Supplementary Written Evidence from Professor Sarah Green, Law Commissioner
Section 67 (challenging jurisdiction)
Questions
1.1 New section 67(3)(e), proposed in clause 10(3), would add an alternative disposal of a challenge to the substantive jurisdiction of the arbitral tribunal:
(e) declare the award to be of no effect, in whole or in part.
1.2 Could Professor Green explain the rationale for this additional power and what it adds to the power to set aside the award in whole or in part.
1.3 It is noted that a parallel power exists under existing section 68(3)(c) in respect of a challenge to an award for serious irregularity, but that no such power exists under section 69 (Appeal on point of law), nor in existing section 67. Could Professor Green explain the thinking behind these differences?
Response
1.4 We assume that there is a difference between declaring an award to be of no effect, and setting aside an award. After all, these two are listed as alternative remedies in the Act (for example, section 68(3)).
1.5 We think the difference is this: setting aside an award allows the tribunal to issue another award, whereas declaring an award to be of no effect means that the award nevertheless continues to exist so that the tribunal cannot revisit it. This is also the view of authors (see the citations in paragraph 8.59 of our first consultation paper).
1.6 Declaring an award to be of no effect is already available under section 67, but only in cases falling under section 67(1)(b). In contrast, it is available across the board in section 68. And case law has assumed that it was meant to be available across the board in section 67 too (see the citations in paragraph 9.141 of our final report).
1.7 We proposed regularising the position, and amending the Act to say explicitly that declaring an award to be of no effect was available across the board in section 67. There were 66 responses to this proposal: 58 agreed, 5 disagreed, and 3 gave other responses (see paragraph 9.130 of our final report).
1.8 The remedies available under section 69 were not part of our inquiry – no stakeholder suggested any need for that reform. Why the drafter chose not to include declaring an award to be of no effect in section 69 (or forgot to include it), we can only speculate. My best guess would be that it was thought that an award which contained an error of law should not be allowed to stand on the record: it should always be varied (corrected), remitted back to the tribunal, or set aside.
Questions
1.9 New section 67(3C)(c), proposed in Clause 11, would contain a provision that “evidence that was heard by the tribunal must not be re-heard by the court, unless the court considers it necessary in the interests of justice”.
1.10 However, proposed sub-sections (3C)(a) and (b), which would prohibit new grounds of objection to the jurisdiction and new evidence not heard by the tribunal, unless the applicant could not with reasonable diligence have discovered the ground or put the evidence before the tribunal respectively, have no such saving provision in similar terms to the proviso “unless the court considers it necessary in the interests of justice”.
1.11 Does Professor Green not consider that such a saving provision might be useful, for example to protect applicants whose legal team missed a possible ground of objection or failed to spot a lacuna in the evidence they adduced before the tribunal? Given that the ultimate arbiter of the merits of a section 67 challenge is the court, if, for instance a new legal team is engaged to present the challenge, is it right that the challenge should necessarily be defeated and a meritorious applicant left to their remedies against the original legal team rather than the court’s having the power to hear the new ground of objection / new evidence?
1.12 There may be other cases where the interests of justice might require the hearing of a new ground of objection or new evidence where the Ladd v Marshall style tests were not met.
Response
1.13 There are three parts to inserted section 67(3C): (a) no new objections; (b) no new evidence; (c) no re-hearing of evidence.
1.14 As for no new objections, section 73(1) already says that a party must make a prompt objection to the jurisdiction of the tribunal, and otherwise they may not raise that objection later unless they could not with reasonable diligence have discovered the grounds for the objection. There is no further saving “unless the court considers it necessary in the interests of justice”. So section 67(3C)(a) merely confirms the current law.
1.15 As for no new evidence, section 67(3C)(b) takes the same approach to new evidence as to new objections. In this way, it aligns with section 73(1). It also aligns with Ladd v Marshall, which said that leave to adduce further evidence on appeal will only be granted if: the evidence could not have been obtained with reasonable diligence for use at the trial; it would probably have an important influence on the result of the case; and the evidence is such as is presumably to be believed. There is no further saving “unless the court considers it necessary in the interests of justice”. Indeed, such a further saving was expressly rejected by Lord Justice Hodson in that case. So section 67(3C)(b) also aligns precisely with current law.
