Chief Justice Sundaresh Menon: Paper delivered at the 8th Judicial Seminar on Commercial Litigation

8TH JUDICIAL SEMINAR ON COMMERCIAL LITIGATION

Transnational Relitigation and the Doctrine of Transnational Issue Estoppel

The Honourable the Chief Justice Sundaresh Menon^*
Supreme Court of Singapore

I.          Introduction

1.               Over the last several decades, the advance of globalisation has brought a proliferation of cross-border business and trade, which in turn has driven a steady increase in the incidence of international commercial disputes. This development has been accompanied by a disturbing trend: the growing number of disputants who attempt to reopen or relitigate issues that have already been decided in another forum. In this paper, such cases will be referred to as cases of transnational relitigation.

2.               Transnational relitigation raises difficult and nuanced issues for the courts, because it engages values that pull in different directions. On the one hand, the interests of litigants, the systemic integrity of the legal framework for international commercial disputes, and international comity¬ argue in favour of restricting or even prohibiting transnational relitigation. Yet, on the other hand, domestic courts have a core duty to safeguard and advance the rule of law and to deliver justice within their jurisdictions, and to uphold the public policy of their national states. These considerations require that courts retain a measure of discretion in determining the effect to be accorded to the outcomes of dispute resolution in other forums. These competing tensions call for a flexible approach that can accommodate the diverse considerations. At the same time, such an approach must be grounded in principle and cohere with existing legal concepts and doctrines.

3.               This paper contends that the doctrine of transnational issue estoppel enables common law courts to address transnational relitigation in an appropriate, flexible, and principled manner. The paper is structured as follows:

1. Section II introduces the problem of transnational relitigation, by examining two notable cases that illustrate the competing considerations in this field: the Putrabali⁽¹⁾ and Kabab-Ji⁽²⁾ cases.


2. Section III turns to relitigation in domestic legal systems, and provides a brief overview of the doctrine of res judicata including the principles relating to issue estoppel.


3. Section IV makes the case that common law courts should adopt and develop the principles of issue estoppel to address cases of transnational relitigation.


4. Finally, Section V focuses on the application of the doctrine of transnational issue estoppel to foreign judgments on the validity of an arbitral award. In particular, it discusses The Republic of India v Deutsche Telekom⁽³⁾ (“Deutsche Telekom”), a recent case in which the Singapore Court of Appeal applied the law of transnational issue estoppel and considered a separate principle conferring primacy to decisions of a seat court (the “Primacy Principle”).

II.         The problem illustrated: Putrabali and Kabab-Ji

4.               To appreciate the issues raised by transnational relitigation, it is useful to begin by considering two leading cases: Putrabali and Kabab-Ji.

5.               In Putrabali, an Indonesian company (“Putrabali”) sold a cargo of white pepper to a French company (“Rena”). The contract of sale provided for disputes to be referred to arbitration before the International General Produce Association (“IGPA”). During shipment, the cargo was lost in a shipwreck. Putrabali demanded payment for the cargo; but Rena refused to make payment.

6.               Putrabali commenced arbitral proceedings against Rena before an IGPA tribunal in London. The IGPA Board of Appeal issued an award in Rena’s favour, finding that Rena had not breached its contractual duties by withholding payment (the “2001 Award”). Putrabali then brought an appeal on certain points of law to the English Commercial Court. The court allowed the appeal, set aside the 2001 Award in part, and remitted the case to the IGPA Board of Appeal to assess the damages due to Putrabali.⁽⁴⁾ Subsequently, in August 2003, the tribunal issued a fresh award that awarded damages to Putrabali (the “2003 Award”).

7.               In the meantime, before the issuance of the 2003 Award,⁽⁵⁾ Rena applied for the 2001 Award to be enforced in France notwithstanding that that award had already been set aside in England. In September 2003, the first-instance French court granted Rena’s application. Putrabali appealed against this decision and separately applied for the 2003 Award to be enforced in France. However, the Paris Court of Appeal upheld the enforcement of the 2001 Award in France and rejected enforcement of the 2003 Award on the ground of res judicata (based on the prior enforcement of the 2001 Award).⁽⁶⁾ These rulings were endorsed by the Court of Cassation of France. In affirming the enforcement of the 2001 Award, the Court of Cassation restated the position under French law that an arbitral award is a “decision of international justice” that is independent of any national legal order, and the validity of the award is to be determined by reference to the law of the enforcement jurisdiction. Under French law, the setting-aside of an arbitral award was not a ground for refusing enforcement of the award. Hence, the court held that Rena was entitled to enforce the 2001 Award in France.⁽⁷⁾

8.               The outcome in Putrabali was problematic for at least three reasons.

1. First, the result was plainly inequitable to Putrabali. It had fought a protracted legal battle to obtain payment under the contract, and ultimately succeeded before both the tribunal and the court of the seat of the arbitration. Yet, as Lord Mance noted in a trenchant critique of the Court of Cassation’s decision, the French courts allowed Rena’s “cunning riposte”⁽⁸⁾ to its multiple legal defeats by granting enforcement of the 2001 Award and then barring enforcement of the 2003 Award. The upshot was that Putrabali was unable to realise the fruits of its victory in the forum in which Rena, a French company, presumably had most of its assets.


2. Second, from a systemic perspective, the outcome in Putrabali led to incoherence in the framework for international commercial dispute resolution.⁽⁹⁾ There were two conflicting decisions on the validity of the 2001 Award in England and France respectively; and worse, there was no cogent or readily appreciable basis for this divergence.


3. Third, the French courts’ refusal to consider the English court’s decision to set aside the award is at odds with international comity – a point also made by Lord Mance.⁽¹⁰⁾ As I explain below,⁽¹¹⁾ comity calls for courts to respect the decisions of other courts; and I suggest that, in the context of international arbitration, comity requires a court that considers whether to enforce an arbitral award (namely, an “enforcement court”), at the minimum, to examine a prior decision on the validity of an arbitral award by the seat court, and explain any departures from that decision. Yet, the Court of Cassation appears to have upheld the enforcement of the 2001 Award without even considering the English court’s decision.

9.               These three troubling aspects of Putrabali reflect a number of factors that count against transnational relitigation, and I will elaborate on these below.⁽¹²⁾ What underlay the outcomes in Putrabali was the different views taken by the French and the English courts to the question of what law governs the status and validity of arbitral agreements and awards. This was even more clearly illustrated in Kabab-Ji. There, a Lebanese company (“Kabab-Ji”) granted a licence to a Kuwaiti company (“Al Homaizi”) to operate Kabab-Ji’s franchise in Kuwait, under a contract which provided for disputes to be settled by arbitration in Paris. When disputes arose between the parties, Kabab-Ji commenced arbitral proceedings in Paris against the Kout Food Group (“KFG”), the parent company of Al Homaizi. A key issue in the arbitration was whether KFG had become a party to the licence contract and the arbitration agreement. This turned on whether the governing law of the arbitration agreement was French law or English law. The majority of the tribunal ruled that French law was the governing law and that, under French law, KFG had become party to the arbitration agreement and the main contract. The majority accordingly issued an award ordering KFG to pay damages to Kabab-Ji.⁽¹³⁾

10.               The parties then began parallel proceedings regarding the award in France and England. KFG applied to the Paris Court of Appeal to set aside the award, on the basis that the tribunal had no jurisdiction over the dispute because KFG was not a party to the arbitration agreement. On the other hand, Kabab-Ji applied for the award to be enforced in England. Ultimately, the French and English courts reached diametrically opposed conclusions on the governing law of the arbitration agreement and the validity of the arbitral award. The French courts held that, under a “substantive rule of international arbitration”, the validity of an arbitration agreement was generally to be assessed based on the common intention of the parties, rather than any national law. KFG had not proved that the parties had a common intention that English law would govern the arbitration agreement. Hence, French law, as the law of the seat, was the governing law of the arbitration agreement. Applying French law, KFG had become a party to the arbitration agreement by participating in the licence contract.⁽¹⁴⁾ Hence, the French courts dismissed Kabab-Ji’s application to set aside the award. In contrast, the English courts found that English law governed the arbitration agreement and under English law, Kabab-Ji was not a party to the licence contract or the arbitration agreement.⁽¹⁵⁾ Accordingly, the English courts refused to recognise and enforce the award in England.

11.               Yet notably, in arriving at different conclusions on the validity of the arbitral award, the English and French courts adopted a fundamentally similar approach. Both sets of courts focused on applying the relevant conflict of laws rules of their jurisdictions to determine the law of the arbitration agreement. Further, and significantly, both courts held that each other’s rulings on the issue were irrelevant to their own decisions.⁽¹⁶⁾ Lords Hamblen and Leggatt, delivering the judgment of the UK Supreme Court, explained this in addressing why the English action should not be adjourned to await the outcome of the French action. They stated as follows:

1. Where there was an application before a foreign seat court to set aside an arbitral award and concurrently, proceedings to enforce the award in England, and both actions involved the same issues governed by the law of the seat, the English action would likely be adjourned pending the seat court’s decision, for two reasons. First, it was sensible for the foreign court to decide issues of foreign law. Second, an adjournment would avoid the risk of conflicting judgments.⁽¹⁷⁾


2. However, these two factors did not apply where the foreign and enforcement courts would apply different systems of law to determine the dispute. This situation arose in Kabab-Ji: the English courts applied English law to determine the dispute, whereas the French courts applied French law. The UK Supreme Court noted that in such cases, the issues before both courts would be different.⁽¹⁸⁾ Hence, the first factor would not apply because the foreign court’s determination of the issue would be irrelevant to the decision of the English court, which “must apply a different system of law to determine the issue”.⁽¹⁹⁾ And the second factor would also not apply because, where the two systems of law applied by the two courts led to different conclusions, “the risk of contradictory judgments cannot be avoided [emphasis added]”⁽²⁰⁾.

