Justice Philip Jeyaretnam: Speech delivered at the International Bar Association Symposium 2024

THE ROLE OF THE COURTS IN ARBITRATION: AN ASIA PACIFIC PERSPECTIVE
26 August 2024

Transnational Issue Estoppel in the Context of International Arbitration

I.     Introduction

1. The courts play an essential role in relation to arbitration, because if an award debtor does not voluntarily perform the award then that award will need to be enforced by a court. Moreover, the award debtor has the opportunity to challenge the award before the courts of the arbitration seat. This challenge is not to the legal or factual merits of the award. It is strictly limited to grounds that concern the foundation and integrity of the process. The most common assertions are that the tribunal lacked jurisdiction or that there was a breach of natural justice or some other due process error. Less common, but just as important in protecting the integrity of the system, are challenges that the award was obtained by fraud or corruption. Lastly, awards may be challenged on the basis of public policy, which is defined narrowly by courts applying national law based on the UNCITRAL Model Law on International Commercial Arbitration (the “Model Law”). 

2.     In order to play an effective role, courts must intervene where mandated by the Model Law. Equally critical is that the court’s own procedure should ensure that challenges are dealt with expeditiously. For Singapore-seated international arbitrations, challenges may be filed in the General Division of the High Court or in the division of the High Court which is the Singapore International Commercial Court. In both divisions, the decision at first instance on a challenge to an arbitration award is typically delivered within six months of filing, and if an appeal is filed then typically that decision would be given within six months of filing of the appeal. 

3.     In this short address, I will focus on the current state of play in relation to situations where a court considers an issue concerning the status or validity of an award after the same issue has been considered by another court. The question is what (if any) preclusive effect should be given to prior decisions made by courts in other jurisdictions. The further question is whether the answer differs depending on whether the first court is the seat court or an enforcing court.

II.     Competing theories of the relationship between international arbitration and national courts 

4.     Broadly speaking, there are two competing theories concerning how arbitration as an adjudicative process relates to national courts. The first is the “delocalisation theory”: arbitration as a transnational legal process operating independently of national law. On this view, “no single state, not even the seat of the arbitration, has the final say on the validity or enforceability of an award.”⁽¹⁾ The arbitral process and award is subject to judicial scrutiny only at the place of enforcement.⁽²⁾ 

5.     There was a time when Belgium took delocalisation to its extremes. In 1985, legislation was enacted in Belgium that precluded the setting-aside of Belgian awards where none of the parties were nationals or residents.⁽³⁾ That experiment was subsequently abandoned in 1998 after the lawmakers realised that it “dissuaded rather than encouraged parties to choose Belgium as the seat of their arbitration”.⁽⁴⁾ 

6.      Today, we see the strongest exponents of the delocalisation theory in France. In the famous Putrabali case, the French Cour de cassation held that “[a]n international arbitral award, which does not belong to any state legal system, is an international decision of justice and its validity must be examined according to the applicable rules of the country where its recognition and enforcement are sought”.⁽⁵⁾

7.     Singapore (and most other common law jurisdictions) take a different view. That view is usually described as the “territorialist” or “jurisdictional” theory of arbitration.⁽⁶⁾ It treats every arbitration as connected to a particular jurisdiction – that is, the seat – so that the process is subject to a dual system of control.⁽⁷⁾ Several consequences follow from this. For one, the setting-aside of an award at the seat will generally be regarded as being universal in effect, so that once set aside at the seat there is no award to enforce.⁽⁸⁾ 

8.     The jurisdictional theory of arbitration also informs Singapore’s approach to the relitigation of issues in the post-award context. Where the national court at the seat has decided on an issue relating to the validity of an award, that decision will ordinarily preclude the relitigation of that issue in Singapore when that award is sought to be enforced here.

III.     The Republic of India v Deutsche Telekom AG

9.     I now turn to the recent Court of Appeal decision in The Republic of India v Deutsche Telekom AG [2024] 1 SLR 56 (“Deutsche Telekom”). In that case, Deutsche Telekom obtained an order permitting the enforcement of an award in Singapore against the Republic of India (“India”). India then applied to set that enforcement order aside on the ground that there was no valid arbitration agreement. The difficulty with India’s attempt at resisting enforcement of the award was that it had previously applied (unsuccessfully) to set the award aside in Switzerland. In those proceedings, the Swiss Federal Supreme Court rejected the same arguments that India was advancing in the Singapore enforcement proceedings. 

