and Mitchell Moranis, WilmerHale
As discussed recently in this forum, the Vienna Convention on the Law of Treaties is the prevailing mechanism for the interpretation of investment treaties (Interpreting Investment Treaties, Roberto Castro de Figueiredo, 21 Oct. 2014). The Vienna Convention, however, was adopted in May 1969. According to the ICSID Database of Bilateral Investment Treaties, there existed 77 BITs at that time. Today there are roughly three thousand investment agreements.
This number – three thousand – makes the investment treaty regime unique. Most treaties are singular within their respective field: there is one Convention on the Law of the Sea, one Rome Statute. In the investment context, however, there are a plurality of individual treaties. And all of these treaties purport to serve the same purpose: to promote and protect investment. The system therefore consists of three thousand discrete documents generally creating the same rights and obligations. Does it make sense to approach these agreements with the tools used to approach large scale conventions?
The Vienna Convention certainly contemplated the existence of bilateral treaties. While there were only a handful of BITs in 1969, bilateral treaties are probably the oldest type. The text of the Vienna Convention reflects consideration of this species of treaty. Article 60 distinguishes between material breaches of bilateral and multilateral treaties and a number of articles refer only multilateral treaties.
However, those articles that provide interpretive guidance, most importantly Articles 31 and 32, do not distinguish between bilateral and multilateral treaties. They also do not distinguish between a treaty and what some refer to as a “treaty regime.” See Jeswald W. Salacuse, The Emerging Global Regime for Investment, 51 Harv. Int’l L. J. 427 (2010).
Article 31(1) provides that “[a] treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose.” It is difficult to take issue with this clear directive. Article 31(1) strikes a balance between the text of a treaty and its context, object, and purpose. The two are intended to complement one another, avoiding notorious battles in other fora between the language and purpose of a given text. Such a practice fits investment treaties, which often clearly enumerate their purpose in preambles.
Article 31(2) provides that the context of a treaty includes the text, preamble, annexes, any related agreements made between all parties in connection with the conclusion of the treaty, and any instrument made by one or more parties in connection with the conclusion of the treaty and accepted by the other parties as a related instrument. Article 31(3) provides that in addition an interpreter shall consider subsequent agreements regarding interpretation or application of the treaty, subsequent practice in treaty application, and “any relevant rules of international law applicable in the relations between the parties.” Finally, Article 32 allows for recourse to supplementary means of interpretation, including the preparatory work of the treaty and the circumstances of its conclusion, to confirm the meaning reached through the process of Article 31 or in cases of ambiguity, obscurity, or manifest absurdity.
In the investment treaty context, the obvious omissions from Articles 31 and 32 are other similar treaties addressing identical subject matter, namely those between one party and a third party or even those between third parties. In the multilateral context, these other treaties simply do not exist. In investment law, they are ubiquitous.
While a tribunal could stretch the concepts of “context,” “related instrument,” and “preparatory work” to incorporate other treaty practice, such reasoning would seem to violate the language and purpose of the Vienna Convention. Articles 31 and 32, above all, embody the principle that treaty parties consent to their specific treaty alone. Hence the need in Article 31(2)(b) of other parties to accept instruments made by one or more parties in connection with the conclusion of the treaty as “an instrument related to the treaty.” Investment treaties, like all contracts, are creatures of consent. In the recent Decision on Jurisdiction and Admissibility in Alemanni v. Argentina (ICSID Case No. ARB/07/8), the tribunal embraced this understanding:
The Ambiente Ufficio tribunal begins by drawing attention to the different, and in some respects inconsistent, approaches taken by investment tribunals, and in international dispute settlement more generally . . ., before recalling that the task of each individual tribunal or judicial body is to apply the particular legal provisions governing the case before it, and to do so within the context of those provisions, including the specific institutional and procedural framework within which they are embedded. The present Tribunal agrees. Its task is not to establish any general legal regime for provisions of this kind, but to decide the very specific question whether Argentina has or has not given its irrevocable consent to arbitrate in the circumstances in which the present Claimants have brought the present arbitration. ¶ 304.
This view seems reasonable in light of the current state of bilateral and plurilateral play. While the investment treaty system may look, talk, and act like a “regime,” it is not a multilateral system.
Yet such principles often do not shape treaty interpretation. Tribunals frequently choose not to follow the Vienna Convention and instead rely on justice, fairness, policy implications, practical consequences, efficiency, or reasonableness. See J. Romesh Weeramantry, Treaty Interpretation in Investment Arbitration 157-164 (Oxford 2012). The common practice of citing previous decisions also implicitly circumvents the strict rules of interpretation. Take the analysis of the Greece-Romania BIT’s “fair and equitable treatment” provision in Spyridon Roussalis v. Romania (ICSID Case No. ARB/06/1). To define FET, the tribunal quoted the “case law” and looked at tribunals’ interpretations in Rumeli v. Kazakhstan (ICSID Case No. ARB/05/16), Parkerings v. Lithuania (ICSID Case No. ARB/05/8), Azinian v. Mexico (ICSID Case No. ARB(AF)/97/2), Tecmed v. Mexico (ICSID Case No. ARB(AF)/00/02), and Saluka v. Czech Republic (UNCITRAL). These previous tribunals did an excellent job of giving content to the words “fair and equitable treatment” and are reasonable persuasive authority as to how to interpret “fair and equitable treatment.” Yet these previous tribunals interpreted that phrase, respectively, in the Kazakhstan-Turkey BIT, the Norway-Lithuania BIT, NAFTA, the Mexico-Spain BIT, and the Netherlands-Czech Republic BIT. And when a tribunal looks to and relies on a previous tribunal’s interpretation, it is necessarily looking to and relying on the treaty underlying that interpretation.
Does “fair and equitable treatment” mean the same thing in all of these treaties? Yes and no. The words are the same. They have the same ordinary meaning, in accordance with Article 31 of the Vienna Convention. But they appear in each treaty with distinct contexts, objects, purposes, texts, preambles, annexes, related agreements, subsequent agreements, related instruments, and preparatory work. Under the Vienna Convention, this means the words should have different interpretations. But FET provisions have come to embody a specific set of protections. Why? Tribunals may wish to avoid the absurdity of the same words meaning three thousand different things. It is certainly easier to accept the legwork of earlier tribunals rather than start from scratch each time. The contexts, objects, purposes, and other tools of interpretations, if not identical, can be similar. But also the resemblance between treaty texts has made the investment treaty system a regime in practice.
Tribunals have navigated between principles and reality. The principles are the rules of the Vienna Convention and the notion of consent on which they are premised. The reality is the simplicity and rationality of looking past a single treaty to many others that purport to do the exact same work. The reality is the benefit of consistency and predictability across the system, for both substantive rules and the interpretive framework that defines those rules. A unique approach to interpreting investment treaties would help resolve these tensions.
Many questions must be answered before states would consent to unique investment treaty interpretation rules. A fundamental issue is whether such rules could be adopted multilaterally or as appendices to individual treaties. Another issue is the extent to which the presence of multiple state parties or a broader free trade agreement, rather than a classic BIT, should affect interpretation of investment provisions. Would a state’s other treaties be more relevant than the treaty practice of third parties? What implications and presumptions would the presence or absence of certain language in other treaties create? These are complex issues that require discussion. But current practice is inadequate to resolve them.