1.16 Also, the point made above “for example to protect applicants whose legal team missed a possible ground of objection or failed to spot a lacuna in the evidence they adduced before the tribunal” merely opens the door to unsuccessful parties attempting to have a second bite of the cherry, and is contrary to the broader principle of finality in litigation, or as here, finality in arbitration.
1.17 As for no re-hearing of evidence, the only guide here is rule 52.21 of the Civil Procedure Rules. Rule 52.21(2) says that, when a court hears an appeal, it will not receive oral evidence, and it will not receive evidence which was not before the lower court, unless the appeal court orders otherwise. When would it order otherwise? Rule 52.21(1) says that an appeal is limited to a review of the decision of the lower court unless “it would be in the interests of justice to hold a re-hearing”. This is why that saving is mirrored in section 67(3C)(c).
1.18 The wording of section 67(3C) was arrived at through an iterative process informed by thorough consultation. In our first consultation paper, we proposed procedural restrictions for a challenge under section 67. 81 consultees responded: 55 agreed, 24 disagreed, and 2 expressed other views. We took all comments into account, and made an amended proposal in our second consultation paper. There were 44 further responses: 31 agreed, 9 disagreed, and 4 gave other responses. We took all those further comments into account, and adjusted accordingly. What appears in section 67(3C) is our third and final iteration.
Inserted section 23A (disclosure)
Questions
1.19 It has been suggested that the words “after reasonable inquiry” be added at the end of sub clause 1 and sub clause 2. Professor Green’s comments would be greatly appreciated.
1.20 Should the clause be amended, as has been suggested, so that the duty to disclose only apples when the person intends to accept or accepts the appointment?
1.21 Is section 23A(1) necessary, as the duty to the parties is contained with section 23A(2)? It has been suggested that the duty is owed only to the parties and not to a person approaching a prospective appointee? It is suggested that if there is a failure to comply with the duty under section 23A(1), but there was subsequent disclosure under section 23A(2) this might cause complications.
1.22 To what extent is it relevant that the provision in section 23A(1) is in line with the Model Law?
Response
1.23 Under inserted section 23A, an arbitrator must disclose “relevant circumstances”. These are defined in section 23A(3) as those which might reasonably give rise to justifiable doubt as to their impartiality; and include circumstances of which the arbitrator ought reasonably to be aware. Adding “after reasonable inquiry” to sections 23A(1) and (2) would risk conflicting with section 23A(3), by having different requirements of reasonableness in different places.
1.24 In our first consultation paper, we suggested that an arbitrator might be required to make reasonable inquiries as to any conflict of interest. We received responses from consultees which questioned that approach (see paragraphs 3.83 to 3.93 in our final report). In light of those responses, we concluded that a positive duty to make inquiries was not warranted. Put simply, sometimes what an arbitrator actually knows is also all that they ought reasonably to know. Not every case will call for further active inquiries. For example, what is required of a sole practitioner on their first appointment might be different from what is required of a partner in a firm of solicitors handling multiple clients and appointments.
1.25 We think that a would-be arbitrator should make disclosure pre-appointment so that the parties can decide whether the arbitrator is suitable for appointment. If the arbitrator only makes disclosure after appointment, the parties might find themselves with an unsuitable arbitrator. Further, removing an arbitrator is more complicated (it can require an application to court under section 24) than appointing properly in the first place.
1.26 An arbitrator who is approached, but who has no intention of accepting appointment, can simply turn down the approach. There is then no need for them to make any disclosure, and no consequences of not making disclosure. For example, clause 2 does not make it a crime not to make disclosure. Yes, it imposes a duty of disclosure, but it does not prescribe any consequences. What happens next is decided by other sections of the Arbitration Act 1996, in particular section 23 (parties remove an arbitrator) or section 24 (court removes an arbitrator). These consequences are irrelevant to a person who turned down an arbitral appointment.