12.               This last point reflects a key norm that prevails in transnational relitigation. Although there is always a desire and often a need for consistent outcomes across the course of international commercial dispute resolution, this cannot always be achieved because national courts retain a core duty to apply the laws of their jurisdictions where this is mandated by their domestic conflict of laws rules and also to uphold the public policies of their states. Hence, in some cases, divergent results may be unavoidable because laws and conceptions of public policy vary. As I elaborate below, the law on transnational relitigation must account for these considerations and preserve sufficient space for national courts to discharge their constitutional responsibilities.⁽²¹⁾

13.               Transnational relitigation will typically engage several competing policy factors and this can have grave implications for the parties. How then can common law courts address such cases in a principled way that accounts for these considerations? I suggest that the answer lies in applying and developing a set of principles which are simply an extension of a familiar doctrine that controls relitigation in the domestic context: the doctrine of issue estoppel.

III.          Relitigation in the domestic context

14.               Relitigation is an age-old challenge that has vexed virtually every legal system. Over time, most legal systems developed some version of the doctrine of res judicata to limit relitigation. The latter doctrine traces back to Roman law and derives from the Latin maxim res judicata pro veritate acciptur (a matter adjudged is taken as true), which was enshrined in the Digest.⁽²²⁾ But this is not just a Western concept. The Privy Council has noted that the principle is found in Hindu jurisprudence,⁽²³⁾ and it was described by Judge Philip Jessup as a “general principle of law recognised by civilised nations”.⁽²⁴⁾

15.               The concept of res judicata is founded on two fundamental policies.⁽²⁵⁾ The first is the private interests of litigants in not being vexed by relitigation of the same cause. This was explained by Lord Coke in Ferrer v Arden,⁽²⁶⁾ who observed that without a doctrine of res judicata, “great oppression might be done under colour and pretence of law …a rich and malicious man would infinitely vex him who hath right by suits and actions; and in the end (because he cannot come to an end) compel him … to … relinquish his right [emphasis added]”.⁽²⁷⁾

16.               The second policy behind res judicata is the public interest in securing the end of litigation. This is based on two considerations underlying the integrity of a legal system.

1. First, scarce adjudicative resources should not be expended in reopening a matter which has already been fairly adjudicated.⁽²⁸⁾ This reflects the ideal of “collective proportionality”, which calls for each case to draw on no more than a proportionate share of adjudicative resources, so that the justice system as a whole can determine all disputes brought to court in an efficient and appropriate way.⁽²⁹⁾


2. Second, and more fundamentally, limiting relitigation is essential for a justice system to perform its key function of resolving disputes. If prior judicial decisions could be subject to unceasing challenge, those decisions would lose their authority, and this would “throw the whole system of justice into disrepute”.⁽³⁰⁾

17.               In common law systems, the law of res judicata comprises at least three main doctrines: cause of action estoppel; issue estoppel; and the “extended” doctrine of res judicata, also known as the rule in Henderson v Henderson⁽³¹⁾.⁽³²⁾

1. Cause of action estoppel bars a party from asserting or denying, as against another party, a “particular cause of action, the non-existence or existence of which has been determined by a court of competent jurisdiction in previous litigation between the same parties”.⁽³³⁾ This is an absolute bar with narrow exceptions for fraud or collusion.⁽³⁴⁾


2. Issue estoppel prevents a party from rearguing a factual or legal issue that was decided by a court of competent jurisdiction in prior litigation between the same parties.⁽³⁵⁾ This is subject to the same exceptions as cause of action estoppel, and a further exception known as the Arnold exception⁽³⁶⁾ which I elaborate on below.⁽³⁷⁾


3. Finally, the “extended” doctrine of res judicata precludes a party, in the absence of “special circumstances”, from raising points that were not decided in prior proceedings, but which should have been raised by the party in those proceedings.⁽³⁸⁾ This is a more flexible mechanism: the court will consider all the circumstances of the case, including “whether there is fresh evidence that might warrant relitigation or whether there are bona fide reasons why a matter was not raised in the earlier proceedings”,⁽³⁹⁾ before holding that a party is barred from raising an issue by virtue of this doctrine.

18.               This paper focuses on the doctrine of issue estoppel. The key principles relating to the domestic variant of this doctrine are well-established.

1. There are four main requirements for issue estoppel to apply: (a) a final and conclusive judgment on the merits, issued by (b) a court of competent jurisdiction, with (c) identity between the parties, as well as (d) identity of issues or subject matter.⁽⁴⁰⁾


2. In general, issue estoppel applies even if the earlier decision was wrong because “the whole point of the doctrine of res judicata is that erroneous decisions must nevertheless be given due effect”.⁽⁴¹⁾


3. That said, issue estoppel is subject to a qualification known as the Arnold exception. This enables a litigant to reopen an issue that was previously determined, in exceptional cases. Under Singapore law, the Arnold exception only applies where (i) the prior decision directly affects the future determination of the parties’ rights; (ii) that decision was clearly wrong; (iii) the error in the decision arose from the fact that a point was not taken before the court that made the decision, and the point could not reasonably have been taken; (iv) any rights accruing under the decision will not be clawed back, and any effects of the decision will not be undone; and (v) great injustice would arise if issue estoppel applied.⁽⁴²⁾

19.               In sum, issue estoppel is a well-established doctrine that limits relitigation within common law legal systems. I suggest common law courts may readily adapt and extend this doctrine to manage transnational relitigation.

IV.          The case for transnational issue estoppel

20.               I develop this point in three steps as follows:

1. I will first describe the compelling and yet nuanced need for finality in international commercial dispute resolution, which reflects a range of competing values.


2. I will then outline the doctrine of transnational issue estoppel under Singapore law.


3. I will finally explain why common law courts should deploy the doctrine of transnational issue estoppel to address and manage transnational relitigation.

A. The need for finality in international commercial dispute resolution

21.               Finality is a fundamental desideratum in international commercial dispute resolution. This flows from the two main policies underlying the doctrine of res judicata (the interests of litigants and the systemic need to limit relitigation),⁽⁴³⁾ which are equally engaged by transnational relitigation, as well as the demands of international comity.

22.               First, as with relitigation within national legal systems, the private interests of litigants in securing the efficient resolution of disputes are undermined by transnational relitigation. Indeed, this concern has even greater force in the transnational context in view of the substantial costs that are typically associated with participating in multiple proceedings across different jurisdictions, and the significant delays to the final resolution of a dispute that are often entailed by such proceedings. The facts of Putrabali illustrate this: more than seven years elapsed between the emergence of the dispute in February 2000⁽⁴⁴⁾ and the final decision of the Court of Cassation in France in June 2007,⁽⁴⁵⁾ and Putrabali must have incurred considerable legal costs in participating in the various French proceedings. In this light, transnational relitigation can often be vexatious and oppressive for successful litigants. This weighs in favour of limits being placed upon this practice.

23.               Further, transnational relitigation undermines the systemic integrity of the legal framework for international commercial disputes. I suggest this framework should be understood as a system, which I refer to as the transnational system of commercial justice (the “TSCJ”), for two principal reasons.

24.               First, this view of the legal framework for international commercial disputes reflects the reality of that framework, which has several systemic features.⁽⁴⁶⁾ One such feature is the growing convergence in various aspects of international commercial law. For example, there is, to a growing extent, substantial similarity in the procedural rules governing international commercial dispute resolution – including the principles on the approach to jurisdiction and arbitration agreements, as well as the law on the enforcement of arbitral awards – across many jurisdictions.⁽⁴⁷⁾ Another systemic aspect of the TSCJ is the body of rules and codes that strives to promote its coherence, by ensuring that it produces consistent or at least compatible outcomes in various jurisdictions. One example of this is the Judicial Insolvency Network Guidelines. This sets out a framework for international collaboration between insolvency courts in cross-border insolvency proceedings, which can significantly reduce inconsistencies in the treatment and outcomes of such cases.⁽⁴⁸⁾ These systemic aspects of the TSCJ unify its parts and justify our viewing it as a system.⁽⁴⁹⁾

25.               Second, the legal framework for international commercial disputes should be seen and approached as a system because this can advance two valuable goals.

1. A systemic perspective of international commercial law and dispute resolution helps promote cross-border trade because such a perspective leads courts and other adjudicative bodies to strive to reduce inconsistencies in legal rules and outcomes, and thereby enhance convergence in international commercial law. This will in turn reduce legal uncertainty and transaction costs for businesses, and so promote cross-border commerce.⁽⁵⁰⁾


2. Further, a systemic view of the legal framework for international commercial disputes would enhance the consistent development of important transnational legal norms. Our world faces several challenges such as climate change, which are giving rise to a range of transnational legal issues that are already being raised before arbitral tribunals and commercial courts.⁵¹⁽⁾ The determination of these issues can generate transnational legal norms to guide businesses and thereby contribute to the global response to challenges like climate change. If the institutions that develop and apply international commercial law were to take a systemic approach to the legal issues arising from such global challenges, this would foster the coherent and coordinated development of the transnational legal norms that will help to address those challenges.

