In what appears to be a first for the English courts, the decision in GPF Gp S.á.r.l v Republic of Poland overturned parts of an award on jurisdiction in an investment treaty arbitration, finding the tribunal had erred in finding no jurisdiction. The claims have been sent back to the same tribunal for continuation towards the liability phase.
This post examines the English Commercial Court’s application of international law principles on the domestic plane, as well as the de novo standard of review adopted by the court to analyse the claimant’s jurisdictional objections.
The challenge: according to GPF, the tribunal erred in finding it lacked jurisdiction in respect of two of its claims
GPF Gp S.á.r.l (GPF) brought an action under section 67 (Challenging the award: substantive jurisdiction) of the English Arbitration Act 1996 (AA 1996) challenging an award on jurisdiction dated 15 February 2017 (the award). The award was rendered by a three-member tribunal in a London-seated arbitration under the Stockholm Chamber of Commerce (SCC) Rules, brought pursuant to the bilateral investment treaty in place between Poland, Belgium and Luxembourg (the BIT).
GPF’s challenge was twofold:
- First, that the tribunal had erred in its determination that the arbitration provision in the BIT was limited to claims for expropriation, and did not extend to GPF’s claims for breach of the fair and equitable treatment (FET) standard.
- Secondly, that the tribunal had mistakenly found that its jurisdiction over GPF’s claims for indirect expropriation was limited to one specific allegedly expropriatory event (a decision rendered by the Warsaw Court of Appeal), and not all “prior measures” by Poland.
Bryan J’s application of international law
Considering GPF’s challenge, Bryan J explained that the starting point for analysing the relevant provisions of the BIT was international law; and, in particular, the principles of interpretation contained in Articles 31 and 32 of the Vienna Convention on the Law of Treaties (1969) (the VCLT), which codifies rules of treaty interpretation under customary international law.
He proceeded, correctly, to apply a “textual” interpretation of the BIT in accordance with Article 31 VCLT (General rule of interpretation), a rule that reflects the fact that the text is presumed to be the authentic expression of the parties’ intention, and is not to be substituted by their presumed intention (a “teleological” approach to interpretation).
Bryan J further observed that the interpretative approach set out in Article 31(1) VCLT brings within it the principle of effet utile (full effect must be given to EU law). This requires that provisions of treaties be interpreted so as to render them meaningful rather than meaningless, but without going beyond what the text of the treaty justifies.
He also explained that recourse to Article 32 (Supplementary means of interpretation) is applicable “only” in the event of there being an ambiguity if the principles in Article 31 are applied. As such, Article 32 “cannot” be used to change or contradict an ordinary meaning that can be derived from the application of Article 31.
Applying the above interpretative approach, it was held by Bryan J that the tribunal did have jurisdiction over GPF’s claim for breach of FET, and over the prior measures comprising part of GPF’s indirect expropriation claim.
While neither party disputed Bryan J’s approach outlined above, his analysis serves as a useful reminder of the proper method for interpreting treaties in accordance with VCLT principles.
A de novo standard of review applied
As regards the standard of review to be applied in a section 67 application, Bryan J held (correctly) that it is “well established” that a section 67 hearing “is in the nature of a rehearing” on the issue of jurisdiction, whether the jurisdictional question is one of ratione personae or ratione materiae. Ultimately, the role of the court in addressing a section 67 application is to consider whether jurisdiction does or does not exist based on a true construction of the arbitration agreement. That ought not to be fettered by the reasoning of the arbitrators or indeed the precise manner in which arguments were advanced before the arbitrators.
Poland argued that, where GPF had not advanced arguments or evidence in the arbitration, GPF had hence waived or lost its right to do so before the court. However, Bryan J concluded that the requirements for a waiver were not met in this case. Indeed, it was difficult to contemplate how a waiver could arise in the context of a section 67 application where there will be a rehearing, and where there is no wording in the AA 1996 restricting parties in this way. In any event, Poland had not demonstrated any evidence to suggest it was prejudiced by having to address new arguments.
This may be contrasted with a decision by the Federal Supreme Court of Switzerland in Recofi v Vietnam, where the judge commented that only in exceptional circumstances would new facts or evidence be taken into consideration in a jurisdictional challenge concerning a treaty award.
Is a de novo standard of review the right standard of review?
As discussed in a previous post, different national courts adopt different standards of review when considering jurisdictional challenges. Some courts will consider the arbitrators’ jurisdiction de novo (even allowing the admission of new evidence) while others will pay the arbitrators’ decision a substantial degree of deference and therefore conduct only a limited review.
Bryan J’s approach in GPF v Poland reflects the firm view under English law that section 67 requires a full review. Internationally, however, this is a relatively controversial position. It is true that, pursuant to any arbitration agreement, including one in an investment treaty (absent an agreement between the parties to the contrary), a tribunal has the power under most national arbitration laws to determine its own jurisdiction. Yet, if national courts then apply a de novo review of a determination by an arbitral tribunal of its jurisdiction then, arguably, the finality of awards (oft cited as one of the chief virtues of arbitration) might be eroded.
The response to this is that where the very question in issue is whether or not a tribunal does indeed have the power to determine the parties’ dispute, a full review is essential. If not, then the national court is left in a worse position to that of the tribunal in properly considering its jurisdiction.
What is more striking in this case, however, is Bryan J’s approach to admitting new arguments. It might be argued that the scheme of the AA 1996 requires all arguments to be made at the earliest opportunity and not “saved up” for challenge proceedings.
While there are other examples of national courts having considered the question of a tribunal’s jurisdiction in a treaty case, English law is unusual for allowing a finding of no jurisdiction to be reviewed. It will be interesting to see how the tribunal deals with this.
It is worth noting though, that the outcome in GPF v Poland would not be possible in International Centre for Settlement of Investment Dispute (ICSID) arbitration, where arbitral awards are not subject to the supervisory power of any national court. In addition, there are supervisory jurisdictions for investment treaty arbitrations such as Switzerland which afford challenges to treaty awards a more limited level of review.
Separately, Poland expressly reserved the right to argue the compatibility of the BIT with EU law and any rights it might have following the decision of the Court of Justice of the European Union in Slovakia Republic v Achmea B.V. The Achmea judgment was handed down a few days after Bryan J’s judgment, and therefore not addressed.