10.     Against that backdrop, the Court of Appeal applied the doctrine of transnational issue estoppel and held that India was precluded from contesting the enforcement of the award on grounds that had already been rejected by the Swiss seat court.⁽⁹⁾ The Court of Appeal also endorsed – albeit in obiter – what it termed the “Primacy Principle”.⁽¹⁰⁾  That principle holds that a seat court’s decision on matters going to the validity of an award should be treated as presumptively determinative, so that the onus is on the party resisting enforcement to prove otherwise.⁽¹¹⁾

11.     The Primacy Principle stems from the notion that the seat court occupies a special position within the system of international arbitration. It is the court that the parties have chosen to vest with supervisory jurisdiction over the arbitration, and so it would follow that the seat court’s decisions on matters pertaining to the validity of an award should be regarded as presumptively determinative.⁽¹²⁾ In Deutsche Telekom, it was said that the basis for the Primacy Principle lies in “the New York Convention read with the Model Law and the [International Arbitration Act], which recognise the special role and function of the seat court”.⁽¹³⁾

12.     The Court of Appeal identified three situations where the seat court’s decision might be held not to be determinative, namely: where that decision conflicts with the public policy of Singapore; where there were serious procedural flaws in the seat court’s decision-making process akin to breach of natural justice; and where the decision is shown to have been perverse. The Court of Appeal stressed that this list was not intended to be exhaustive.⁽¹⁴⁾ 

13.     Where the Primacy Principle is grounded in the scheme of the New York Convention and the Model Law, transnational issue estoppel is a common law doctrine of general application beyond the context of international arbitration. To be precise, it reflects a particular application of the issue estoppel doctrine which, together with cause of action estoppel and the rule in Henderson v Henderson (the “Henderson principle”), gives the law of res judicata most of its content.⁽¹⁵⁾ There is nothing internal to the logic of issue estoppel that compels a distinction between the seat court and enforcement courts. On that view, a prior decision of an enforcement court may conceivably give rise to an issue estoppel precluding the relitigation of issues not only before parallel enforcement courts, but also before the seat court.

14.     If that is correct, then the result would be incongruous with the primacy accorded to the seat court within the scheme of international arbitration. It would also be incongruous with the prerogative of other enforcement courts to consider issues afresh in deciding whether to enforce an award within its territory. Practical difficulties follow from this: there is the fear of award creditors first seeking enforcement in a jurisdiction most inclined to allow it, before using that first decision to bind courts elsewhere.

15.     The Court of Appeal acknowledged these difficulties in Deutsche Telekom and suggested that if the doctrine of transnational issue estoppel is to be disapplied in relation to prior enforcement court decisions, then that may be a result defensible on policy grounds.⁽¹⁶⁾ I will come back to the Court of Appeal’s observations on this point, but it suffices for now to note that whether a transnational issue estoppel can arise out of a prior enforcement court decision remains an open question in Singapore. 

IV.     The approaches taken in other jurisdictions

16.     I now turn to briefly consider how other jurisdictions have approached this issue. 

17.     I begin with the English courts. English law has fully embraced conventional res judicata principles in relation to the relitigation of issues post-award. Under English law, an issue estoppel may arise out of not only prior seat court decisions,⁽¹⁷⁾ but also prior enforcement court decisions (so far as the issues in question relate to the validity of the award).⁽¹⁸⁾ The English courts have also endorsed the Henderson principle as a further control which is “consistent with the policy of sustaining the finality of decisions of the supervisory courts”.⁽¹⁹⁾ 

18.     This approach of applying existing res judicata doctrine instead of adopting the Primacy Principle is what Jonathan Mance IJ endorsed by his concurring opinion in Deutsche Telekom. Mance IJ would have preferred to “rely on the tools which [were] already to hand, and not to give decisions of courts of the seat a specially elevated status in law in case of repeat challenges".⁽²⁰⁾ 

19.     Australian law, by contrast, has articulated and adopted a doctrine akin to the Primacy Principle. Following the Federal Court of Australia’s decision in Gujarat NRE Coke Ltd v Coeclerici Asia (Pte) Ltd (2013) 304 ALR 468 (“Gujarat NRE”), it appears that the Primacy Principle is the tool of choice for the Australian courts, with the court holding that “it will generally be inappropriate for this court, being the enforcement court of a Convention country, to reach a different conclusion on the same question of asserted procedural defects as that reached by the court of the seat of arbitration”⁽²¹⁾ In Gujarat NRE, the Federal Court declined to rule on whether a prior seat court decision could give rise to an issue estoppel in Australian enforcement proceedings.⁽²²⁾

20.     In an extra-judicial speech from 2014, then Federal Court Chief Justice James Allsop (and now member of the international bench of the SICC), who was also on the coram that decided Gujarat NRE, warned that “[a] doctrine that gives primacy to the first court (chosen by either party) to produce a decision about an award (whether that court is the supervising court or not) may undermine the internationality of the award, undermine the autonomy of the parties choosing the seat, and undermine the authority of the arbitrator and of the arbitral process."⁽²³⁾ 

V.     The way forward

21.     As has been seen, there is general agreement that weight should be accorded to the prior decisions of courts from other jurisdictions concerning the status or validity of an award. Unresolved reservations remain where the prior decision is one of an enforcing court and the matter now comes before the seat court on a challenge to the award. In this connection, I now touch on a distinction between two types of issues that has been explored in the case law.⁽²⁴⁾

A.     Forum-connected issues

22.     The first category of issues are issues that turn on the legal position in the forum court and are therefore uniquely within the forum court’s sphere of competence. This includes questions of arbitrability and public policy. I would describe these as “forum-connected” issues.