26.               For these reasons, the legal architecture that has emerged to deal with international commercial disputes should be seen as a system. And once we adopt this view, it is plain that transnational relitigation undermines the integrity of that system at two levels.

1. At the first level, similar to relitigation within national legal systems,⁽⁵²⁾ transnational relitigation can lead to the disproportionate and wasteful use of scarce adjudicative resources, which could be spent more productively on resolving fresh disputes. In other words, transnational relitigation reduces the capacity of the overall system to determine disputes in an appropriate and efficient way.⁽⁵³⁾


2. At the second level, transnational relitigation impairs the coherence of the TSCJ. This is because the order and unity of the system are somewhat compromised if a dispute that has been determined by one arm of the system can be reopened before a different arm of that system. And, the problem is exacerbated if these efforts result in conflicting decisions because such discordant results undercut the integrity of the overall system.

27.               Further, there is a third factor that counts against transnational relitigation: international comity. Comity requires that the courts respect the decisions of other courts, which are made in the exercise of the sovereign power of their respective states.⁽⁵⁴⁾ Hence, if a dispute has already been determined by a foreign court, the forum court should generally be slow to allow the same issues to be relitigated before it, because this would not accord due respect to the prior decision.

28.               In sum, there are three compelling reasons to place limits on transnational relitigation: the interests of litigants, the systemic need to control relitigation, and international comity. But critically, there is a factor that pulls in the other direction. This arises from the fact that the TSCJ is not a supranational legal order, but a decentralised edifice that is superintended by national courts, who play a central role in operating the system.⁽⁵⁵⁾ As national institutions, domestic courts have a foremost constitutional duty to apply the law and to deliver justice in their jurisdictions, and to advance the public policies of their respective states.⁽⁵⁶⁾ This is a central judicial responsibility, which cannot be abrogated.

29.               It follows that where the laws of the forum require a domestic court, as in Kabab-Ji, to take a different approach to the issues than that adopted by the foreign court, or to reopen the dispute in view of its own conceptions of justice or public policy or of the governing law, the domestic court must hear and determine the dispute even if this would permit the transnational relitigation of seemingly similar issues. Further, in such cases, inconsistent legal results across different jurisdictions may be inescapable because there are variations in laws, conceptions of justice and public policy between different states. This is an inevitable consequence of the decentralised nature of the TSCJ and the central role of domestic courts in operating that system.

30.               Hence, while there is a strong need for finality in the resolution of international commercial disputes, this must be balanced against the role of the courts in upholding the rule of law, delivering justice, and advancing the public policy of their jurisdictions.

B. The elements and contours of transnational issue estoppel

31.               In Singapore, the doctrine of transnational issue estoppel was closely examined by a five-judge bench of the Court of Appeal in Merck Sharp & Dohme Corp (formerly known as Merck & Co, Inc) v Merck KGaA (formerly known as E Merck)⁽⁵⁷⁾ (“Merck Sharp”). In that case, the parties were two distant branches of a German business that was established in the 17^th century under the name “E Merck”. In the 1970s, the parties entered into two documents including one (the “1970 Agreement”) to govern their use of the “Merck” name. Subsequently, the parties became embroiled in legal proceedings in at least nine jurisdictions over their use of the “Merck” name, including in England.⁽⁵⁸⁾

32.               Between 2014 and 2017, the English courts handed down three decisions in the proceedings in England between the parties. Among other things, the courts (a) made rulings on the meaning and effect of the 1970 Agreement (in particular, clause 7 of the 1970 Agreement) and (b) found that the appellant in Merck Sharp (“MSD”) had breached the 1970 Agreement (the “English Rulings”).⁽⁵⁹⁾

33.               The respondent in Merck Sharp (“MK”) and another party then began proceedings in Singapore against MSD and three other parties for trademark infringement, passing off and breach of contract.⁽⁶⁰⁾ This gave rise to the question of whether MSD was bound by the English Rulings in the Singapore proceedings. On appeal, the sole issue was whether MSD was bound by the English courts’ interpretation of clause 7 of the 1970 Agreement,⁽⁶¹⁾ which in turn depended on whether the doctrine of transnational issue estoppel applied.

34.               The Court of Appeal began by affirming the existence of the doctrine, and noting the competing tensions raised by transnational relitigation.⁽⁶²⁾ The court opined that a proper balance could be struck between these tensions by (a) applying the elements of transnational issue estoppel carefully, (b) delineating the outer limits of the doctrine and (c) potentially taking a different approach to the Arnold exception in cases of transnational relitigation.⁽⁶³⁾ The court elaborated on each of these aspects in turn.

35.               First, the court set out the elements of transnational issue estoppel. The first element is a foreign judgment that is capable of being recognised, under the conflict of laws rules of the forum. Under the common law,⁽⁶⁴⁾ this required (a) a final and conclusive decision on the merits, by (b) a court of competent jurisdiction (meaning a court with international jurisdiction over the parties and the subject-matter of the dispute) and that (c) no defences to recognition of the judgment apply.

36.               The court elaborated on the first element as follows:

1. Where there were multiple foreign judgments, the judgment that was issued first would be recognised for the purpose of transnational issue estoppel.⁽⁶⁵⁾


2. A foreign judgment would not be recognised if there was a prior or subsequent inconsistent local judgment between the same parties. The court left open the proper approach where a foreign judgment is issued when local proceedings on the same or substantially the same subject matter have been commenced and are pending.⁽⁶⁶⁾


3. The court also left open the issue of whether reciprocity (meaning that recognition of Singapore judgments would be accorded by the foreign forum) was a condition for the recognition of a foreign judgment at common law.⁽⁶⁷⁾

37.               The second and third elements of transnational issue estoppel are identity of issues and identity of parties. The court held that the legal principles regarding these elements are the same as those that apply in the context of domestic issue estoppel.⁽⁶⁸⁾ That said, a court should take care in identifying and analysing the issues determined in a foreign decision, in assessing whether there was identity of issues and, more broadly, whether transnational issue estoppel should apply.

1. First, the court should determine whether the specific issue said to be the subject matter of an issue estoppel was necessary, rather than being merely collateral, to the foreign judgment. Issue estoppel would only arise in the former scenario.⁽⁶⁹⁾


2. Second, the court should evaluate whether the specific issue in question had been finally and conclusively determined in the foreign jurisdiction, because a foreign judgment should not be given greater effect in the forum than it has in its home jurisdiction (such as would be the case where the foreign forum does not recognise a concept of issue estoppel).⁽⁷⁰⁾ I note that this last proposition has been criticised.⁽⁷¹⁾ Yet, it reflects settled law⁽⁷²⁾ and is sound because the factors that count against transnational relitigation⁽⁷³⁾ have much less force where the foreign court’s determination of the issues may be reopened in that forum.


3. Third, the court should determine whether the party against whom issue estoppel is invoked had the opportunity to raise the issue. If not, issue estoppel might not arise.⁽⁷⁴⁾

38.               Having set out the three main elements of transnational issue estoppel, the court opined that these were necessary but insufficient conditions for transnational issue estoppel to arise.⁽⁷⁵⁾ Beyond these conditions, the doctrine of transnational issue estoppel had certain outer limits, and no estoppel would arise if these limits were breached. The court identified four potential outer limits:

1. First, the doctrine should not apply to an issue that the forum court ought to determine under its own law.⁽⁷⁶⁾


2. Second, the doctrine should not apply to territorially limited foreign judgments (such as findings on the public policy of the foreign forum).⁽⁷⁷⁾


3. Third, the doctrine might not apply to bar a defendant in foreign proceedings from raising an issue that it had bona fide reasons for not (fully) contesting in the foreign forum.⁽⁷⁸⁾


4. Fourth, the doctrine might not apply where (recognition of) the foreign judgment would conflict with the forum’s public policy. In this regard, the court left open whether such a public policy defence could be invoked where the foreign court had erred in identifying or applying the law of the forum, or whether such an error would constitute a standalone exception to issue estoppel.⁽⁷⁹⁾

39.               Finally, the court considered whether and how the Arnold exception might apply in the context of transnational issue estoppel. This is a vexed issue because under Singapore law, the Arnold exception only applies if the earlier decision was “clearly wrong”;⁽⁸⁰⁾ yet there is an established principle that a foreign judgment cannot be questioned on its merits once it has been recognised by the forum court (the “Godard principle”).⁽⁸¹⁾ The court discussed four possible approaches to this conundrum, which were canvassed by the amicus curiae:⁽⁸²⁾

1. First, the Arnold exception might simply not apply to foreign judgments on the basis that it was not possible to establish that a prior foreign decision was “clearly wrong” given the Godard principle. Yet on this approach, foreign judgments would have stronger preclusive effect than domestic judgments. The court observed that this was undesirable in principle, because it seemed incompatible with the court’s role as the custodian of the rule of law to yield readily to a foreign judgment which appeared to have been wrongly decided.


2. On the second approach, the Godard principle might be construed as applying only to the cause of action, so that a litigant could argue that the foreign court’s decision on a specific issue was “clearly wrong” even if the foreign court’s determination of the cause of action was unimpeachable. The court noted the amicus curiae’s submission that this approach was also problematic because, among other reasons, the line between a cause of action and an issue is not always easily drawn.


