The international arbitration community has lately been occupied with various proposals to reform investor-state disputes. On the interstate level, a consensus seems to be building that several aspects of the current system need to be modified in order for the system to safeguard its own legitimacy.

In this context, there are various reform proposals floated in different fora. Most notably – and most concretely – the EU Commission has been clear in its ambition to establish a more “court like” system to solve investment disputes. Although this work goes on in parallel on different levels, the proposal has been implemented in the signed but not yet ratified treaty with Canada, the CETA. Among other novelties, the reforms include the establishment of a permanent roster of arbitrators (Art. 8.27), an appellate tribunal (Art. 8.28) and the express opening for a future “multilateral” tribunal and appellate body (Art. 8.29).

Although there are many details still to be worked out with respect to this new dispute settlement mechanism, this blog post discusses one specific aspect, which is presumably relevant for most other reform plans, namely the overlooked issue of the place of arbitration.

Similar to many other investment treaties, Art. 8.23.2 CETA allows an investor to choose between different arbitration rules. The three available regimes are the ICSID Convention (with its associated arbitration rules), ICSID Additional Facility and the UNCITRAL Rules. There are numerous differences between these three rules. One such difference is that the two latter require a legal place of arbitration (lex arbitri) in a national jurisdiction. This aspect is integral to the current structure of the ICSID Additional Facility Rules and the UNCITRAL Rules and nothing in CETA suggests an intention to deviate from this fundamental feature.

The lex arbitri governs the outer procedural frame for the arbitration, and is also integral for the question of set-aside proceedings (determining which court has jurisdiction), as well as enforcement under the New York Convention (which requires that the award has a “nationality”). Arbitration under the ICSID Convention, by contrast, is “self-enclosed” in the sense that all such matters are regulated within the Convention itself. In fact, one stated intention behind the ICSID Convention was to remove domestic courts from the arbitration proceedings entirely.

Given this importance of the lex arbitri, remarkably little attention seems to have been paid to which domestic jurisdiction should govern non-ICSID proceedings under CETA and, more crucially, how the new features described above interact with the lex arbitri and the courts at the place of arbitration.

CETA does not specify the place of arbitration in case the ICSID Additional Facility Rules or the UNCITRAL Rules are used. Therefore, the choice rests with the disputing parties (in the case of the UNCITRAL Rules, which also authorize the tribunal to determine the seat if the parties cannot agree) or directly with the tribunal (in the case of ICSID Additional Facility Rules, which require the tribunal to consult with the disputing parties prior to determining the seat).

Presumably, in the CETA context, the contracting states envisioned – to the extent any such details were envisioned at all – that the disputing parties or the tribunal would choose a neutral and suitable place of arbitration (yes, that means you, Switzerland) where the more creative features of CETA would be respected. Alternatively, the assumption might be that the proceedings might be anchored in either Canada or the EU. In any event, the place of arbitration – or at least a limited list of available options – could easily have been specified directly in the treaty (which is the case, for example, with the NAFTA). As it stands now, it is at least theoretically possible for the disputing parties or the tribunal to choose any seat it deems appropriate, which might be in Singapore, Dubai or Moscow.

So why does the place of arbitration matter in this context? Generally speaking, it matters because the law at the place of arbitration determines the procedural frame for the arbitration and therefore, as international arbitration lawyers are well aware, may have important consequences. The lex arbitri can come into play in everything from fundamental matters such as arbitrability and the tribunal’s general powers, to more detailed procedural questions such as evidence production and court assistance during the proceedings.

More specifically, however, the proposed reforms potentially push up against that procedural frame in an unprecedented way with respect to the post-award stage, in the sense that the proposed appellate mechanism is a novelty whose compatibility with mandatory provisions of domestic arbitration law remains to be tested. At the post-award stage, will domestic courts accept that CETA deprives courts of jurisdiction to hear a challenge against the award, in favor of an international “appellate” court?

It is very possible that this will be a non-issue, in the sense that the contracting states’ intentions, as expressed in the CETA, will be respected by domestic courts. It is submitted, however, that the lack of regulation introduces (further) elements of uncertainty into the ambitious reform efforts. Imagine, for example, a disputing party which loses a CETA arbitration under the UNCITRAL rules. The UNCITRAL tribunal is seated in third state X and the losing party moves to challenge the award before that state’s courts, instead of under the CETA appellate mechanism. It is not particularly far-fetched that the courts in state X would not cede its jurisdiction to hear the challenge in such a case, given that the proceedings are governed by the state’s arbitration statutes which generally give state court’s exclusive jurisdiction over such challenges. In fact, only a few jurisdictions allow for parties to an arbitration to contract out of set-asides at the place of arbitration (although the number seems to be growing, see this earlier post on this blog). By way of illustration, Article 34 of the UNCITRAL Model Law provides that:

(1) Recourse to a court against an arbitral award may be made only by an application for setting aside in accordance with paragraphs (2) and (3) of this article.

(2) An arbitral award may be set aside by the court specified in article 6 [the court specified as competent by the domestic statute] only if […]

The Model Law approach is generally echoed in most arbitration statutes, which seem to presuppose that any post-award challenge against an award can only be brought before the courts in the state.

Finally, given the general calls to include domestic courts to a larger degree in future investor-state proceedings, it is surprising that this particular detail has not been discussed further in the arbitration community. What is the role of domestic courts vis-à-vis the appellate mechanism? How does the place of arbitration affect any subsequent enforcement efforts? Is it desirable to allow for both UNCITRAL/ICSID Additional Facility arbitration (which require domestic lex arbitri and domestic courts) and ICSID arbitration (which does not)? These questions should probably be discussed as part of any further reform effort. This applies especially in the EU context, given the well-known tension between tribunals and EU law: what difference does it make if the place of arbitration is in an EU jurisdiction, compared to in a “third state“ jurisdiction?

It is clear that the interaction between various lex arbitri and the new wave of “reformed” investor-state arbitration has not been researched, or even discussed, sufficiently. At least there is ample room for PhD proposals…


To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here. To submit a proposal for a blog post, please consult our Editorial Guidelines.

Kluwer Arbitration
This page as PDF


  1. A potential Pandora box, which could also have significant effects in terms of applicable law, given the arguable public policy status of certain provisions of EU law (e.g. competition law). The choice of the seat of arbitration could become a battlefield…

  2. I suggest that no legal arrangement can exclude the disputants from making their own arrangements – even if one is a State or State Entity. Arguably, any investment agreement could be drafted to permit conventional arbitration outwith any treaty ‘court’ provision.

    Is the CETA arrangement arbitration for the purposes of the New York Convention 1958. or is it a court?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.