Kluwer Arbitration Blog | 11 February 2021
Could an appellate review mechanism “fix” the ISDS system?
by Dr. Margie-Lys Jaime
In the past few years, the world has been following the Investor-State Dispute Settlement (ISDS) reform debate under the aegis of the United Nations Commission on International Trade Law (UNCITRAL). This discussion started in 2017, when the UNCITRAL Working Group III began its work on ISDS reform.
Among the proposals submitted by member States and other stakeholders, two systemic reform proposals call for the particular attention of academics and practitioners, as they may involve the replacement (total or partial) of the ISDS system. These are proposals for: (i) the establishment of a multilateral investment court; and (ii) the creation of an appellate mechanism.
Already, the European Union has successfully negotiated the inclusion of a permanent first instance tribunal and an appellate tribunal in its recent International Investment Agreements (IIAs). For some commentators, disputes arising out of the new EU IIAs could end in the destruction of the basic principles of neutrality, finality, and party autonomy. For others, the incorporation of an appeal mechanism in IIAs is not novel. In fact, in 2004 the International Centre for Settlement of Investment Disputes (ICSID) explored the possibility of creating an ICSID appeals facility, noting that at least 20 countries may have included an appeal mechanism in their IIAs.
I have argued elsewhere that the establishment of an appeal mechanism is more likely to be accepted by the international community than the creation of a multilateral investment court. This is in part because the proposal for a multilateral investment court presently envisages giving investors little role in appointing a first instance tribunal. Indeed, an appeal mechanism would allow the parties to continue selecting the decision-makers at first instance, while ensuring consistency and predictability in the system. But, while an appeal mechanism might be preferable over a multilateral investment court, is it the panacea for addressing ISDS concerns?
In this post, I contend that an appeal mechanism is not the panacea to all primary concerns associated with ISDS; namely, concerns pertaining to: (i) the lack of consistency, coherence, predictability, and correctness of arbitral decisions by ISDS tribunals; (ii) the lack of impartiality, independence and diversity of arbitrators and decision-makers; and (iii) the cost and duration of ISDS cases. As one commentator has noted, there is a need for holistic reform. Indeed, a holistic approach could address the broader legitimacy concerns over ISDS more effectively.
Why Implement an Appeal Mechanism?
The overarching purpose behind the establishment of an appellate review framework for investment arbitration is to enhance coherence and consistency in the ISDS system by creating an appellate body able to review manifest errors in the interpretation and application of treaty law. As highlighted by some commentators, such an appeal mechanism is expected to improve the quality and consistency of investment arbitration awards by moving towards a precedent-based system.
That being said, the establishment of a multilateral appeal mechanism is controversial. For some, the creation of an appellate mechanism could have a negative impact on the duration and costs of proceedings, thereby creating uncertainty in ISDS. I disagree.
While the appeal proceedings may entail additional time and costs (particularly if the tribunal is vested with the power of reviewing findings of fact), the appeal mechanism would substitute the annulment/set-aside proceedings, without adding an additional layer of control to the process. In other words, once the investor chooses to appeal the award, it would not be subject to annulment under the ICSID Convention nor could it be set aside or annulled by a court in the seat of the arbitration under the ICSID Additional Facility Rules or the UNCITRAL Arbitration Rules.
Furthermore, contrary to what has been argued, an appeal mechanism could avoid the additional costs produced by a second arbitration following the annulment of the decision of a first instance tribunal. Indeed, unlike in the domestic context where the remand of the issue to the lower instance is the general rule in some jurisdictions, an appeal tribunal in the arbitration context should be able to modify or reverse the legal findings and conclusions of the first tribunal without the need to remand the issue to a new tribunal.
What Should the Scope and Standard of Review Be?
In treaty-based arbitrations, arbitral tribunals are called on to interpret a wide range of standards included in IIAs, as well as a larger spectrum of principles and norms under public international law. The concerns are not the same where the basis for ISDS is a contract or domestic legislation. These factors would justify limiting the appeal option to treaty-based ISDS.