3. Third, the Arnold exception could be seen as a broad discretionary exception in the context of foreign judgments (in contrast to the restricted nature of the exception in relation to domestic issue estoppel). However, this could threaten the integrity of the doctrine of transnational issue estoppel, and lead to legal uncertainty which might undermine the interests of litigants and make it more difficult for courts to apply the doctrine.


4. Finally, the Godard principle could be interpreted as permitting a foreign decision on an issue to be challenged in the forum court, if such a challenge could also be mounted in the foreign court.

40.               The court observed that all these four approaches faced certain difficulties and left the issue of the proper approach for determination in a future case.⁽⁸³⁾

41.               Returning to the facts in Merck Sharp, the arguments on appeal centred on (a) whether there was identity of issues between the English and Singapore proceedings and (b) whether the English Rulings were territorially confined to England, and thus incapable of giving rise to transnational issue estoppel. On the first point, the Court of Appeal found there was identity of issues because, in relation to the interpretation of clause 7 of the 1970 Agreement, the issue in the court below was the same as that determined in the English proceedings.⁽⁸⁴⁾ Further, the court found that the English court’s ruling on the interpretation of clause 7 was not territorially limited to England.⁽⁸⁵⁾ Hence, the court found that the doctrine of transnational issue estoppel applied and dismissed the appeal.

C. The utility of the doctrine of transnational issue estoppel

42.               I suggest that common law courts should deploy and develop the doctrine of transnational issue estoppel to address and control transnational relitigation, for three principal reasons.

43.               First, as a matter of authority, transnational issue estoppel is supported by an established line of cases which have recognised that foreign judgments can give rise to issue estoppel. This principle has been recognised since the 1967 decision of the House of Lords in Carl Zeiss Stiftung,⁽⁸⁶⁾ and its contours and limits have been developed in later cases.⁽⁸⁷⁾ Hence, transnational issue estoppel is not a novel or unorthodox concept, but a well-established common law doctrine.

44.               Second, as demonstrated by the analysis of the doctrine above, transnational issue estoppel is a fact-sensitive and flexible tool. Hence, it can typically enable the courts to strike an appropriate balance between the applicable policy considerations in a given case. In particular, the doctrine affords the courts sufficient space to discharge their constitutional role of safeguarding the rule of law, delivering justice, and securing the applicable public policy. To elaborate:

1. The doctrine would not bar transnational relitigation in an extreme case like Kabab-Ji: it would permit the forum court to reopen an issue that was determined by the foreign court applying foreign law, but which the forum court ought to decide applying its own law, on two grounds. First, as the UK Supreme Court noted in Kabab-Ji,⁽⁸⁸⁾ issue estoppel would not arise in such a case because there would be no identity of issues since the issues raised in the two forums would be framed differently, by reference to different laws. Second, the first potential outer limit outlined in Merck Sharp⁽⁸⁹⁾ would preclude transnational issue estoppel from arising in such cases.


2. The doctrine could enable the forum court to correct (manifest) legal errors made by the foreign court regarding the law of the forum, and address any injustice arising from such errors. This could be achieved under (i) the potential public policy exception or an independent exception or (ii) a modified version of the Arnold exception.⁽⁹⁰⁾


3. The doctrine would empower the forum court to safeguard the public policy of its jurisdiction at two levels.
   1. The incorporation of the forum’s conflict of laws rules into the doctrine – in particular, the defences to recognition of a foreign judgment⁽⁹¹⁾ – enables the forum court to give effect to domestic public policy. For example, a foreign judgment that was procured by fraud or one that was the product of an unfair trial would not be recognised by the forum court, and would therefore not give rise to an estoppel.
   
   
   2. To the extent that the defences to recognition of a foreign judgment do not capture all the applicable policy factors, these could be accommodated under the standalone public policy exception proposed in Merck Sharp.⁽⁹²⁾

45.               Third, transnational issue estoppel is not an ad hoc or unstructured tool, but a principled doctrine comprising a clear and defined set of principles. Hence, it can provide a measure of legal certainty to international commercial parties and promote broad consistency in how the courts determine cases of transnational relitigation. In this light, the doctrine of transnational issue estoppel emerges as a well-suited control mechanism for transnational relitigation.

V.          Transnational issue estoppel in the context of international arbitration

46.               This leads to the application of transnational issue estoppel in the specific context of international arbitration. One important and vexed issue in international arbitration is the approach that an enforcement court should take to a prior decision of the seat court or another enforcement court on the validity of an arbitral award. I will address this issue in three main parts:

1. I will first outline the legal and conceptual background to this issue, before providing an overview of the approaches that various common law jurisdictions have taken when addressing this issue.


2. I will then discuss the recent decision of the Singapore Court of Appeal in Deutsche Telekom where the court applied the doctrine of transnational issue estoppel and also considered the Primacy Principle in the context of a prior decision by a seat court on the validity of an arbitral award.


3. I will finally outline some considerations that may be relevant in assessing whether the doctrine of transnational issue estoppel should apply to decisions of an enforcement court (an issue that was left open in Deutsche Telekom).

A. Background

i. The legal and conceptual background

47.               The starting point is Article V(1)(e) of the New York Convention.⁽⁹³⁾ This provides that recognition and enforcement of an arbitral award “may” be refused where the award has been “set aside … by a competent authority of the country in which, or under the law of which, that award was made”. On a plain reading, this provision permits an enforcement court to refuse to enforce an award on the basis that it has been set aside, but leaves the enforcement court the discretion to decide whether to do so.⁽⁹⁴⁾ But Article V(1)(e) does not provide any guidance as to how this discretion should be exercised. Further, the New York Convention does not address how the seat court or another enforcement court should treat the prior decision of an enforcement court on the validity of an arbitral award.

48.               In grappling with this conundrum, many commentators have referred to the continuing conceptual debate in international arbitration between two schools of thought, which contributed to the outcomes in Putrabali and Kabab-Ji.⁽⁹⁵⁾ On the “territorialist” view, the legal validity of an arbitral award derives from the legal order of the seat of the arbitration. Hence, if the award is set aside, it becomes a legal nullity and cannot be enforced in any jurisdiction because there is nothing left to enforce. By contrast, on the “delocalisation” theory, the legal force of an arbitral award stems from an international legal order that transcends national legal systems. This entails that the effect of a seat court’s decision to set aside an award is confined to the jurisdiction of the seat alone – the award remains enforceable in other jurisdictions.⁽⁹⁶⁾ This conception of international arbitration underpinned the Court of Cassation’s decision in Putrabali and was relevant to why the UK Supreme Court in Kabab-Ji considered that different systems of law applied in England and France to govern the question of whether KFG had become a party to the arbitration agreement.⁽⁹⁷⁾

49.                This paper does not propose to venture into this debate, save to note that the position set out in Deutsche Telekom indicates that the correct position lies in between the two extremes.⁽⁹⁸⁾ In a nutshell, on this approach, the enforcement court will generally give significant weight to a seat court’s decision on the validity of an arbitral award, by applying the doctrine of transnational issue estoppel or the Primacy Principle, save in exceptional cases. This reflects a position between territorialism and delocalisation, though one that falls much closer to the former. Before I explain the position in Singapore, I set out the law in three common law jurisdictions – England, Australia, and the United States (the “US”) – on the approach to a foreign judgment on the validity of an arbitral award.

ii. The law in England, Australia, and the US

50.               In England, the courts apply the doctrine of transnational issue estoppel to prior decisions of a foreign court on the validity of an arbitral award, regardless of whether the foreign court was the seat court or an enforcement court.

51.               This approach was adopted in the leading case of Diag Human SE v The Czech Republic.⁽⁹⁹⁾ There, the claimant applied for an arbitral award issued in the Czech Republic to be enforced in England. Before initiating the English action, the claimant had filed a similar application for enforcement of the award in Austria. The Supreme Court of Austria dismissed that application, on the basis that the award had not become binding on the parties (see Article V(1)(e) of the New York Convention) because a process for review of the award, which was stipulated in the arbitration agreement, had not concluded. In the English proceedings, the defendant argued that issue estoppel arose from the Austrian court’s finding that the award had not become binding, so that the claimant was precluded from contending otherwise before the English court. Eder J accepted this submission and accordingly dismissed the enforcement application.⁽¹⁰⁰⁾

52.               In Diag Human, the prior decision on the arbitral award was made by an enforcement court. The English courts have since held that decisions of a seat court may similarly give rise to transnational issue estoppel. For example, in Carpatsky Petroleum Corp v PJSC Ukrnafta (“Carpatsky”),⁽¹⁰¹⁾ the claimant applied for an arbitral award issued in an arbitration seated in Sweden to be enforced in England. The defendant resisted the application on the basis of an alleged breach of due process by the arbitral tribunal. However, a similar complaint had been rejected by both a Swedish court (the seat court) and a US District Court (another enforcement court). Butcher J held that issue estoppel arose from the decisions of both courts, to bar the defendant from raising its due process challenge to the award in England.⁽¹⁰²⁾

53.               However, in England, transnational issue estoppel will generally not arise from decisions of a foreign court on issues of public policy. This was recognised in Yukos Capital Sarl v OJSC Rosneft Oil Co (No 2) (“Yukos Capital”),⁽¹⁰³⁾ where the English Court of Appeal ruled that a Dutch decision that found that a Russian judgment was tainted by a partial judicial process, and was therefore contrary to public policy, did not give rise to an issue estoppel in England. The English court reasoned that whether a judgment was the product of a partial process raised issues of public order or public policy, which were “inevitably different in each country”.⁽¹⁰⁴⁾ There could be no identity of issues between decisions in different jurisdictions about such issues, and issue estoppel would therefore not arise.

