitive advantage’, as the panel understood this term.Footnote ³⁹ On the basis of application of this standard, the panel found that the US had not demonstrated that Guatemala's failure to effectively enforce its labour laws was in a manner affecting trade.

Notably, both WTO panels and the CAFTA-DR panel have stressed that there is no need for actual trade effects in ascertaining whether a measure affects trade.Footnote ⁴⁰

The wisdom of associating ‘affecting trade’ with conferring a ‘competitive advantage’ by the CAFTA-DR panel could be called to question. Is it necessary that the failure to enforce labour laws conferred an advantage on the domestic industry (above a de minimis level that also) in order for it to ‘affect’ trade? Where the lack of enforcement concerns sectors or industries in which there is bilateral trade, or (based on the tariff lines that have been committed under the FTA) where there could potentially be bilateral trade between the FTA partners, would the lack of enforcement ‘affect’ trade even if such lack of enforcement had not conferred a competitive advantage to the domestic industry? The panel appears to have substituted the requirement to affect trade for a requirement to affect the conditions of competition in the market.

The US administration however appears to prefer the former. In FTAs concluded post the CAFTA–DR panel report, they have incorporated a clarification by way of a footnote that, ‘for greater certainty, a failure is “in a manner affecting trade or investment between the Parties” if it involves: (i) a person or industry that produces a good or supplies a service traded between the Parties or has an investment in the territory of the Party that has failed to comply with this obligation; or (ii) a person or industry that produces a good or supplies a service that competes in the territory of a Party with a good or a service of another Party’.Footnote ⁴¹

In sum, it appears safe to conclude that there has been no uniformity in addressing the trade-nexus test. Whereas WTO panels have answered this question on a case-by-case basis, the CAFTA-DR panel has applied an impact on ‘competitive advantage test’, and the USMCA has clarified that it prefers an approach which takes into consideration whether there is bilateral trade-relations between the FTA partners in the concerned sector or industry. However, what is crucial to note is that there is nothing inherent to labour measures that establish this relation to trade as suggested by the EU–Korea FTA panel. It is something that needs to be established on a case-by-case basis, whatever the ultimate test or benchmark would be.

7. Conclusions– Where to from Here?

An increasing number of FTAs have started incorporating provisions to protect labour standards in the territories of FTA parties. An ILO database estimates that 113 out of 375 RTAs that were surveyed in 2021 had a labour dimension.Footnote ⁴² Further, as illustrated by the Korea–EU FTA and the CAFTA-DR FTA, FTA parties are also considering enforcing these obligations through dispute settlement mechanisms envisaged under the FTAs. Considering that such labour clauses continue to proliferate and begin to get adjudicated, it is important to seek clarity on the fundamental question of what are the objectives and motivations behind incorporating labour (or sustainability) standards into FTAs? Are they incorporated to protect against the erosion of such standards in order to reinforce a comparative advantage in trade?

In the case of FTAs that recognize that liberalization of trade could exert negative externalities in the labour market and beget a race to the bottom, it is crucial that treaty interpreters and adjudicators appreciate that labour disciplines have been incorporated to remedy such potential market failures that could arise from trade liberalization. In order to adjudicate upon labour measures, therefore, it would be crucial to first demonstrate that there is such a trade-nexus for such measures.

Attempts, however, to de-hyphenate trade and labour linkages and to transform the FTA into a vehicle to require countries into ratcheting up their labour standards could have severe policy implications. As a starting point, it is important to recognize that there is already a fair amount of suspicion, particularly amongst low-income countries, surrounding the need and effectiveness of labour disciplines in FTAs, who view it as a protectionist measure in the guise of humanitarian concerns. Expansive views that adherence to ILO standards is required, even where there is no demonstrable trade-nexus, could further entrench these views and negate any positive advances in labour standards that were being achieved through soft-law (i.e., best endeavor) mechanisms in FTAs such as knowledge-sharing on best practices.

Further, such an expansive approach is predicated upon the assumption that FTAs, and specifically dispute settlement mechanisms within FTAs, constitute the appropriate platform to bring about such labour reforms in countries. However, there appears to be nothing to suggest that FTAs have a relative advantage over other forums exclusively dedicated for such work, such as the ILO.

Establishing a baseline, where only labour measures that have a trade-nexus fall within the scope of an FTA review, is however only the first step. The more difficult next step would be determining the test for measures that have such a trade-nexus. This has to be an FTA-specific approach, given the divergence in treaty language for this standard (e.g., measures affecting tradeand the trade-related aspects of labour). Of course, the adjudicator must first turn to the terms of the treaty itself and consider the ordinary meaning of these terms in their context, and in light of their object and purpose, within the meaning of Article 31 of the Vienna Convention on the Law of Treaties. Express definition of what constitutes such measures in the treaty text itself, as is the case in the USMCA, could go far in removing ambiguities for the interpreter. A case-by-case analysis of whether the specific measure at issue affects or is related to trade, as the case may be, could be preferable to a strait-jacketed approach in this regard.