Symposium on the VCLT, ILC and Investor-State Disputes: Legal Questions Concerning the Temporal Application of Treaties in International Investment Arbitration Cases

Introduction 

International investment tribunals face various challenges when conceptualising the temporal scope of the various treaties they interpret and apply. A review of case law developed under international investment tribunals evidences that many unsettled questions remain in this regard. This post draws on our recently-published chapter on these topics to examine how international investment tribunals have sought to address aspects of temporality in international investment arbitration. It draws on three different elements of temporality and assesses the singularities of international investment arbitration vis-à-vis general international law: (a) entry into force and nonretroactivity of the treaty; (b) provisional obligations pending the entry into force of the treaty; and (c) temporal issues pertaining to the termination of a treaty.

Entry into Force and Non-Retroactivity of Treaties

The precise timing of the entry into force of a treaty is crucial for determining the temporal scope of a treaty: a State can, through actions or omissions, only breach obligations that are in force and binding on that State at the time of the action or omission. This general rule of non-retroactivity is provided by Article 28 of the Vienna Convention on the Law of Treaties (VCLT), stipulating that ‘[u]nless a different intention appears from the treaty or is otherwise established, its provisions do not bind a party in relation to any act or fact which took place or any situation which ceased to exist before the date of the entry into force of the treaty with respect to that party’. Article 28 codified an already existing rule of international law, which has been confirmed by the ICJ.

However, intertemporal law as defined in the ILC’s Draft Articles of Responsibility of States for Internationally Wrongful Acts (ARSIWA) does not prevent facts or acts occurring prior to the entry into force of a particular obligation to be considered in the analysis of the breach of that obligation (Articles 14-15 ARSIWA). Thus, acts occurring before the entry into force of the treaty can be considered in order to establish a factual basis for later breaches or to provide evidence of intent, and acts of a continuous or composite character can trigger responsibility for acts that took place even before the entry into force of the treaty. International investment tribunals have fully endorsed those principles, having found that non-payment of amounts specified in contractcontinuing delay by national courts, and the continuous withholding of permits to constitute continuous acts .  See e.g., SGS Societe Generale de Surveillance SA v. Republic of the Philippines, ICSID Case No. ARB/02/6 (29 January 2004) paras 166-167; Chevron Corporation (USA) and Texaco Petroleum Corporation (USA) v. The Republic of Ecuador , PCA Case No. 34877, Interim Award (1 December 2008) para. 298; Pac Rim Cayman LLC v. Republic of El Salvador, ICSID Case No. ARB/09/12, Decision on the Respondent’s Jurisdictional Objections (1 June 2012) para. 3.43.

International investment tribunals have also considered facts which occurred before the entry into force of a treaty when determining whether the treaty was subsequently breached. In Tecmed v. Mexico, whilst noting that acts which took place in their entirety before the treaty became operative could not violate the terms of the treaty, the tribunal observed that:

it should not necessarily follow from this that events or conduct prior to the entry into force of the Agreement are not relevant for the purpose of determining whether the Respondent violated the Agreement through conduct which took place or reached its consummation point after its entry into force … [C]onduct, acts, or omissions of the Respondent which, though they happened before the entry into force, may be considered a constituting part, concurrent factor, or aggravating or mitigating elements of conduct or acts or omissions of the Respondent (paras 66- 68). 

Furthermore, investment tribunals have endorsed the principle provided for by Article 28 VCLT also in the context of disputes arising before the entry into force of a treaty. In Feldman v. United Mexican States, the tribunal stated that ‘[g]iven that NAFTA came into force on January 1, 1994, no obligations adopted under NAFTA existed, and the Tribunal’s jurisdiction does not extend before that date. NAFTA itself did not purport to have any retroactive effect’. This line of reasoning is perfectly consistent with Article 28 VCLT: as no different intention appears from NAFTA or is otherwise established, the treaty cannot be accorded retroactive effect. 

The most controversial issues with respect to the non-retroactivity of investment treaties have perhaps arisen when tribunals have sought to identify the real cause and source of origin of a dispute. In Luchetti v. Peru, the tribunal found that the facts that gave rise to the earlier dispute continued to be central to the later dispute, and accordingly rejected the claim as the real dispute arose before the BIT entered into force (paras 36, 48-53). A more lenient approach is evident in the case law of the ICJ when, for instance, in Certain Properties, the ICJ found that certain decisions by German judicial institutions were ‘inextricably linked’ to the interpretation of the Settlement Convention with regard to Germany in 1954 and the Czechoslovakia Benes Decrees of 1945, which were considered to be the ‘real cause’ of the dispute (para. 28).

Such approach was also evident in Jan de Nul v. Egypt, where the investment tribunal concluded that although the claimant had already challenged the actions before Egyptian domestic courts, the tribunal found itself competent to exercise jurisdiction and concluded that the first dispute was based on domestic law, and the second was based on the relevant BIT, wherefore the second dispute was covered by the BIT (para. 117). Thus, in essence, where jurisdiction over a dispute arises under a BIT containing a clause specifying that the treaty does not apply to disputes or claims originating before its entry into force, arbitral tribunals have focused on the specific timing as to when the dispute arose, and whether the facts that gave rise to an earlier dispute continued to be central to the later dispute. 

Provisional Obligations of States Pending the Entry into Force of the Treaty

A pending treaty is capable of imposing certain obligations on States notwithstanding the fact that the treaty is not yet in force and, typically, would not yet have acquired any legal effects. This usually occurs under two circumstances: either by virtue of what is known as the ‘interim obligation’under Article 18 VCLT or through the provisional application of the treaty under Article 25 VCLT.  