23.     Let me illustrate this by reference to the grounds of challenge raised in the Singapore High Court’s recent decision in Sacofa Sdn Bhd v Super Sea Cable Networks Pte Ltd and another [2024] SGHC 54 (“Sacofa”). The award there was made in a Singapore-seated arbitration involving a Malaysian project. As it happened, the Malaysian courts – specifically, the Kuala Lumpur High Court – had already decided that the award could be enforced in Malaysia.  In short, the award creditor had obtained an order permitting the registration and enforcement of the award in Malaysia before the award debtor applied to set the award aside in Singapore. The setting-aside application was brought on two grounds: the first was that the arbitrator had acted in excess of his jurisdiction; the second was that the award was contrary to the public policy of Singapore because it contained certain findings that were allegedly illegal under Malaysian law. The award debtor contended that it was contrary to the public policy of Singapore to enforce an award that was illegal under Malaysian law. In response, the award creditor submitted that the same arguments on illegality had already been considered and rejected by the Malaysian enforcement court, in which case there was a transnational issue estoppel precluding the relitigation of the same arguments in Singapore.

24.     The Judge agreed with the award creditor and held that the Malaysian enforcement court’s findings in respect of the illegality arguments gave rise to a transnational issue estoppel. He had no difficulty in accepting that the award debtor’s public policy objections were in reality “premised upon issues of Malaysian law and Malaysian public policy which the Malaysian courts [were] best placed to deal with.”(25) Accordingly, the Kuala Lumpur High Court having held that there was no illegality under Malaysian law barring enforcement of the award, the respondent to the arbitration was estopped from raising a ground premised on there being such illegality under Malaysian law. 

B.     Forum-neutral issues

25.    The second category of issues, which I will refer to as “forum-neutral issues”, are those that do not depend on the specific law of the forum for their determination. Forum-neutral issues have been referred to as “issues pertaining to the validity of the award”,⁽²⁶⁾ and they include questions of compliance with the agreed arbitral procedure, and whether an award was made in excess of jurisdiction.  

26.     It is not controversial that decisions of a seat court in relation to forum-neutral issues should be accorded significant – if not conclusive – weight. Even in jurisdictions where res judicata principles operate as the primary control against post-award relitigation, the seat court’s primacy within the scheme of international arbitration will almost invariably be acknowledged as a factor that must be given proper weight in the res judicata analysis. 

27.     More difficult is the question of whether and when an enforcement court’s prior decision on a forum-neutral issue should be regarded as preclusive. The answer for any jurisdiction may depend on its judicial policy. Ultimately, it would reflect a balance between considerations of finality, comity, and the forum court’s prerogative to determine issues for itself.

28.     The position in Singapore is not yet settled, but there is support for the view that enforcement court decisions on forum-neutral issues would not give rise to a transnational issue estoppel. 

29.     Returning to the case of Sacofa, I mentioned that the setting-aside application was also brought on grounds that the arbitrator had acted in excess of his jurisdiction. The award creditor pointed out that the same jurisdictional objections had been thrice considered and rejected by the Malaysian courts (ie, in the award debtor’s unsuccessful pre-award application for an anti-arbitration injunction; in the appeal against that decision; and again in the enforcement proceedings). On that footing, the award creditor argued that there was likewise a transnational issue estoppel precluding the relitigation of the same jurisdictional objections in Singapore.

30.     The Judge however rejected that argument. He reasoned that “[giving] preclusive effect to a prior enforcement (or non-enforcement) decision would undermine the role of the seat court and subvert the scheme underlying the New York Convention”.⁽²⁷⁾ He also took the view that the primacy of the seat court justified a departure from conventional issue estoppel doctrine.⁽²⁸⁾ 

VI.     Conclusion

31.     The question of how to weight prior court decisions in other jurisdictions concerning issues relating to the enforceability of awards is an important one. Relitigation of issues not only increases costs but is potentially a hindrance to expeditious justice. Unless and until this question is addressed by bodies like UNCITRAL, it is for national courts to grapple with it. I have described the emerging but not yet settled approach in Singapore and hope that this has been helpful both to your understanding of this issue and as a seed for the discussions to come this afternoon. 

32.     Thank you very much, and I wish this conference every success.