Appeals based upon errors of law are necessary in order to correct errors in treaty interpretation. Yet, this power should be limited and carefully drafted. There is, as has been highlighted, a need to clarify the scope of the ʻlawʼ in ISDS. Should it be limited to an error of law that is material and prejudicial? Should it include the application or interpretation of any applicable law (not restricted to investment treaties and notions in public international law)? Or instead, should it list specific treaty standards such as expropriation, fair and equitable treatment, and non-discrimination? The latter seems to be the most appropriate path to follow.
Similar to the notion of ʻlawʼ, there is a need to clarify the meaning of ʻfactʼ. I am of the view that errors in the appreciation of facts should be limited to manifest errors in the appreciation of the relevant domestic law and the assessment of damages, and not to any other error in the appreciation of the facts, even if this error is manifest. This would provide a substantial degree of deference to the findings of the first instance tribunal, reducing unnecessary costs and delays. Indeed, a complete de novo review of both law and facts could have a negative impact on the cost and duration of the proceedings.
In addition, the appellate tribunal should be able to grant or deny the appeal based on the annulment grounds of the ICSID Convention (Art. 52). This would ensure that the option for appeal was integrated into the ICSID system. By contrast, grounds for annulment or setting aside under the New York Convention may not be applicable to treaty-based arbitration.
UNCITRAL Working Group III is currently discussing possible scopes and standards of review for an appellate mechanism. In this context, the consolidated draft provisions on an appellate mechanism and enforcement include a reference to Article VI(1) of the New York Convention, which could leave room for intervention on domestic grounds of arbitrability and public policy. The New York Convention was not designed for investment arbitration. Keeping both lists of grounds (under ICSID and New York Conventions) could create a double standard, allowing the appellant to argue the most advantageous provision or combination of provisions to its case.
If the Commission decides to include grounds from the New York Convention that are not contained in the ICSID Convention, then it would be preferable to enumerate such grounds instead of referring to the relevant provisions of both Conventions, avoiding duplications. For instance, a departure from a fundamental rule of procedure under the ICSID Convention may cover Article V(1)(b) and V(1)(d) grounds of the New York Convention.1)
When Should the Appellate Tribunal Dismiss the Appeal or Order Security for Costs?
Modelled on Rule 41(5) of the ICSID Arbitration Rules, the appellate tribunal should be able to dismiss the appeal where it is clear that the appeal is manifestly unfounded or frivolous. This has been included in recent negotiated IIAs (see, for example, the European Union-Singapore FTA).
Moreover, the tribunal should consider all relevant circumstances in determining whether to order the appellant to provide security for costs and the amount to be provided. Yet, the security for costs order should not unduly undermine the appellant’s ability to pursue its claim, as this may constitute a denial of justice. This would ensure a balanced approach, taking into account the different interests at stake.
As I concluded in a recently published chapter focused on reshaping ISDS through an Appellate Review Mechanism,2) the introduction of an appeal system should favour “a balance between the finality and the correctness of arbitral awards.” Furthermore, the consistency and correctness of arbitral awards should not unnecessarily undermine the finality of arbitral awards by increasing the cost and duration of the proceedings.
This does not mean that an appeal mechanism would be the panacea for all the concerns related to ISDS, but it could at least be a significant step towards the improvement of the arbitration process. Other initiatives that are currently under consideration by Working Group III – for instance, a code of conduct for adjudicators, the adoption of specific rules on treaty interpretation, security for costs and frivolous claims, and third party funding – are also useful tools to further confirm the legitimacy of the ISDS system.
1. ↑ Pursuant to Article V(1), the recognition and enforcement of the award may be refused, if: (b) The party against whom the award is invoked was not given proper notice of the appointment of the arbitrator or of the arbitration proceedings or was otherwise unable to present his case;… or (d) The composition of the arbitral authority or the arbitral procedure was not in accordance with the agreement of the parties, or, failing such agreement, was not in accordance with the law of the country where the arbitration took place.
2. ↑ For further information, see The Investor-State Dispute Settlement System. Reform, Replace or Status Quo?, by Alan M. Anderson & Ben Beaumont (eds), (Kluwer International Law, 2020), 528 pp.
To view all formatting for this article (keywords, tables, etc.), please read it on the original page