On Thursday, the arbitral tribunal in Croatia/Slovenia rendered a final award on the merits of the dispute despite what a previous post on this blog called “severe breaches of duty of confidentiality and impartiality” during the arbitral proceedings. In commercial and investment arbitration, where procedural irregularities arise, either party can seek annulment of the award before an annulment body, such as a national court or an ICSID ad hoc committee. In other words, commercial and investment arbitral awards are subject to a compulsory control mechanism. The case of Croatia/Slovenia, however, is an interstate arbitration. Two questions thus arise. First, as a matter of lex lata, are interstate arbitral awards subject to a compulsory control mechanism? Second, as a matter of lex ferenda, should interstate arbitral awards be subject to such a mechanism? This post addresses these two questions in turn.
As a matter of lex lata, interstate arbitral awards are not subject to a compulsory control mechanism. This might seem obvious to many readers. After all, as a matter of practice, States very rarely seek the annulment of interstate awards. The only two modern cases where States have done so, Arbitral Award Made by the King of Spain (Honduras v. Nicaragua) and Arbitral Award of 31 July 1989 (Guinea-Bissau v. Senegal), are both cases where the parties consented to the jurisdiction of the International Court of Justice (ICJ) to settle the dispute. Absent such consent, States simply do not seek the annulment of interstate arbitral awards.
This practice is in line with the legal framework governing interstate arbitration. One might wonder why there is no compulsory control mechanism for interstate awards, but there is for commercial and non-ICSID investment awards (i.e., annulment at the national courts of the seat of arbitration). After all, commercial, non-ICSID investment, and interstate arbitration all have quite similar legal frameworks. In all three, the arbitral award is “final and binding”. In all three, State entities can be parties to the arbitrations. And in all three, the arbitral tribunal is often seated in a national jurisdiction where the national law expressly grants national courts the power to annul international arbitral awards. So why are national courts able to annul commercial and non-ICSID awards, but not interstate awards? To be more concrete, as Croatia/Slovenia is seated in Belgium, why can’t Croatia go to Belgian courts to annul the award rendered on Thursday?
One commonly cited reason is that interstate arbitral awards are completely “delocalized”: they are subject only to public international law, not to any national law, so it would be absurd for national courts to have the power to annul them. Nevertheless, there are two important counterarguments to consider. First, many scholars argue that commercial and non-ICSID arbitral awards are similarly “delocalized”, but they are undoubtedly subject to annulment (the effect of the annulment on enforcement is another question). And second, interstate arbitrations today, like commercial and non-ICSID arbitrations, generally have a seat of arbitration, so it would be inaccurate to argue that they are completely detached from national jurisdictions.
A more practical reason why national courts cannot annul interstate arbitral awards is State immunity. Indeed, in Chagos Marine Protected Area (Mauritius v. United Kingdom), the United Kingdom invoked State immunity to counter Mauritius’s claim that the award could be annulled by Dutch courts (Reasoned Decision on Challenge, para. 117). Article 5 of the United Nations Convention on Jurisdictional Immunities of States and Their Property (which has not yet entered into force, but largely reflects customary international law) provides that “a State enjoys immunity … from the jurisdiction of the courts of another State subject to the provisions of the present Convention”. The provision concerning arbitration, Article 17, provides that “a State cannot invoke immunity … in a proceeding which relates to … the setting aside of the award”, but the provision applies only to arbitrations where (1) the arbitration agreement is between the State and “a foreign natural or juridical person”; and (2) the dispute is one “relating to a commercial transaction”. As a result, Article 17 applies to commercial and investment arbitrations. Nevertheless, as the ILC Commentary makes clear (p. 55), Article 17 does not apply to interstate arbitrations because (1) the arbitration agreement is formed between States; and (2) interstate disputes often (though not always) concern sovereign rights rather than commercial transactions. Article 12 of the European Convention on State Immunity (which has entered into force, but is not widely ratified or acceded to) contains the same two requirements.