54.               Yukos Capital concerned the effect of a foreign decision that recognising a foreign judgment would be contrary to public policy. But the reasoning in that case has been extended in England to foreign decisions that recognise or refuse to recognise an arbitral award on public policy grounds. The English courts have found that such decisions do not give rise to issue estoppel in England, because they would not and cannot determine issues of English public policy which are matters for the English court to decide.⁽¹⁰⁵⁾

55.               It follows that foreign court decisions on issues of arbitrability would generally not give rise to issue estoppel in England, because arbitrability is fundamentally a matter of public policy,⁽¹⁰⁶⁾ and domestic public policy differs between states and are for the courts of each jurisdiction to decide.⁽¹⁰⁷⁾ This was noted by Eder J in Diag Human.⁽¹⁰⁸⁾ Further, in England, a foreign judgment that is regarded as “perverse”, on the ground that the law that was applied in and formed the basis of the foreign judgment is “at variance with generally accepted doctrines of private international law”, will not found an issue estoppel.⁽¹⁰⁹⁾

56.               Next, in Australia, it appears that the courts do not favour the doctrine of transnational issue estoppel in approaching foreign court decisions on the validity of an arbitral award. Rather, they accord primacy to or, in effect, place very significant weight on decisions of the seat court. This is what was referred to as the Primacy Principle in Deutsche Telekom.Under this approach, an enforcement court will not depart from decisions of the seat court save in exceptional cases.

57.               This was endorsed by the Full Court of the Federal Court of Australia in Gujarat NRE Coke Ltd and another v Coeclerici Asia (Pte) Ltd (“Gujarat (Full Court)”).⁽¹¹⁰⁾ There, the respondent had obtained an arbitral award against the appellant in a London-seated arbitration. The appellant applied unsuccessfully in England for the award to be set aside on the basis of an alleged breach of due process in the arbitration. The respondent then applied for the award to be enforced in Australia. This was opposed by the appellant, who raised the same due process challenge that had been ventilated in England.

58.               At first instance, Foster J found that issue estoppel arose from the English court’s decision so as to bar the respondent from raising its due process challenge in the Australian court.⁽¹¹¹⁾ Further, Foster J held that even if no issue estoppel arose, it would generally be inappropriate for an enforcement court to reach a different conclusion from the seat court on the same question.⁽¹¹²⁾ In other words, Foster J applied both the doctrine of transnational issue estoppel and recognised the Primacy Principle. On appeal, the Full Court of the Federal Court of Australia endorsed the Primacy Principle but expressed no view on the applicability of the doctrine of transnational issue estoppel in Australian proceedings regarding the enforcement of an arbitral award. The court explained that the issue of whether that doctrine applied in this context was “of importance and of potential difficulty”, and had not been settled by any higher authority in Australia.⁽¹¹³⁾ It was unnecessary to decide this issue, because the Full Court agreed that the Primacy Principle should apply. The court approved the following account of that principle set out in Minmetals Germany GmbH v Ferco Steel Ltd:⁽¹¹⁴⁾

59.               This endorsement of the Primacy Principle was recently reaffirmed by the Federal Court of Australia in Hub Street Equipment Pty Ltd v Energy City Qatar Holding Company (“Hub Street Equipment”).⁽¹¹⁵⁾ It is not completely clear why the Australian courts have not gone further to apply the doctrine of transnational issue estoppel to decisions of the seat court on the validity of an arbitral award. But in two extra-judicial lectures, the then-Chief Justice James Allsop (who was part of the coram in Gujarat (Full Court) and Hub Street Equipment) suggested that if this approach were adopted, the doctrine of transnational issue estoppel would also have to apply to similar decisions of an enforcement court as well. His Honour thought this might be undesirable because it might incentivise a race to certain enforcement courts on tactical grounds, and “undermine the autonomy of the parties in choosing the seat”.⁽¹¹⁶⁾ On this, it may be open to the courts to apply the doctrine to decisions of a seat court, but not to decisions of an enforcement court, if there is a sound normative basis for drawing such a distinction – a point made by the majority of the Singapore Court of Appeal in Deutsche Telekom.⁽¹¹⁷⁾ It remains to be seen if such a basis can be found.⁽¹¹⁸⁾

60.               In the US, similar to Australia, the courts invoke something akin to the Primacy Principle and generally do not apply the doctrine of transnational issue estoppel. In TermoRio SA ESP and LeaseCo Group LLC v Electranta SP,⁽¹¹⁹⁾ the appellant applied for an arbitral award made in an arbitration seated in Columbia to be enforced in the US District of Columbia (“DC”). The award had been set aside in Columbia on the “rather parochial”⁽¹²⁰⁾ ground that Columbian law did not expressly permit the arbitration rules of the International Chamber of Commerce, which the tribunal had adopted, to be used in arbitration in Columbia. The Court of Appeals for the DC Circuit upheld the lower court’s dismissal of the application to enforce the award. The court held that the seat court had “primary jurisdiction” over the award, and the enforcement court, which had “secondary jurisdiction”, would generally not enforce an award that had been set aside by the seat court unless this was contrary to public policy. The court also observed that the public policy exception is very narrow: it would only apply where following the seat court’s decision would be “repugnant to fundamental notions of what is decent and just in the US”.⁽¹²¹⁾

61.               These principles have been followed by other US courts.⁽¹²²⁾ And, in the recent decision of Esso Exploration & Production Nigeria Ltd. v. Nigerian National Petroleum Corp 40 F 4th 56 (“Esso Exploration”), the Court of Appeal for the 2^nd Circuit set out the following principles regarding the public policy exception:

1. In applying that exception, the enforcement court could consider the merits of the seat court’s decision, but would only engage in a “light touch” review. It would not disturb the seat court’s decision if this “appear[ed] susceptible to reasonable disagreement”, but would only intervene if the seat court’s decision was “so facially deficient in [its] substantive analysis that [it] merit[s] no respect”.⁽¹²³⁾


2. The public policy exception will not be established based on “broad, non-specific evidence”. For example, allegations that a judicial system is unfair must be supported by specific evidence of this.⁽¹²⁴⁾

62.               For completeness, I note that some US courts have applied concepts of issue estoppel in determining the effect of decisions of other courts on the validity of an arbitral award.⁽¹²⁵⁾ However, these appear to be outlier cases and the predominant approach appears to involve the application of something akin to the Primacy Principle, as outlined above.

B. The Deutsche Telekom decision

63.               I turn to the current position in Singapore, which was set out in Deutsche Telekom. There, the parties were the Republic of India (“India”) and a German company (“DT”). In 2011, India terminated an agreement between an Indian state-owned entity (“Antrix”) and a company (“Devas”) in which DT had invested (the “Devas-Antrix Agreement”). DT then began arbitral proceedings against India, alleging that the termination of the Devas-Antrix Agreement was in breach of a bilateral investment treaty between India and Germany (the “India-Germany BIT”). The seat of the arbitration was Geneva, Switzerland.⁽¹²⁶⁾

64.               In 2017, the arbitral tribunal issued an interim award (“Interim Award”) that dismissed India’s jurisdictional objections and found India liable for breaching its obligations under the India-Germany BIT to accord fair and equitable treatment to DT’s investment in Devas. India then applied to the Federal Supreme Court of Switzerland (the “Swiss Federal Supreme Court”) for the Interim Award to be set aside. However, in 2018, the Swiss Federal Supreme Court dismissed India’s application for the setting-aside of the Interim Award (the “Swiss Setting-Aside Decision”).⁽¹²⁷⁾ This decision proved to be critical in the proceedings in Singapore.

65.               In 2020, the arbitral tribunal issued a final award on quantum (the “Final Award”). Subsequently, DT applied for and obtained an ex parte order for leave to enforce the Final Award in Singapore (the “Leave Order”). India applied for the Leave Order to be set aside, but this application was dismissed by the Singapore International Commercial Court. India appealed against this decision.⁽¹²⁸⁾

66.               A central issue on appeal was whether India was precluded from raising its objections to the Leave Order, on the basis that those objections had already been determined in the Swiss Setting-Aside Decision. This raised the question of how a Singapore enforcement court should treat an earlier decision of the seat court regarding the validity of an arbitral award. A five-judge bench of the Singapore Court of Appeal held that a Singapore enforcement court should apply the doctrine of transnational issue estoppel in determining the effect of a seat court’s decision on the validity of an arbitral award.⁽¹²⁹⁾ Further, the majority of the court indicated its support for the Primacy Principle albeit without conclusively deciding this point because it was unnecessary to do so.⁽¹³⁰⁾

67.               The court held that the doctrine of transnational issue estoppel applies to a seat court’s decision on the validity of an arbitral award, for the following reasons:

1. In the recognition and enforcement of an arbitral award, the regime established by the New York Convention does not operate in isolation. The domestic law of the enforcement court, including its conflict of laws rules, applies as well. In Singapore, as established in Merck Sharp,those rules include the principles of transnational issue estoppel. Hence, as a starting point, the doctrine of transnational issue estoppel would apply in the context of the recognition and enforcement of an arbitral award.⁽¹³¹⁾


2. Applying the doctrine of transnational issue estoppel would respect the parties’ choice of the arbitral seat. In making that choice, the parties choose the jurisdiction and the system of law that will regulate many matters concerning the arbitration. In line with this, an enforcement court should ordinarily give effect to a prior decision of the seat court on the validity of the arbitral award.⁽¹³²⁾


