Mishcon de Reya page structure
Site header
Main menu
Main content section

Supreme Court judgment in Halliburton v Chubb: named but not shamed

Posted on 27 November 2020

This morning the Supreme Court finally handed down its highly anticipated judgment on an issue close to the heart of every arbitrator: bias. In Halliburton Company v Chubb Bermuda Insurance Ltd (UKSC 2018/0100), the Supreme Court Justices unanimously dismissed Halliburton's appeal from the Court of Appeal, and spared the blushes of the man at the centre of the controversy. Penning the lead judgment was Lord Hodge (joined by three others) while Lady Arden  added additional points.

The case raised hugely important principles relating to the circumstances in which an arbitrator in an international arbitration may appear to be biased by not disclosing the fact of having been appointed in related arbitrations. It brought together three key duties, not easily reconciled: impartiality, disclosure and confidentiality.

Due to the impact of the judgment on international arbitrations conducted under the laws of England & Wales, the Supreme Court granted leave to intervene to five prominent arbitration institutions. Mishcon de Reya LLP's partner Louis Flannery QC and associate Sophia Louw represented the Chartered Institute of Arbitrators.

The Underlying Facts

The dispute arose out of the Gulf of Mexico BP oil spill disaster that happened in in 2010. Transocean owned the Deepwater Horizon oilrig and Halliburton (H) provided services like cementing. Both BP and H settled various legal claims relating to the oil spill.

H made a claim on its insurance policy with Chubb (C). It sought repayment of the sum it had paid out in settlement. C refused to pay, saying that the settlement was unreasonable and that it had not been consulted, so it had no obligation to indemnify H. The policy was on the now ubiquitous "Bermuda Form", i.e. subject to New York law but providing for London arbitration. H started arbitration proceedings in London. Both sides appointed an arbitrator. They could not agree on the chair of the tribunal. In 2015, the English High Court (Flaux J, as he then was) appointed the candidate proposed by C as chair. He had been known only as "M" during the course of the case, but the Supreme Court has now named him as Kenneth Rokison QC.

A few months after his appointment in the H-C arbitration, Rokison was appointed as an arbitrator in two other arbitrations also arising out of the oil spill and both essentially considering the same or similar issues. First, a claim by Transocean against C; secondly, a claim by Transocean against another insurer. In the first arbitration, Rokison was appointed by C. Neither it nor its solicitors (the same firm) notified H or its solicitors. In the second arbitration, Rokison was appointed as chairman by agreement of the parties. Crucially, he failed to disclose these appointments to H while the arbitration against C was ongoing.

When H learned (by reason of court proceedings in the US revealing Rokison's involvement in the other arbitrations) of the later appointments, it challenged Rokison’s impartiality. R denied any wrongdoing, blamed it on an innocent oversight, and explained that it had not occurred to him that he had a duty to disclose the appointments. H applied to the High Court to have Rokison removed as chair under section 24(1)(a) of the Arbitration Act 1996 (the "1996 Act"). This provides that a party may apply to the court to remove an arbitrator if “circumstances exist that give rise to justifiable doubts as to his impartiality”. The application was founded on a submission that Rokison’s conduct in not disclosing the later appointments gave rise to an appearance of bias.

The High Court concluded that Rokison's acceptance of the appointments did not give rise to any justifiable concerns as to his independence and he was under no obligation to disclose the appointments. The Court of Appeal upheld the High Court's judgment. H appealed to the Supreme Court.

The Supreme Court ruling

The core issues were:

  • whether and to what extent an arbitrator may accept appointments in multiple references concerning the same or overlapping subject matter with only one common party without giving rise to an appearance of bias; and
  • whether and to what extent an arbitrator may accept the multiple references described in the first issue without making disclosure to the party who is not the common party.

Lord Hodge, writing for the majority, made the following key observations in reaching the decision:

1) Impartiality is a cardinal duty of an Arbitrator

The obligation of an arbitrator to act fairly and impartially is a core principle in arbitration. Indeed, it is the first principle set out in section 1 and enshrined in section 33 of the 1996 Act.

Under English law, the duty of impartiality applies equally to party-appointed arbitrators, sole arbitrators and presiding arbitrators.  English law should therefore uphold the rules that support the integrity of international arbitration.  (See paras 49, 63 and 151)

2) The test to establish whether there is a real possibility of bias is objective

The Supreme Court confirmed that the test under English law to establish whether an arbitrator has a real possibility of bias is objective (unlike some institutional rules, which seem to import a subjective test – i.e. "in the eyes of the parties"). The "fair-minded and informed observer"  should take into account a host of considerations including the factual matrix of the case, the perceptions of the role of the arbitrator in the case and expectations regarding the degree of independence of an arbitrator. An arbitrator's failure to make a disclosure of acting in multiple cases is a factor of consideration for the objective observer to take into account.

For example, inter-related appointments are not common in ICC arbitrations, but are common in GAFTA and LMAA arbitrations. As such, the practice of accepting multiple appointments in overlapping cases does not inherently give rise to a real possibility of bias, as long as the arbitrator can approach each arbitration objectively and with an open mind. In other words, the test is highly fact-specific. 

This assessment should be made at the date of the hearing to remove the arbitrator, and not at the date of the arbitrator's acceptance of the second appointment. (See paras 127 – 131; and 152)

3) The legal duty of disclosure:

The Court clarified that the legal duty of disclosure does not override the arbitrator's duty of privacy and confidentiality in English law. Unless the parties to the arbitration have expressly or implicitly waived their right to disclosure, an arbitrator sitting over multiple related arbitrations must make disclosure of the existence of the arbitration and the identity of the common party.  This disclosure is not just a question of best practice but is a matter of legal obligation. It is encompassed in the statutory duties under section 33 of the 1996 Act.  This was in line with submissions made by the ICC, LCIA and CIArb. 