1.27 Note that UNCITRAL Model Law, article 12, similarly says that a person approached in connection with their possible appointment as arbitrator “shall disclose” any conflicts of interest. The UNCITRAL Model Law has been adopted in 118 jurisdictions worldwide. The same rule applies under the Arbitration (Scotland) Act 2010, schedule 1 rule 8. There does not appear to be any practical problem here.
1.28 When a would-be arbitrator is approached, there may be no arbitral proceedings yet on foot, and so no “parties” to arbitral proceedings. And anyway, the person approaching the arbitrator for appointment might not be one of the parties – for example, it might be an arbitral institution tasked with making the appointment, or the court (see section 18), or another arbitrator seeking to appoint a chairperson. An arbitrator should make disclosure to an appointing authority. Again, this is the position under the UNCITRAL Model Law, and in Scotland.
1.29 More generally, section 23A(1) is the duty of disclosure for appointment, and section 23A(2) is the continuing duty of disclosure post-appointment. It is possible to comply with one but breach the other. In particular, complying with (2) does not necessarily cure a breach of (1). It will be up to the parties to decide whether, in light of any breach, they still have confidence in the arbitrator. For example, they can remove an arbitrator (section 23) or apply to court for the court to remove the arbitrator (section 24).
1.30 The dual approach in section 23A(1) and (2) aligns with the UNCITRAL Model Law, article 12(1), and the Arbitration (Scotland) Act 2010, schedule 1 rule 8. It is the same in Sweden, and in Switzerland, and under the ICC Arbitration Rules, and under the LCIA Arbitration Rules (see paragraph 3.23 of our final report). In short, section 23A aligns with international best practice.
Clauses 3 and 4 (arbitrator immunity)
Question
1.31 It has been suggested that the provisions in relation to immunity are unclear. In particular it is said that the term “unreasonable” to be inserted in a new provision in section 29 lacks precision. Professor Green’s comments would be appreciated.
Response
1.32 We think that clauses 3 and 4 are clear. They are further explained in the Explanatory Notes, and in chapter 5 of our final report.
1.33 Inserted section 29(4) says that an arbitrator incurs no liability for resignation unless the resignation is shown to be unreasonable. We do not think that this lacks precision. The law of England and Wales has long applied a standard of reasonableness to professional conduct.
1.34 Some consultees suggested that the Act should provide a list of when resignation might be reasonable (see paragraph 5.29 of our final report). Suggestions included bereavement, illness, or a need to avoid subsequent international sanctions (for example, following war in Ukraine). We concluded that it would not be possible to come up with a list which would be exhaustive now and for all time. “Reasonable” provides flexibility, and is a very familiar test in the law of England and Wales.
Inserted section 41A (emergency arbitrators)
Questions
1.35 It has been suggested that clause 41A(2) be amended so that it read “order or award” rather than “order or directions”. If directions are made they would ordinarily be contained in an order; the omission of “award” would suggest no award can be made by an emergency arbitrator. Could Professor Green please comment.
1.36 A similar point has been made about section 39 in the light of the decision in YDU v SAB and BYH [2022] EWHC 3304. We would appreciate Professor Green’s comments.
Response
1.37 Inserted section 41A(2) refers to “order or directions” of an emergency arbitrator because section 41(5) refers to “order or directions” of a normal arbitrator. The purpose of section 41A is to extend to emergency arbitrators the scheme already available for normal arbitrators under section 41 (see chapter 8 of our final report). The language of the two sections matches.
1.38 Section 41 does not deal with awards by normal arbitrators. The scheme for enforcing awards (as opposed to orders or directions) is separate. Awards are enforced under section 66 (or challenged under sections 67 to 69). Since section 41A replicates for emergency arbitrators what section 41 does for normal arbitrators, section 41A does not refer to awards either.
1.39 We question whether an emergency arbitrator can issue an award properly so called. Awards are final and binding. Emergency arbitrator decisions are provisional – they can be revisited by the normal tribunal once the normal tribunal is operational. However, if an emergency arbitrator is somehow empowered to issue an award, that would be enforceable like any award under section 66. No provision is needed in section 41A (just like there is no matching provision in section 41).
