The investor-State dispute settlement provisions of the EU’s proposed new free trade agreements with the US (the Transatlantic Trade and Investment Partnership – TTIP), Canada (the Comprehensive Economic and Trade Agreement – CETA) and Singapore (the EU – Singapore Free Trade Agreement) are receiving considerable coverage in the popular media as well as within academic circles (note the recent EJIL: Talk! posts here and here). However, these agreements include not only investor-State dispute settlement but also inter-State dispute settlement provisions which should be equally interesting to international lawyers. On the one hand these provisions provide incentives to resolve trade disputes involving big trading players bilaterally, outside the WTO’s multilateral system, while on the other hand their provisions also reflect an attempt to address problems of fragmentation within the international legal system.
As with many areas of international law, world trade law has historically been concerned with the dangers of fragmentation. While the WTO may be viewed as a ‘fragment’ of the international legal system, within the WTO there is particular worry over the role of new free trade agreements. Why, when you have a multilateral institution with a comparatively clear set of obligations and relatively effective dispute system, would you conclude free trade agreements which contain substantively similar (though often wider ranging) obligations? The fear here is of what Jagdish Bhagwati called the ‘spaghetti bowl’ – a mass of regional or bilateral agreements concluded without consideration for each other or their implications for trade, potentially increasing costs, regulation and distorting conditions of competition for traders.
The concern is not only economic (i.e. that free trade agreements will undermine the non-discriminatory backbone of the current trade settlement), it is also legal: the ‘spaghetti bowl’ can distort the coherence of a legal system and its attendant expectation of certainty as much as it can distort the conditions of competition in trade. Nonetheless, institutional deadlock at the WTO has led to a number of free trade agreements being concluded globally. Until now, cases which have involved overlapping free trade agreement and WTO obligations have been resolved on the system in question’s own terms: for example, the Argentina v Brazil, Pork Subsidies MERCOSUR tribunal using more detailed WTO provisions to interpret a Decision (para. 57) or the Appellate Body in Mexico – Soft Drinks acknowledging NAFTA obligations but not seeing any basis for adjudicating upon them directly (paras. 54-56).
While some free trade agreements such as NAFTA contain fork-in-the-road provisions (Article 2005:6) or primacy clauses (Article 103:2), many do not and the practice of those that do has been mixed (note, the discussions over the Tuna/Dolphin II dispute between Mexico and the US). The free trade agreements that are of particular interest for this debate take a different approach. The TTIP (yet to be finalised, though EU proposals are available online in an attempt to increase transparency), the EU-Singapore Free Trade Agreement and CETA(both awaiting ratification) all offer a ‘softer’ and more transparent way of trying to accommodate obligations under both the WTO and free trade agreements.
Under these three agreements, the provisions on inter-State dispute settlement (in particular the interpretative obligations of adjudicators) demonstrate sensitivity to the need to accommodate the multilateral elephant in the room. Under the EU’s proposal for Article 20TTIP, the ‘panel shall also take into account relevant interpretations in reports of panels and the Appellate Body adopted by the WTO Dispute Settlement Body’ when interpreting the agreement. CETA Article 14:10 directs similarly that the ‘panel shall also take into account relevant interpretations in reports of Panels and the Appellate Body adopted by the WTO DSB.’ The EU-Singapore Free Trade Agreement is somewhat more limited: Article 15.8 requires that ‘[w]here an obligation under this Agreement is identical to an obligation under the WTO Agreement, the arbitration panel shall take into account any relevant interpretation established in rulings of the WTO Dispute Settlement Body’ (emphasis added).
Superficially this seems like a step forward. If WTO members are going to continue concluding free trade agreements (and it appears that they will), the least they can do is to try to (1) maintain conformity of these agreements with WTO law (specifically Art XXIV GATT and Art V GATS) and (2) introduce some form of meaningful conflict avoidance or coordination mechanism into their free trade agreement that stands some chance of working. If CETA and TTIP panels are to ‘take into account’ WTO interpretations of comparable legal provisions this would seem to aid in the quest for some sort of coherence in world trade law without requiring more demanding and politically contentious primacy claims to be made on the behalf of the WTO system.
The difficulty arises when we take into account the way in which trade law might develop. Let us imagine that the TTIP were successfully concluded and ratified. EU and US disputes over trade obligations arise from their more specific obligations under the TTIP. After all, the bilateral resolution of disputes, especially under a system tailored to the needs of the two parties, can offer many advantages. A number of disputes that previously might have gone to dispute settlement under WTO instead are resolved through the TTIP, and a number of disputes that arise under TTIP relate directly to WTO obligations (the suggested TTIP definition of a subsidy, for example, which is linked to Article 1.1 of the WTO SCM Agreement).
This would already constitute a marked change. Historically a large number of WTO disputes have been transatlantic: genetically modified organisms, beef hormones, aircraft subsidies (on both sides), poultry, steel, bananas and corporate tax arrangements have all been subject to protracted litigation at the WTO, and subsequently influenced its law.
Of course, the WTO resolves disputes between many members, not only the EU and US. If disputes between the EU and US were resolved under the TTIP rather than the WTO, a significant number of disputes would not result in a WTO panel or Appellate Body report. Note that the EU and US are still the most active claimants and respondents at the WTO and inter-US/EU disputes constitute some 36% of their complaints. Canada is the third most active claimant at the WTO (figures are available from the WTO and Horn et al).
While the EU, US and Canada would still engage in litigation with other WTO members, there would exist an increasing body of case-law that develops without their direct participation. One wonders whether either side would be comfortable with panels within the TTIP or CETA taking the lead from WTO panel or Appellate Body reports that arose from China-Mexico or Indonesia-Brazil disputes, for example. Given the importance that the argumentation adopted by the parties can have on the outcomes of disputes, as well as the report itself, it is hard to imagine that they would accept this with equanimity. Reduced participation ought not matter if we consider the WTO a truly multilateral institution with a global mandate, yet when reduced participation reduces influence on outcomes, Members’ views may change.
What, then, are the possible outcomes? It may be that most EU-US disputes will still be litigated at the WTO in spite of a potential bilateral solution. This is the pattern at NAFTA where litigation takes place under both agreements (indeed, at times simultaneously as inTuna/Dolphin II). It could also be that third party participation by the US and EU in disputes will satisfy their need to participate in the judicial processes that (indirectly) affect them.
There are, however, two more extreme alternatives: (1) for panels established by these free trade agreements to ignore de facto their interpretive obligations under the treaty and (perhaps unintentionally) to continue to regionalise the multilateral trade system both legally and economically; or (2), acknowledge the development of a truly global ‘common law of international trade’ and move into a pattern of coordination rather than separation in trade law. As is often the case with the interpretative process in judicial decision-making, it may not be clear which of these options a panel takes: how far does the obligation to ‘take into account’ extend? Note Simon Lester’s identification of the ambiguity inherent in this term.
What is of particular interest here is the way the acknowledgment of a problem (increasingly institutionalised regionalism within a multilateral trading system) and an attempt to offer a remedy (here through interpretative conflict avoidance) may unintentionally initiate a new stage in the development of trade law. Which way it goes, of course, remains to be seen