Whether one agrees more with the first reason, the second reason, or perhaps another reason not mentioned above, the conclusion is the same: as a matter of lex lata, interstate arbitral awards are not subject to a compulsory control mechanism.
As a matter of lex ferenda, in the opinion of the present author, interstate arbitral awards should be subject to a compulsory control mechanism. The policy debate concerning the propriety of annulment centers around the reconciliation of two principles: finality and fairness. On the one hand, arbitral awards should be final, such that a dissatisfied party cannot relitigate the dispute simply because it disagrees with the outcome. On the other hand, arbitral awards should be fair, such that a dissatisfied party can relitigate the dispute if there are legitimate reasons for doing so.
In commercial and investment arbitration, this balance between finality and fairness has been struck in the mandates of annulment bodies. In order to promote fairness, the rules governing commercial and investment arbitration (e.g., the ICSID Convention, the UNCITRAL Model Law, and national laws) provide for annulment as a compulsory control mechanism. In order to promote finality, these rules set forth very limited grounds for annulling arbitral awards, which are generally procedural rather than substantive. Although one cannot conclude that this balance is perfect, it is widely accepted.
In interstate arbitration, the balance has been struck with greater emphasis on finality. This is because, although the annulment of interstate awards remains a possibility, there is no compulsory control mechanism. One can argue that finality is more important in interstate arbitration because the continued existence of interstate disputes can lead to serious consequences like the loss of life (e.g., in territorial boundary disputes featuring armed conflict). Nevertheless, Croatia/Slovenia and South China Sea (Philippines v. China) both show that concerns over fairness can also arise, and where these concerns are not addressed, even a “final” award will not be treated as final by all the parties, such that the aforementioned serious consequences can still arise.
Indeed, there are at least two reasons why a compulsory control mechanism for interstate awards would be desirable. First, there needs to be an impartial forum to determine whether procedural irregularities have occurred. In South China Sea, China expressed serious concerns not only over the jurisdiction of the tribunal, but also over its composition and constitution, yet no impartial forum could hear this claim on a compulsory basis. In the two most recently filed UNCLOS arbitrations Enrica Lexie (Italy v. India) and Ukraine v. Russia, similar concerns could arise regarding the propriety of the appointing authority’s self-appointment, yet again no compulsory impartial forum can hear the claim. Second, there needs to be an impartial forum to determine the consequences of procedural irregularities. In Croatia/Slovenia, there is no question that procedural irregularities occurred, but there is no compulsory impartial forum to determine their consequences.
Although the arbitral tribunal itself could be the one to decide on these matters (as the Croatia/Slovenia tribunal did in its Partial Award of 30 June 2016), under the principle of nemo judex in causa sua, the tribunal should not be the one, or at least not the only one, deciding on the consequences of its own procedural irregularities. Doing so not only undermines the legitimacy of the arbitral award, but also may lead the dissatisfied State to unilaterally declare that the award is “null and void”, as China did in South China Sea, or that the arbitration simply “does not exist”, as Croatia did in Croatia/Slovenia.
In conclusion, interstate arbitral awards are not subject to a compulsory control mechanism, but, in the opinion of the present author, they should be. On a practical level, how could this compulsory control mechanism be established? First, as a matter of customary law, one could envisage an evolution of the law of jurisdictional immunity so as to extend the exception contained in Article 17 of the United Nations Convention on Jurisdictional Immunities of States and Their Property to interstate arbitrations, such that at the very least State immunity could not be invoked in interstate annulment proceedings. Second, as a matter of conventional law, States could come together to conclude a multilateral convention that provides for a compulsory control mechanism for interstate arbitrations between State parties, as has been proposed at least once in the past. Third, as a matter of State practice, States could simply begin to offer to accept the jurisdiction of the ICJ on the basis of forum prorogatum for disputes over the validity of interstate arbitral awards.
The views expressed in this post are solely those of the author, and do not necessarily reflect the views of any institution with which the author is affiliated. The author (LinkedIn, SSRN) may be contacted at email@example.com.