1.40 The Arbitration Bill does not reform section 39. We considered “minor” reform of section 39, but concluded against it. The full reasons are set out in paragraphs 11.62 to 11.79 of our final report. In short, we said that, since the publication of our first consultation paper, new case law showed that the interpretation of section 39 was a live issue of expanding scope, and that it would be unwise to tinker with section 39 in those circumstances. The case of YDU v SAB reinforces the point that section 39 is a matter of widening live debate.
Section 44 (court orders)
Question
1.41 It has been suggested that section 44 be amended to make it clear that the court can enforce the order of the Emergency Arbitrator. Professor Green’s comments would be greatly appreciated.
Response
1.42 In this context, the Arbitration Act 1996 has two pathways, as follows.
(1) Either an arbitrator issues an order, which, if ignored, is followed by a peremptory order, which can then be enforced by the court. This is section 41. This scheme has now been extended to emergency arbitrators by section 41A.
(2) Or the parties apply to court for the court to issue its own order. This is section 44. For non-urgent cases, an application under section 44 can require the permission of the tribunal. Clause 8 of the Bill allows instead the permission of an emergency arbitrator (when it is the emergency arbitrator, not yet the normal tribunal, which has the reins of the arbitral proceedings).
1.43 It would disrupt this scheme if section 44 provided for the court to enforce the order of an emergency arbitrator. It would render section 41A redundant. And why would section 44 enforce the order only of an emergency arbitrator, and not a normal arbitrator? Yet that change would render section 41 redundant. But there is no need for any of this. Section 44 is about the court making its own orders. Meanwhile, the court can enforce the orders of normal arbitrators and emergency arbitrators under the other pathway provided by section 41 and section 41A.
Section 103 (New York Convention)
Questions
1.44 It has been suggested that there is a lack of clarity on the effect of the provisions of new clause 6A in the context of section 103 (and the effect of the New York Convention) in the light of the decision of the Supreme Court in Katb; Professor Green’s comments would be greatly appreciated
1.45 Does the Bill in its present form align with prevailing international jurisprudence on these provisions of the New York Convention?
Response
Compatibility of section 6A with the New York Convention
1.46 We discuss the compatibility of section 6A with the New York Convention in paragraphs 12.46 to 12.52 of our final report. In short, we say as follows.
1.47 The New York Convention (1958) provides for signatory states to uphold arbitration agreements and enforce in-coming foreign-seated arbitral awards. There are 168 state signatories to the New York Convention.
1.48 Under article V.1(a) of the New York Convention, enforcement of an arbitral award can be refused where the arbitration agreement is not valid under the law to which the parties have subjected it, or failing any indication thereon, under the law of country where the award was made. This is given effect by section 103(2)(b) of the Arbitration Act 1996.
1.49 In other words, in determining the law applicable to the arbitration agreement, the New York Convention looks first to party choice, then provides a default rule which identifies the law of the seat. This is the same as section 6A.
1.50 A survey of 80 jurisdictions recorded that 51% similarly apply law of the seat (as does the default rule in Scotland), 34% apply the law of the main contract, 9% adopt a validation approach (like Switzerland), and 6% adopt an “international rules” approach (like France) (see paragraph 12.43 of our final report). So section 6A is not only compatible with the New York Convention, it also aligns with the majority approach worldwide (and including Scotland).
1.51 As for the case of Kabab-Ji v Kout [2021] UKSC 48, that involved an arbitration seated in France, with English law as the express choice of law to govern the main contract. The Supreme Court held that English law would govern the arbitration agreement, and refused enforcement of the in-coming French arbitral award. Then the French courts held that the arbitration agreement was governed by French law as the law of the seat, and upheld the arbitral award.
1.52 The Supreme Court in Kabab-Ji said that a choice of law to govern the main contract was consistent with “subjecting” the arbitration agreement to that law too. We do not disagree. But it is also consistent with the New York Convention for a domestic law to prescribe that an arbitration agreement is “subjected” to a governing law only when that choice is made expressly by reference to the arbitration agreement itself, and not indirectly by reference to a choice of law to govern the main contract. Which is to say that the New York Convention wording is open-textured, and compatible with a range of approaches. For example, no-one suggests that the approach in France or Switzerland or Scotland is a breach of the New York Convention.