3. This approach would cohere with the notion that courts are part of an international legal order, within which they should “respect each other’s decisions in the fullest sense, and so far as possible avoid duplication, repetition and inconsistency in decision-making”.⁽¹³³⁾ In other words, applying transnational issue estoppel would advance international comity and the systemic need to limit relitigation.⁽¹³⁴⁾


4. Moreover, the doctrine of transnational issue estoppel could also be accommodated readily under the law of most common law jurisdictions. Hence, the Singapore enforcement court’s approach would align with prevailing common law approaches to the recognition and enforcement of foreign judgments.⁽¹³⁵⁾


5. Further, the doctrine was a flexible tool which could be applied in a way that safeguards the domestic concerns of the enforcement court, while advancing international comity as far as possible.⁽¹³⁶⁾

68.               Next, the Court of Appeal made the following remarks on the application of the doctrine of transnational issue estoppel in this context:

1. First, the court reiterated the holding in Merck Sharp that issue estoppel would only arise in respect of a foreign judgment if that decision had conclusively determined the issue in question in the foreign forum.⁽¹³⁷⁾ In this regard, it was noted that there may often be a need for expert evidence on whether a foreign court’s findings, including findings of fact, would give rise to issue estoppel under its own law. This was important because in civil law jurisdictions, the law of res judicata may only bar a party from relitigating the conclusions in the operative part of the judgment (meaning the actual decision on the validity of the award), but not the reasons or findings underlying that decision.⁽¹³⁸⁾


2. Second, the majority of the Court of Appeal left open the issue of whether a Singapore enforcement court should apply the doctrine of transnational issue estoppel to decisions of an enforcement court on the validity of an arbitral award.⁽¹³⁹⁾ The majority noted two concerns in this regard. First, it had been argued that this approach would elevate the first enforcement court to a status akin to that of the seat court – yet this seemed to be inconsistent with the New York Convention. Second, it had been suggested that this approach would incentivise forum shopping in that the parties might seek to bring enforcement proceedings in a favourable jurisdiction to obtain a finding that would then bind subsequent courts. The majority did not express a concluded view on this point. Nevertheless, the majority observed that any departure from the applicability of transnational issue estoppel in international arbitration had to be grounded in principle, and this might or might not lie in the policy of the scheme of international arbitration which places an emphasis on the seat court.⁽¹⁴⁰⁾


3. Third, the court noted that under English law, transnational issue estoppel will not arise from a “perverse” foreign judgment.⁽¹⁴¹⁾ A foreign judgment might be perverse where the foreign court applied its own law in determining the dispute in a way that was contrary to generally accepted doctrines of private international law (such as where the foreign court had disregarded the relevant applicable law or refused to apply other laws which ought to have regulated the rights of the parties, resulting in a judgment that was wrong).⁽¹⁴²⁾ The court did not expressly adopt this as a limit to the doctrine in Singapore. The court also did not resolve the issues left open in Merck Sharp as to the limits of the Arnold exception.

69.               The court then addressed the Primacy Principle. The majority of the court noted that it was not necessary to resort to the Primacy Principle in Deutsche Telekom, but indicated their receptiveness to that principle. The majority explained that the Primacy Principle accords with the widely accepted view in international arbitration that the seat court enjoys a position of primacy. Further, the Primacy Principle would advance the interests of comity, minimise or even avoid inconsistencies in judicial decisions and, most importantly, would ensure the finality and effectiveness of international commercial arbitration. And it would advance these ends in a way that accords due weight to party autonomy because the seat court will typically have been chosen by the parties.⁽¹⁴³⁾

70.               The majority held that the doctrinal basis of the Primacy Principle could be found in the principle that the Singapore courts are bound to interpret domestic legislation and hence, to develop the common law of Singapore, as far as permissible, in a way that advances Singapore’s international law obligations, and in the norm that the common law should, as far as possible, be developed in a way which coheres with relevant legislation. In this context, the New York Convention read with the Model Law⁽¹⁴⁴⁾ and the International Arbitration Act 1994 (2020 Rev Ed) (the “IAA”) provided the basis for the Primacy Principle.⁽¹⁴⁵⁾

71.               Further, the majority suggested that the Primacy Principle might operate as follows (subject to further elaboration as the law develops):

1. Where transnational issue estoppel does not arise from a prior decision of a seat court on the validity of an arbitral award, such as where the strict technical requirements of the doctrine were not met, or where the parties chose not to incur the cost or trouble of proving the elements of the doctrine, a Singapore enforcement court should nonetheless accord primacy to that decision, by treating it as presumptively determinative of the matters dealt with in the decision pertaining to the validity of the award. The onus would then shift to the party seeking to persuade the enforcement court to come to a different view to establish a sufficient basis for doing so.⁽¹⁴⁶⁾


2. The Primacy Principle would not be an absolute or preclusive principle. This was because the instruments that underpin that principle – the New York Convention, the Model Law, and the IAA – provide for most arbitral awards to be subject to the rules for enforcement and challenge in the courts of the country in which enforcement is sought. Further, in line with the principle of “double-control”, it was for each enforcement court to determine the weight to be accorded to decisions of the seat court.⁽¹⁴⁷⁾


3. The Primacy Principle would only apply where there had been a prior decision of a seat court (as opposed to a prior decision of an enforcement court). This was because the scheme for the recognition and enforcement of international arbitral awards accords a special status to the seat court which supervises the conduct of proceedings, and which alone has the power to grant the active remedies. Enforcement courts do not enjoy any such status, though the decision of a prior enforcement court on an arbitral award would be carefully considered because it was the pronouncement of another court dealing with the same issues.⁽¹⁴⁸⁾


4. There were at least three possible limits to the Primacy Principle. First, that principle might not apply if this would conflict with the public policy of the enforcement jurisdiction. Second, the principle might not apply if “there were serious procedural flaws in the decision-making process of the seat court”, akin to failures of natural justice, or where to uphold the seat court’s decision would be repugnant to fundamental notions of what was just. Third, the principle might not apply where the seat court’s decision was “so wrong as to be perverse”. As to whether an error falling short of rendering the seat court’s decision perverse would suffice, the majority observed that it would not suffice for the enforcement court to merely disagree with a decision on which reasonable minds might differ. The majority left open the question of where between perverseness and a mere disagreement with the decision of the prior court, the threshold lay.⁽¹⁴⁹⁾

72.               Having set out the aforementioned principles and observations, the Court of Appeal considered whether India was estopped from raising its objections to the Leave Order. The court found that all the requirements of transnational issue estoppel were satisfied, for the following reasons.

1. First, the requirement of a final and conclusive decision on the issues in respect of which issue estoppel was said to arise was met. In this regard, the court considered the parties’ expert evidence on the effect of the Swiss Setting-Aside Decision under Swiss law. Based on that evidence, the court found that a Swiss court would not reconsider India’s objections to the arbitral award that had been considered and rejected by the Swiss Federal Supreme Court in the Swiss Setting-Aside Decision.⁽¹⁵⁰⁾


2. Second, the requirement of identity of parties was satisfied.⁽¹⁵¹⁾


3. Third, the requirement of identity of issues was met because India’s objections to the Leave Order were the same as those that had been considered and dismissed by the Swiss Federal Supreme Court in the Swiss Setting-Aside Decision.⁽¹⁵²⁾

73.               Further, the Court of Appeal found that none of the limits to transnational issue estoppel suggested in Merck Sharp⁽¹⁵³⁾ applied. In the circumstances, the court found that India was estopped from raising its objections to the Leave Order and therefore dismissed India’s appeal.⁽¹⁵⁴⁾

C. The applicability of the doctrine to decisions of enforcement courts on the validity of arbitral awards

74.               As I have noted, the majority of the Court of Appeal in Deutsche Telekom did not express a view on whether transnational issue estoppel might arise from an enforcement court’s decision on the validity of an arbitral award.⁽¹⁵⁵⁾ This is a complex and vexed issue, and this paper will not express a view on this point. That said, the following considerations may be notable.

75.               To begin with, as the majority in Deutsche Telekom observed, given that transnational issue estoppel may arise from a decision of a seat court, any ruling that such an estoppel would not arise from a decision of an enforcement court would have to be grounded in principle. This justification might perhaps be found in the policy reflected in the legal framework governing international arbitration, which places an emphasis on the seat court.⁽¹⁵⁶⁾

76.               Yet the courts might be able to secure the primacy of the seat court, while extending the doctrine of transnational issue estoppel to decisions of enforcement courts, by developing principles that promote this objective.

1. One possible principle could be that where there are conflicting decisions by the seat court and a foreign enforcement court, only the former would give rise to transnational issue estoppel (even if it were issued after the decision of the foreign enforcement court). Notably, the decision in Carpatsky provides some support for such a principle. There, the Ukrainian courts had found that there was no valid arbitration agreement and refused to recognise the arbitral award on this basis.⁽¹⁵⁷⁾ However, the Swedish courts (the courts of the seat jurisdiction) found that a valid arbitration agreement had been formed. Butcher J held that, in these circumstances, it would “not be in accordance with justice” for the English court to apply issue estoppel based on the rulings of the Ukrainian court.⁽¹⁵⁸⁾


2. Further, where an enforcement court has issued a decision on the validity of the arbitral award, but a decision of the seat court is pending, the court in which transnational issue estoppel is invoked could impose a case management stay on the local proceedings, pending the decision of the seat court. This approach would be consistent with Art VI of the New York Convention, which empowers a court to adjourn proceedings to enforce an arbitral award if a setting-aside application has been made in the seat jurisdiction.⁽¹⁵⁹⁾ Once the seat court renders its decision, the enforcement court could then decide whether to recognise an issue estoppel based on that decision or to apply the Primacy Principle to that decision.