Article 18 VCLT imposes an obligation on States to refrain from acts which would defeat the object and purpose of a treaty when a State has signed the treaty or expressed its consent to be bound by the treaty, pending the entry into force of the treaty. In Tecmed, the tribunal sought to ascertain what sort of conduct would reach the threshold of defeating the object and purpose of a treaty:

Article 18 does not only refer to the intentional acts of States but also to conduct which falls within its provisions, which need not be intentional or manifestly damaging or fraudulent to go against the principle of good faith, but merely negligent or in disregard of the provisions of a treaty or of its underlying principles, or contradictory or unreasonable in light of such provisions or principles (para 71). 

It is doubtful whether this is an entirely accurate analysis of Article 18 VCLT. The provision may indeed prohibit fraudulent as well as negligent behaviour. At the same time, it is imperative to bear in mind that by incorporating the word ‘defeat’, Article 18 sets a very high threshold. The provision does not refer to an obligation to comply with or act consistent with the provisions of a pending treaty, or even the object and purpose of individual treaty provisions but prohibits conduct which defeats the very raison d’etre of the treaty.

The provisional application of a treaty entails that a State must comply with some or all provisions of a treaty that is not yet in force (Article 25 VCLT). A treaty applied provisionally is legally binding and enforceable under international law. The issue of provisional application arose in Ioannis Kardassopoulos v. Georgia, where a Greek investor argued that Georgia had expropriated a pipeline construction concession and failed to reimburse him for the loss of his investment. Both Greece and Georgia signed the ECT on 17 December 1994 and applied the treaty on a provisional basis. Georgia argued that it had no legal obligation to refrain from expropriation. It relied on Article 1(6) ECT, which stipulates that the ECT only applies to matters affecting investments after the treaty enters into force. The  tribunal disagreed and stated that Article 45(1) ECT obliged both States to apply the whole treaty as if it had entered into force on 17 December 1994, the date of their respective signatures. Hence, consistent with the VCLT, arbitral tribunals have confirmed the legally binding nature of a provisionally applied treaty(para. 220). 

Temporal Issues pertaining to Treaty Termination 

Just as a State cannot breach a treaty not yet in force, a State cannot breach a treaty after its termination. Article 70 VCLT provides that unless the treaty otherwise provides or the parties otherwise agree, the termination of a treaty releases the parties from any obligation further to perform the treaty. The question of the consequences of termination has arisen in investment arbitration principally in relation to the applicability of the ICSID Convention after a State’s withdrawal. Article 71 of the ICSID Convention provides that ‘[a]ny Contracting State may denounce this Convention by written notice to the depositary of this Convention. The denunciation shall take effect six months after receipt of such notice’. Furthermore, Article 72 makes clear that notice of withdrawal shall not affect the rights or obligations under the ICSID Convention of the withdrawing arising out of consent to the jurisdiction of the Centre given before notice was received by the depositary.

However, the ICSID Convention does not define the conditions for terminating consent. This has raised certain questions, including whether a claim can be initiated after the treaty has been terminated with respect to that State. Tribunals have disagreed on the scope and effect of Articles 71 and 72. For instance, in Venoklim v. Venezuela, the tribunal found that during the six-month period after denunciation, Venezuela was in fact still a Contracting State to the ICSID Convention. As such, its consent to arbitration was still in existence during the six-month period. This consent was capable of being accepted and acted upon by an investor who wished to bring a claim for breach (para. 63).

In contrast, the tribunal in Fabrica de Vidrios v. Venezuela considered ‘consent’ under Article 72 to entail consent that was already perfected. As such, a State which had notified the treaty depositary of its withdrawal from the treaty could not be viewed, during the six-month termination period, to still have consented to arbitration. As soon as the notification of withdrawal had been deposited, an investor could no longer act upon consent previously given (paras 250-306).

Conclusion 

The temporal application of treaties has proven to be a complex undertaking. With respect to the non-retroactivity of the treaty, investment tribunals have repeatedly endorsed the principle provided for by Article 28 VCLT and Article 13 ARSIWA. Acts occurring before the entry into force of the treaty have only been considered in order to clarify and understand the background, causes, and scope of violations of the relevant treaty once in force, to provide a factual basis for breaches occurring through later conduct, and to provide context to actions occurring subsequent to entry into force. 

As concerns provisional obligations, the broad reading of Article 18 VCLT as embraced by the tribunal in Tecmed stands at odd with the threshold and scope of the provision as anticipated by the VCLT drafters and upheld under general international law. The reactions of States during the negotiations on the subject of Article 18 emphasised the role of the object and purpose of the treaty, which would be defeated through adverse action of States. In contrast, arbitral tribunals have confirmed the legally binding nature of a provisionally applied treaty, consistent with Article 25 VCLT. Lastly, the issue of consent to jurisdiction in the context of termination of treaties remains unclear. Inconsistency in case law raises uncertainty as to the legal effects of Articles 71 and 72 of the ICSID Convention, and it is yet to be determined whether consent to jurisdiction would remain effective and valid even after the lapse of the six-month termination period. That said, the temporal issues discussed are complicated, and it should be borne in mind that the solutions offered by many international investment tribunals have demonstrated due regard for the principles laid down in the VCLT and ARSIWA. This attitude serves as a good starting point in the endeavour to further develop the cohesiveness of international investment law with public international law.

The promo code 20VCLT22 offers a 20% discount on the retail price of The Vienna Convention on the Law of Treaties in Investor-State Disputes: History, Evolution and Future in the Wolters Kluwer eStore. The code is valid through to 28 February 2023. 

Agnes Rydberg is Lecturer in International Law at the University of Sheffield. Malgosia Fitzmaurice is Professor of Public International Law at the Queen Mary University of London.