The Cold War era brought to life, in a strange way, a number of all-encompassing treaties dealing with major subjects such as international treaty-making, diplomatic relations, law of seas, etc. Even among the topics covered by these treaties were enforcement of foreign arbitral awards and investment disputes. However, it seems like the world has already left the age of wide-ranging multilateral treaties governing basic State-to-State international relations. Nowadays we can see relative fragmentation and reluctance to multilateral treaties, e.g. the USA pulling out of some multilateral agreements and institutions.
The Vienna Convention on the Law of Treaties (VCLT) is an instrument of key significance to the system of international relations and international law making. Few other multilateral documents provide such fundamental guidance, as the VCLT regulates cornerstone issues such conclusion, validity, termination and, very importantly, interpretation.
Rules of interpretation as per the VCLT have lately become a point of focus for one more reason. Investment treaty arbitration requires the interpretation of international treaties governing investment regimes in an environment where other international instruments function. The Achmea saga is a prime example of a situation that an investment tribunal may face where a number of treaty regimes run in parallel or cross each other so that it becomes crucial to apply Article 31 VCLT to reach a particular reading of the treaty at hand.
One of the arguments put forward in the Vattenfall jurisdictional objection by Germany and the European Commission is that EU law forms part of the analysis of the Tribunal’s jurisdiction via Article 31(3)(c) VCLT, as “relevant rules of international law applicable in the relations between the parties”. To the extent that EU law or the Achmea European Court of Justice (ECJ) judgment fall under Article 31(3)(c) VCLT, they could become relevant to take into account when interpreting the Energy Charter Treaty (ECT). The position of the European Commission is that this approach would lead to a “harmonious interpretation” of the ECT.
The Tribunal noted that the correct starting point for the interpretation of the relevance of the ETC is the general rule of interpretation in Article 31(1) VCLT, i.e., “[a] treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms in their context and in the light of its object and purpose”. The Tribunal observed that, contrary to the European Commission’s position, Article 31(3)(c) is not the starting point of the interpretation exercise under the VCLT.
According to the Tribunal, EU law may come into the interpretation analysis if it is to be “taken into account, together with the context” under Article 31(3)(c). Thus, the Tribunal held that it is not the proper role of Article 31(3)(c) VCLT to rewrite the treaty being interpreted, or to substitute a plain reading of a treaty provision with other rules of international law, external to the treaty being interpreted, which would contradict the ordinary meaning of its terms.
The Tribunal considered that allowing for different interpretations of the same ECT treaty provision would be unacceptable, determining that this would be an incoherent and anomalous result and inconsistent with the object and purpose of the ECT and with the rules of international law on treaty interpretation and application.
What could be inferred from these arguments? When States enter into international legal obligations under a multilateral treaty, the principles of pacta sunt servanda and good faith require that the terms of that treaty have a single consistent meaning. State parties to a multilateral treaty are entitled to assume that the treaty means what it says, and that all State parties will be bound by the same terms. It cannot be the case that the same words in the same treaty provision have a different meaning depending on the independent legal obligations entered into by one State or another, and depending on the parties to a particular dispute. The need for coherence, and for a single unified interpretation of each treaty provision, is reflected in the priority given to the text of the treaty itself over other contextual elements under Article 31 VCLT. The need to maintain a uniform meaning of treaty terms is also clearly reflected in Article 33 of the Vienna Convention, dealing with treaties authenticated in two or more languages.
That is not to say that all differences in the substantive obligations of ECT Contracting Parties resulting from differences in the range of other relevant treaties to which each Contracting Party has signed up are entirely ironed out by the ECT. The ECT does make provision for variations among the substantive obligations of the Contracting Parties. For example, ECT Article 10, paragraph 1 identifies the obligations of each Contracting Party towards investors and includes the “treaty obligations” of the State and “any obligations it has entered into with an Investor”, and paragraph 10 identifies obligations under “the applicable international agreements for the protection of Intellectual Property Rights to which the respective Contracting Parties are parties”. But the crucial point is that while the application of those paragraphs may result in the substantive obligations of one Contracting Party differing from those of another, those paragraphs themselves have a uniform meaning for every ECT Contracting Party. Equally important, there is no similar provision in the ECT for differing obligations in relation to the ECT dispute settlement procedures.
The Vattenfall Tribunal found that the effects of such an interpretation in the manner proposed by the European Commission would not ensure “systemic coherence”, but rather its exact opposite. It would create one set of obligations applicable in at least some “intra-EU” disputes and another set of different obligations applicable to other disputes. This would bring uncertainty and entail the fragmentation of the meaning and application of treaty provisions and of the obligations of ECT Parties, contrary to the plain and ordinary meaning of the ECT provisions themselves.
Investment Treaty Interpretation In The Light of Art. 31 VCLT: Lessons From Vattenfall
The Vattenfall award on jurisdiction teaches several important lessons concerning use of the VCLT’s interpretative tools.
First, the Tribunal does not see the different paragraphs of VCLT Article 31(2) as separate approaches to treaty interpretation. The Tribunal assumes that one cannot “cherry pick” what approach to undertake. Instead, the Tribunal points out that every interpretative attempt should start at the very fundamental premise of interpretation which is set in Article 31(1): the first component of interpretation is ordinary meaning. Moreover, the means of interpretation is good faith, while the end of interpretation should be its object and purpose. This is outlined as the basis from where each Tribunal should start, including in investment treaty arbitration cases.
Second, the Vattenfall tribunal implies that the interpretative exercise may start but also end there. If the meaning becomes clear at this primary stage, there is no need to look further. If the plain wording of the investment treaty at hand in the particular case reveals in a satisfactory way the meaning underlying the rules and concepts subject to construction, the Tribunal should not move on to Article 31(3).
Third, this may not always be the case, but where it is not the case, this should not at all mean that the approaches enshrined under Article 31(2) should be applied light-heartedly. Paragraph 2 of Article 31 provides additional instruments to help the interpretative exercise. Vattenfall seems to suggest that these should be used cautiously in order not to produce a result which does not correspond to the very ends of Article 31. For instance, Vattenfall might imply that one should not “embed” entire regimes within a particular treaty system which will not help the interpretative task – on the contrary, this will lead to further incoherence as some of the treaty parties would be bound by the treaty at hand, e.g. ECT, while others will have infused within their relations other treaties, e.g. EU instruments. This may lead to contrasting interpretations of one and the same concept used in one and the same document. Further, the Vattenfall Tribunal seems to be quite cautious in using the concept of “relevant rules of international law”. The suggested approach is rather not to undertake an expansive view and look for a relevant rule of law high and low; instead, there should be a precise and defined rule which is relevant. One step further, it may be considered whether the Vattenfall Tribunal actually might have intended to state that not each and every potential rule should be used as a tool for interpretation: this should rather be an established and developed one of strong importance.
Although the Vattenfall award on jurisdiction does not claim to be an exhaustive manual for application of Article 31 of VCLT, it may provide precious guidance on how to make use of the VCLT. It outlines the key approach to interpretation and demonstrates how it can operate in investment treaty disputes so that investment treaty arbitration is not a self-contained island but a breathing part of the larger body of international treaty law.
To see our full series of posts celebrating the 50th jubilee anniversary of the Vienna Convention on the Law of Treaties, click here.