77.               These are but a few preliminary and tentative observations, and it remains to be seen what position will ultimately be adopted in Singapore.

          VI. Conclusion

78.               This paper has suggested that the doctrine of transnational issue estoppel is a valuable tool that common law courts can deploy to address transnational relitigation. It can accommodate the policy factors that count against relitigation, and yet preserve space for the court to fulfil its constitutional role of advancing the rule of law, delivering justice, and securing the public policy of its jurisdiction. Common law courts should therefore adopt this doctrine, and take up the task of developing it in the light of the applicable policy factors. Several questions remain unresolved and will in time require careful consideration, including the following:

1. whether a foreign judgment should be recognised (as capable of giving rise to transnational issue estoppel) where:
   1. local proceedings on the same or substantially the same subject matter have been commenced and are pending;⁽¹⁶⁰⁾
   2. the foreign jurisdiction that issued the foreign judgment does not recognise the judgments of the forum court;⁽¹⁶¹⁾ or
   3. in the context of international arbitration, the foreign judgment is one of an enforcement court;⁽¹⁶²⁾
2. the nature and scope of the outer limits of the doctrine of transnational issue estoppel;⁽¹⁶³⁾


3. whether and, if so, how the Arnold exception applies in the context of transnational issue estoppel;⁽¹⁶⁴⁾ and


4. whether the Primacy Principle should be recognised as a supplementary doctrine in the context of international arbitration.⁽¹⁶⁵⁾

_{* I am deeply grateful to my former law clerks, Darien The and Lu Huiyi, and my colleagues, Assistant Registrars Tan Ee Kuan and Wee Yen Jean, for all their assistance in the research for and preparation of this paper.}