However, disclosure can be made only if the parties to whom the obligations are owed give their consent. Such consent may be inferred from the arbitration agreement in the context of the practice in the relevant field. Thus, an arbitrator will have to decline a second appointment if the arbitrator does not obtain the consent of the parties in a prior related arbitration to disclose confidential matters to the prospective arbitration and vice versa.

While the Supreme Court acknowledged that in some forums such as GAFTA and LMAA overlapping and multiple appointments are an accepted feature, under English law multiple appointments must be disclosed in the absence of an agreement to the contrary.  Such disclosure is in the public interest in order to uphold the integrity of arbitration as a system of alternative dispute resolution.

Where there is real possibility of bias, the arbitrator is under a legal duty to disclose the overlapping appointments, unless the parties have agreed otherwise.  As such, an arbitrator’s duty of privacy and confidentiality would not prevent limited disclosure such as the identity of the common party, a statement that the second arbitration arose out of the same set of facts and whether the proposed appointment was by the parties or by another means.

Conclusion:

On the facts of this case, while confirming that (a) English law imposed a duty of disclosure on all arbitrators and prospective candidates; (b) that Rokison breached that duty in not disclosing the later appointments to H, the Supreme Court followed Popplewell J and the Court of Appeal in rejecting the application.  The Court was not persuaded that the objective observer would have found a real possibility of bias on Rokison's part at the date of the hearing of his removal, which is the key date on which his conduct and all the circumstances were to be judged. Lord Reed provided his reasons as follows:

  1. By the date of hearing of his removal, Rokison had provided an explanation of his oversight to disclose his subsequent appointments, which was accepted by H.
  2. His second appointment followed six months after his first, which explained why he did not see the need to inform H of his appointment in the second arbitration.
  3. The Court also accepted what it described as Rokison's "measured" view that the second and third arbitrations would not overlap in legal evidence or submissions with the first. As a result, there was no likelihood of C gaining an advantage.
  4. Rokison had received no secret financial benefit from his subsequent appointments.
  5. The Court was satisfied that there is no basis for inferring unconscious bias on Rokison's behalf in the form of subconscious ill-will in response to his challenge as arbitrator.
  6. Lastly, the Supreme Court even allowed for the lack of clarity on the law as a basis for not treating Rokison as having crossed the line.

The result is, to say the least, disappointing, and not just in the time it took to be propounded. It effectively upheld the Court of Appeal in saying that Rokison had a duty to disclose the later appointments to H, but like the Court of Appeal, denuded that duty of any teeth at all. In all these cases, the in-built weakness of "the system" is in not being able to open up the mind of the arbitrator, or of those who were involved in the multiple appointments. The nearest we would ever get is to put them in the hot seat and have their testimony examined in court. However, there is no reported case under section 24's bias heading of an arbitrator being cross-examined. Their "innocence" is therefore not only presumed; it is effectively unchallengeable except if it is countered by compelling documentary evidence, which rarely exists e.g. to show a bribe or other financial reward or promise of future appointments, etc. There is no basis on which any suggestion of dishonesty or indeed bias can be made against an arbitrator without compelling evidence. Our ethical code as solicitors and counsel prohibits our making such an allegation without clear evidence in support. And without the main witness being available to be cross-examined, the chances of discovering wrongdoing are always going to be slim.  Rokison's "measured" response to the challenge should, with respect, never have been relevant. He had been appointed in later related arbitrations by a common party (in one case) and did not disclose those appointments, despite the exact same issue arising in all three arbitrations (namely the reasonableness of the settlement made by the policy holder). Rokison had even made clear when challenged that he had considered the IBA Guidelines on Conflicts of Interest, which he "did not think" required him to disclose the later appointments. In other words, he had quite obviously and quite properly given it careful consideration before deciding – for whatever reason – not to reveal the full truth to H.

Given that the same law firm that had pushed Rokison into position as presiding arbitrator in the first arbitration appointed him in the two later arbitrations, and in all three references for the insurer, they and he alone were privy to the multiple appointments (and the potential overlap of issues), but neither said anything to H or its lawyers. Inexplicably, neither of these points were run by H in the appeal. It did at least run one point, which concerned the asymmetry of information, i.e. Rokison learning something in one arbitration that would infect his thinking in another. But the Court ruled that with the benefit of hindsight that concern was misplaced. CIArb argued that it was nothing to the point. Hindsight should never be relevant. At the point of the later appointments, Rokison could not possibly have guaranteed to H that he would never be infected by knowledge gained in one of the other two arbitrations. It was open to the Court to set a bright line example, and make clear that once the breach of the duty to disclose was established, there should be a presumption of apparent bias, which should mean the onus passing to the arbitrator to demonstrate otherwise. It is the only way that those with something to hide will know that they are at risk immediately of being removed if later discovered (as should have happened to Rokison).   

At least on one issue, the judgment produced a new and important test for what an arbitrator, or prospective arbitrator, was required to disclose. The answer is: any fact or circumstance which might reasonably give rise to justifiable doubts as to impartiality (see para 108). This is, we think, a broader and more appropriate test, as put forward by Merkin & Flannery on the Arbitration Act 1996 at pp.286-287, which was cited by Lord Hodge with approval.

How can we help you?
Help

How can we help you?

Subscribe: I'd like to keep in touch

If your enquiry is urgent please call +44 20 3321 7000

Crisis Hotline

I'm a client

I'm looking for advice

Something else