Interpretation of the New York Convention in light of section 6A
1.53 So far we have discussed whether section 6A is compatible with the New York Convention. A slightly different question (and a new question) is whether the New York Convention should be interpreted to align with section 6A: that is, whether section 103 of the Arbitration Act (which enacts article V.1(a) of the New York Convention) should be interpreted as requiring an express choice of governing law for the arbitration agreement (and not merely an implied choice).
1.54 The court in Kabab-Ji acknowledged that there was no international consensus on how to interpret article V.1(a) of the New York Convention [at 32]. Indeed, what is especially notable about Kabab-Ji was how opposite conclusions were reached by the courts in England and France. In the absence of international consensus, the Supreme Court applied “first principles” – being those it articulated in Enka v Chubb.
1.55 On the one hand, the courts could interpret section 103 independently of section 6A, in light of the fact that section 103 applies, not a domestic rule, but an international convention. On the other hand, section 6A marks a deliberate departure from Enka v Chubb, which in turn means that any case which relied on Enka v Chubb has limited precedential value so that, should a case like Kabab-Ji arise again, it could be considered afresh.
1.56 All of which is to say that section 6A is no impediment to a proper interpretation of section 103 – whatever the courts, after full argument, think that should be.
Confidentiality
Questions
1.57 Professor Green explained why the Bill did not address confidentiality. Is there any further point that Professor Green would like to make in the light of the written evidence of the City of London Law Society?
1.58 Should statutory safeguards be introduced to give effect to the “public interest” referred to in section 1(b) of the Arbitration Act as has been suggested?
1.59 In particular should the Bill clarify, especially in the light of Federal Republic of Nigeria v Process and Industrial developments Ltd [2023] EWHC 2638 (Comm):
(1) The duties of arbitrators when faced with allegations of corruption;
(2) The powers of arbitrators in such circumstances;
(3) The ways in which abuse of confidentiality might facilitate or be used to conceal corrupt practices?
Response
No general rule of confidentiality
1.60 Some stakeholders, including the City of London Law Society Arbitration Committee, suggested that the Arbitration Act 1996 should contain a default rule that arbitrations are confidential. We did not recommend that. Our reasons are set out in chapter 2 of our first consultation paper, and in chapter 2 of our final report. 85 consultees responded on this topic: 56 agreed with us, 23 disagreed, and 6 gave other responses.
1.61 In short, we did not think that a one-size-fits-all rule would work. In practice, some arbitrations are confidential, and some are not, and in both cases the boundary of confidentiality can vary according to context. In some areas, there is a live debate as to where to draw those boundaries – which is to say, there is not yet a definitive consensus. The case of Nigeria v P&ID itself has fuelled the debate, and is often given as an argument against confidentiality. The rules of arbitral institutions often address confidentiality, and again in a variety of ways. If parties agree that their arbitration is confidential, that provides the maximum protection under English law already. But English law also recognises exceptions when an agreement of confidentiality can be overridden. This includes where justice or public interest requires disclosure. Overall, the law seems to be in a sensible place, and there do not appear to be any particular problems in practice.
The case of Nigeria v P&ID
1.62 Nigeria v P&ID [2023] EWHC 2638 (Comm) is an extreme case, and we should be careful of over-reactions.
1.63 In that case, Mr Justice Knowles set aside an arbitration award against the Nigerian government worth US$11 billion. The arbitration process was tainted by corruption. A bribed Nigerian official had a role in bringing about the main contract and the arbitration agreement. And when the arbitration itself was underway, Nigeria’s legal team were bribed to pass privileged information to the other side, P&ID. This only came to light when Nigeria changed its legal team to challenge the arbitration award, and when P&ID appointed new lawyers who abided by their own professional duties. Also, P&ID’s witness evidence in the arbitration was knowingly false. At the conclusion of the case, Knowles J referred P&ID’s previous lawyers to the disciplinary processes of the Bar Council and the Law Society.