_{(1)       PT Putrabali Adyamulia v Société Est Epices [2003] 2 Lloyd’s Rep 700 (England) (“Putrabali (English Commercial Court)”); PT Putrabali Adyamulia (Indonesia) v. Rena Holding, et al (2007) 32 ICCA Yearbook Commercial Arbitration 299 (Court of Cassation of France, 29 June 2007) (“Putrabali (Court of Cassation of France)”).
(2)       Kabab-Ji SAL v Kout Food Group [2022] 2 All ER 911 (England) (“Kabab-Ji (UK Supreme Court)”); Kout Food Group v Kabab-Ji SAL (2022) 47 ICCA Yearbook Commercial Arbitration 603 (Court of Cassation of France, 28 September 2022) (France) (“Kabab-Ji (Court of Cassation of France)”).
(3)       The Republic of India v Deutsche Telekom AG [2023] SGCA(I) 10 (“Deutsche Telekom”).
(4)       Putrabali (English Commercial Court) at [33].
(5)       Michael Polkinghorne and Marie-Aude Ziadé, “French Courts Confirm That An Award Set Aside at the Seat of the Arbitration Can Be Enforced” (2008) 13(1) IBA Arbitration Newsletter 23 (“Polkinghorne and Ziadé”).
(6)       Polkinghorne and Ziadé.
(7)       Putrabali (Court of Cassation of France) at 302.
(8)       Jonathan Mance, “Arbitration: a Law unto itself?” (2016) 32(2) Arbitration International 223 (“Mance”) at 232.
(9)       Mance at 232 (noting the “idiosyncratic and fragmenting” effects of the Court of Cassation’s decision).
(10)       Mance at 231.
(11)       See paragraph 27 below.
(12)       See paragraphs 22–27 below.
(13)       Kabab-Ji (UK Supreme Court) at [5]–[6].
(14)       Kabab-Ji (Court of Cassation of France) at 605–607.
(15)       Kabab-Ji (UK Supreme Court) at [18].
(16)       Kabab-Ji (Court of Cassation of France) at 605 (referring to the decision of the Paris Court of Appeal); Kabab-Ji (UK Supreme Court) at [87], [89] and [90].
(17)       Kabab-Ji (UK Supreme Court) at [87].
(18)       Kabab-Ji (UK Supreme Court) at [87].
(19)       Kabab-Ji (UK Supreme Court) at [89] and [90].
(20)       Kabab-Ji (UK Supreme Court) at [90].
(21)       See paragraphs 28–29 below.
(22)       KR Handley, Spencer Bower and Handley: Res Judicata (LexisNexis, 5th Ed, 2019), para 27.02.
(23)       Sheoparasan Singh v Ramnandan Singh (1916) 43 Indian Law Reports 694 at 706.
(24)       South West Africa Case (Liberia v South Africa) [1966] ICJ Rep 4 at 332–333 (per Jessup J).
(25)       Lockyer v Ferryman (1877) 2 App Cas 519 at 530; The Royal Bank of Scotland NV (formerly known as ABN Amro Bank NV) and others v TT International Ltd (nTan Corporate Advisory Pte Ltd and others, other parties) and another appeal [2015] 5 SLR 1104 (“TT International”) at [98].
(26)       Ferrer v Arden (1599) 6 Co Rep 7a (“Ferrer”).
(27)       Ferrer at 9a.
(28)       Sirko Harder, “The Effects of Recognized Foreign Judgments in Civil and Commercial Matters” (2013) 62 ICLQ 441 (“Harder”) at 450.
(29)       John Sorabji, English Civil Justice after the Woolf and Jackson Reforms: A Critical Analysis (Cambridge University Press, 2014) at pp 167–168.
(30)       Kho Jabing v AG [2016] 3 SLR 1273 at [2].
(31)       Henderson v Henderson (1843) 3 Hare 100; 67 ER 313.
(32)       TT International at [98].
(33)       Thoday v Thoday [1964] P 181 at 197, cited in TT International at [99].
(34)       Arnold v National Westminster Bank plc [1991] 2 AC 93 (“Arnold”) at 104D–104E, affirmed in TT International at [103] and [139].
(35)       Watt (formerly Carter) v Ahsan [2008] 1 AC 696 at [31], cited in TT International at [100].
(36)       Arnold at 109A–109B, cited in TT International at [103].
(37)       See paragraph 18(c) below.
(38)       TT International at [101].
(39)       TT International at [104].
(40)       Lee Tat Development Pte Ltd v Management Corporation of Strata Title Plan No 301 [2005] 3 SLR(R) 157 at [14]–[15].
(41)       TT International at [71], citing Mulkerrins v PricewaterhouseCoopers (a firm) [2003] 1 WLR 1937 at [10].
(42)       TT International at [190].
(43)       See paragraphs 15–16 above.
(44)       Putrabali (English Commercial Court) at [2].
(45)       Putrabali (Court of Cassation of France).
(46)       Sundaresh Menon CJ, “The Transnational System of Commercial Justice and the Place of International Commercial Courts”, Lecture in Bahrain (9 May 2023) (“The TSCJ and the Place of ICCs”) at paras 17–28.
(47)       The TSCJ and the Place of ICCs at paras 20 and 22.
(48)       The TSCJ and the Place of ICCs at paras 24 and 26.
(49)       The TSCJ and the Place of ICCs at para 28.
(50)       Sundaresh Menon CJ, “The Law of Commerce in the 21st Century: Transnational commercial justice amidst the wax and wane of globalisation”, Lecture hosted by the University of Western Australia Law School and the Supreme Court of Western Australia (27 July 2022) (“The Law of Commerce in the 21st Century”) at para 17.
(51)       The TSCJ and the Place of ICCs at para 16.
(52)       See paragraph 16(a) above.
(53)       Yeo Tiong Min, “The Changing Global Landscape for Foreign Judgments”, Yong Pung How Professorship of Law Lecture 2021 (6 May 2021) (“Yeo”) at para 8.
(54)       Adrian Briggs, Private International Law in English Courts (Oxford University Press, 2014) at para 3.139; Harder at 449–450; Yeo at para 6.
(55)       The TSCJ and the Place of ICCs at paras 29–47.
(56)       Merck Sharp & Dohme Corp (formerly known as Merck & Co, Inc) v Merck KGaA (formerly known as E Merck) [2021] 1 SLR 1102 (“Merck Sharp”) at [33].
(57)       Merck Sharp.
(58)       Merck Sharp at [3], [4] and [7].
(59)       Merck Sharp at [8].
(60)       Merck Sharp at [10].
(61)       Merck Sharp at [14].
(62)       See Section IV.A above.
(63)       Merck Sharp at [25] and [34].
(64)       There are slightly different conditions for the recognition of a foreign judgment under other statutory regimes in Singapore: see, eg, Part 3 of the Choice of Court Agreements Act 2016, which gives effect to the Hague Convention of 30 June 2005 on Choice of Courts Agreements, and ss 4 and 5 of the Reciprocal Enforcement of Foreign Judgments Act 1959.
(65)       Merck Sharp at [36(a)].
(66)       Merck Sharp at [36(b)] and [38].
(67)       Merck Sharp at [39].
(68)       Merck Sharp at [40].
(69)       Merck Sharp at [40(a)], citing Carl Zeiss Stiftung v Rayner & Keeler Ltd [1967] 1 AC 853 (“Carl Zeiss Stiftung”)at 918.
(70)       Merck Sharp at [40(b)] and [42]–[43].
(71)       Harder at 451–452.
(72)       See Carl Zeiss Stiftung at 919 (per Lord Reid), and the cases cited in Merck Sharp at [42].
(73)       See paragraphs 22–27 above.
(74)       Merck Sharp at [40(c)].
(75)       Merck Sharp at [52].
(76)       Merck Sharp at [55].
(77)       Merck Sharp at [56].
(78)       Merck Sharp at [57].
(79)       Merck Sharp at [58] and [61].
(80)       See paragraph 18(c) above.
(81)       Merck Sharp at [63], citing Godard v Gray (1870) LR 6 QB 139.
(82)       Merck Sharp at [63(a)]–[63(d)].
(83)       Merck Sharp at [65].
(84)       Merck Sharp at [47].
(85)       Merck Sharp at [48].
(86)       Carl Zeiss Stiftung; Sundaresh Menon CJ, “The Role of the National Courts of the Seat in International Arbitration”, Keynote Address at the 10th Annual International Conference of the Nani Palkhivala Arbitration Centre (17 February 2018) (“The Role of the National Courts of the Seat”) at para 24.
(87)       Apart from Merck Sharp, relevant cases include The Sennar (No 2) [1985] 1 WLR 490 (which held that issue estoppel may arise from a foreign court’s decision on a jurisdictional issue) and Desert Sun Loan Corp v Hill [1996] 2 All ER 847 (which held that issue estoppel may arise from a foreign court’s interlocutory decision on a procedural issue, eg, whether a party had submitted to proceedings in the foreign court): see Peter Barnett, Res Judicata, Estoppel, and Foreign Judgments (Oxford University Press, 2001) at paras 5.35–5.71.
(88)       See paragraph 11(b) above and Kabab-Ji (UK Supreme Court) at [87].
(89)       See paragraph 38(a) above.
(90)       See paragraphs 38(d) and 39 above.
(91)       See paragraph 35 above.
(92)       See paragraph 38(d) above.
(93)       Convention on the Recognition and Enforcement of Foreign Arbitral Awards (10 June 1958) 330 UNTS 3 (entered into force 7 June 1959).
(94)       BAZ v BBA and others and other matters [2020] 5 SLR 266 at [36]; Gary B Born, International Commercial Arbitration (Kluwer Law International, 3rd Ed, 2020) at §26.05[C][8][a][i]).
(95)       See, eg, Mance; Manu Thadikkaran, “Enforcement of Annulled Arbitral Awards: What Is and What Ought to Be?” (2014) 31(5) Journal of International Arbitration 575 at 577–590.
(96)       The Role of the National Courts of the Seat at paras 6–8.
(97)       See paragraphs 7, 10 and 11(b) above.
(98)       See the discussion of the problems with both theories in The Role of the National Courts of the Seat at paras 11–20. A similar view was expressed by Lord Mance: see Mance at 235.
(99)       Diag Human SE v The Czech Republic [2014] 2 Lloyd’s Rep 283 (“Diag Human”).
(100)       Diag Human at [59] and [63].
(101)       Carpatsky Petroleum Corp v PJSC Ukrnafta [2020] 1 Lloyd’s Rep 566 (“Carpatsky”).
(102)       Carpatsky at [130], [134] and [135].
(103)       Yukos Capital Sarl v OJSC Rosneft Oil Co (No 2) [2014] QB 458 (“Yukos Capital”).
(104)       Yukos Capital at [151] and [156].
(105)       See, eg, Stati and others v Republic of Kazakhstan [2017] 2 Lloyd’s Rep 201 at [84], [87] and [89].
(106)       Tomolugen Holdings Ltd v Silica Investors Ltd [2016] 1 SLR 373 at [75]; Anupam Mittal v Westbridge Ventures II Investment Holdings [2023] 1 SLR 349 (“Anupam Mittal”) at [47].
(107)       In Anupam Mittal, the Singapore Court of Appeal held that in determining whether a dispute was arbitrable in Singapore, a Singapore court would consider both domestic public policy (the public policy of Singapore) and the public policy of the law governing the arbitration agreement: see Anupam Mittal at [55]. Hence, a foreign judgment which held that a dispute was not arbitrable under the governing law of the arbitration agreement (being the law of that foreign forum) might be capable of giving rise to transnational issue estoppel in Singapore.
(108)       Diag Human at [58].
(109)       Air Foyle Ltd and another v Centre Capital Ltd [2003] 2 Lloyd’s Rep 753 at [36(5)]; Deutsche Telekom at [93]–[95].
(110)       Gujarat NRE Coke Ltd and another v Coeclerici Asia (Pte) Ltd (2013) 304 ALR 468 (“Gujarat (Full Court)”).
(111)       Coeclerici Asia (Pte) Ltd v Gujarat NRE Coke Ltd [2013] FCA 882 (“Gujarat (Single Judge)”) at [102].
(112)       Gujarat (Single Judge) at [103].
(113)       Gujarat (Full Court) at [64].
(114)       Minmetals Germany GmbH v Ferco Steel Ltd [1999] 1 All ER (Comm) 315.
(115)       Hub Street Equipment Pty Ltd v Energy City Qatar Holding Company (2021) 396 ALR 1 at [77].
(116)       James Allsop, “The Authority of the Arbitrator” (2014) 30(4) Arbitration International 639 at 651; James Allsop, “International Commercial Arbitration – the Courts and the Rule of Law in the Asia Pacific Region” (2015) 81(2) Arbitration 169 at 172. See also Matthew Barry, “The Role of the Seat in International Arbitration: Theory, Practice, and Implications for Australian Courts” (2015) 32(3) Journal of International Arbitration 289 (“Barry”) at 319.
(117)       See paragraph 68(b) below.
(118)       See paragraphs 74–77 below.
(119)       TermoRio SA ESP and LeaseCo Group LLC v Electranta SP 487 F 3d 928 (DC Cir, 2007) (“TermoRio”).
(120)       Barry at 308.
(121)       TermoRio at 937–939.
(122)       See, eg, Thai-Lao Lignite Co Ltd v Government of the Lao People’s Democratic Republic 864 F 3d 172 (2nd Cir, 2017); Esso Exploration & Production Nigeria Ltd. v. Nigerian National Petroleum Corp 40 F 4th 56 (2nd Cir, 2022) (“Esso Exploration”).
(123)       Esso Exploration at 75.
(124)       Esso Exploration at 77.
(125)       See, eg, Belmont Partners, LLC v. Mina Mar Group, Inc. 741 F Supp 2d 743 (WD Va, 2010) at 752–753.
(126)       Deutsche Telekom at [5], [15] and [16].
(127)       Deutsche Telekom at [17]–[18].
(128)       Deutsche Telekom at [19], [32] and [33].
(129)       Deutsche Telekom at [4], [193], [195] and [196].
(130)       Deutsche Telekom at [4] (judgment of the majority); cf. Deutsche Telekom at [221] (concurring opinion of Jonathan Hugh Mance IJ).
(131)       Deutsche Telekom at [97].
(132)       Deutsche Telekom at [98].
(133)       Mance at 239–240.
(134)       Deutsche Telekom at [99] and [100].
(135)       Deutsche Telekom at [100].
(136)       Deutsche Telekom at [101]; see further paragraph 44 above.
(137)       See paragraph 37(b) above.
(138)       Deutsche Telekom at [88]–[90].
(139)       Jonathan Hugh Mance IJ, who delivered a concurring opinion, considered that transnational issue estoppel could also arise from decisions of an enforcement court on the validity of an arbitral award: see Deutsche Telekom at [201], [214], [215] and [221].
(140)       Deutsche Telekom at [91]–[92].
(141)       See paragraph 55 above.
(142)       Deutsche Telekom at [93]–[95].
(143)       Deutsche Telekom at [121].
(144)       United Nations Commission on International Trade Law, Model Law on International Commercial Arbitration, 1985, UN Doc A/40/17.
(145)       Deutsche Telekom at [122].
(146)       Deutsche Telekom at [122], [125] and [130(a)].
(147)       Deutsche Telekom at [123], citing PT First Media TBK (formerly known as PT Broadband Multimedia TBK) v Astro Nusantara International BV and others and another appeal [2014] 1 SLR 372 (“Astro”) at [75].
(148)       Deutsche Telekom at [124].
(149)       Deutsche Telekom at [127]–[129] and [130(b)].
(150)       Deutsche Telekom at [172]–[173].
(151)       Deutsche Telekom at [175].
(152)       Deutsche Telekom at [154] and [175].
(153)       See paragraph 38 above; Deutsche Telekom at [177].
(154)       Deutsche Telekom at [176] and [188].
(155)       See paragraph 68(b) above; Deutsche Telekom at [92].
(156)       See paragraph 68(b) above; Deutsche Telekom at [92].
(157)       Carpatsky at [22] and [31]–[32].
(158)       Carpatsky at [110]; Richard Garnett, “Estoppel and Enforcement of International Arbitration Awards” (2021) 95 Australian Law Journal 337 at 347.
(159)       Deutsche Telekom at [76]; The Role of the National Courts of the Seat at paras 40–45.
(160)       See paragraph 36(b) above.
(161)       See paragraph 36(c) above.
(162)       See paragraphs 74–77 above.
(163)       See paragraphs 38(a)–38(d) above.
(164)       See paragraphs 39–40 above.
(165)       See paragraphs 69–71 above.}