1.64 Had this occurred in court proceedings rather than arbitration proceedings, it is difficult to see why anything would have been different. The arbitrators were very experienced, including Sir Anthony Evans (former Lord Justice of Appeal), and Lord Hoffmann (former Law Lord) as chair. It is difficult to see why they would have acted differently as judges rather than as arbitrators. The bribery, breach of legal professional privilege, and knowingly false evidence was not known to the tribunal. However, Knowles J said that he did not consider that the tribunal did “all that it could to find out more” about some issues which seemed to be him to raise flags. But this is a criticism of the arbitrators, not of arbitral law.
1.65 The arbitration clause provided for London as the “venue” but for Nigerian arbitration legislation to apply to any dispute. Nevertheless, Nigeria applied to the Commercial Court to set aside the arbitral award under section 68 of the Arbitration Act 1996 on the ground of serious irregularity. That application was successful. To that extent, it shows that section 68 works. The Arbitration Bill makes no change to section 68, and none appears necessary.
1.66 Knowles J made the following final observations. Major commercial contracts involving states need to be properly drafted (unlike this one). Disclosure of documents during dispute resolution is very important (and was not done honestly here in the arbitration). Parties need good legal representation (when here it was poor in the arbitration). All this is true – but is nothing which the Arbitration Bill could appropriately mandate or ensure.
Arbitrator duties
1.67 So what are the powers and duties of arbitrators when faced with allegations of corruption? First, note that the arbitrators in Nigeria v P&ID were not faced with allegations of corruption – that was the problem (both sides were corrupted to hide the truth).
1.68 In Nigeria v P&ID, the court was faced with allegations of corruption, and this is what the court did: issued a judgment which prevented the corrupt party from benefitting from the corruption; published the judgment; referred the corrupt lawyers to their professional bodies.
1.69 Arbitrators can do all these things already.
1.70 First, arbitrators can issue an award which prevents the corrupt party from benefitting from their corruption. Arbitrators are already under a statutory duty to be impartial and to reach a fair resolution of the dispute (section 1 and section 33).
1.71 Second, arbitrators can refer corrupt lawyers to their professional bodies – as anybody can. And we might add: arbitrators too can be reported to their professional bodies or appointing institutions if the arbitrators fail to live up to their professional duties.
1.72 Third, arbitrators can apply to court for permission to publish their award, as follows.
1.73 In Emmott v Michael Wilson & Partners Ltd [2008] EWCA Civ 184, there had been a confidential arbitration. One party then brought legal proceedings. The other party applied for permission to disclose documents from the arbitration in the legal proceedings. This was approved, so that the courts, which were considering the same issues as arose in the arbitral proceedings, would not be misled by the presentation of distorted evidence.
1.74 The Court of Appeal said that a party can apply to court to restrain any disclosure in breach of confidentiality, or (a mirror image application) for permission to disclose.[1] It would presumably be open to an arbitrator to make that application, because an arbitrator is bound by the duty of confidentiality as much as the parties.
1.75 The Court of Appeal said that disclosure will be permissible in a range of circumstances, including:
…the first is where there is consent, express or implied; second, where there is an order, or leave of the court (but that does not mean that the court has a general discretion to lift the obligation of confidentiality); third, where it is reasonably necessary for the protection of the legitimate interests of an arbitrating party; fourth, where the interests of justice require disclosure, and also (perhaps) where the public interest requires disclosure.[2]
1.76 We add as follows. On the one hand, it may be necessary for publication to be redacted in parts, so that only the public interest elements are published, rather than all the details. On the other hand, an arbitrator is immune from liability anyway if they act in good faith (section 29), and publishing what is relevant in the public interest would surely be an act of good faith.
1.77 Emmott was endorsed by the Supreme Court in Halliburton v Chubb.[3] Disclosure in the interests of justice is also an exception acknowledged by arbitration authors.[4]
Conclusion
1.78 Our overall conclusion is that there is no need for any law reform here. The common law already provides a nuanced and flexible approach with sufficient coverage. Whereas we still think that a one-size-fits-all rule risks being too blunt and perhaps soon out of date.
21 February
