When is an arbitrator conflicted? Court of Appeal decides

April 24, 2018

Update from Imran Benson, 23rd April 2018

Many insurers, law firms and blue-chip companies have favourite arbitrators who they regularly recommend for appointments. If an arbitrator is in the happy position of being regularly appointed by a particular user, is this something which should be disclosed to the user’s opponent in the arbitration and/or can it constitute a conflict on the grounds of bias?

In Halliburton v Chubb [2018] EWCA Civ 817, the English Court of Appeal had to resolve this question. Halliburton had settled claims arising out of the Deepwater Horizon disaster in 2010 (where a BP oil rig had exploded causing a vast oil leak). It sought to make a claim against its insurers, Chubb, who resisted the claim. The dispute went to a 3-panel arbitration with Halliburton and Chubb each appointing one member and, since they could not agree, the Court appointing the third arbitrator, M, as chairman. M was Chubb’s preferred choice. M told everyone that he had been appointed by Chubb on arbitrations in the past and was acting in two references where Chubb was involved. M was then appointed by Chubb in a further arbitration also arising from the Deepwater Horizon disaster (but not involving Halliburton). He was then appointed in a third arbitration connected to the disaster (but not involving Chubb or Halliburton).

Halliburton eventually found out about the two other appointments and asked the arbitrator to explain. By this stage the other two arbitrations had only gotten as far as legal issues, rather than factual findings. The arbitrator wrote some thoughtful letters expressing regret at not giving full disclosure and said he would consider resigning from the other two appointments once the legal issues had been resolved. Halliburton was dissatisfied and brought a challenge to M’s appointment. In the meantime the arbitration continued and handed down an award in Chubb’s favour. Halliburton’s appointed arbitrator concurred in the result but issued a document expressing profound disquiet about the arbitration’s fairness.

Halliburton argued that the non-disclosure was evidence of an unconscious bias and M should be removed and the award set-aside. This challenge failed.

The CA restated the test on bias: would the fair-minded and objective observer, having considered the facts, conclude that there was a real possibility of bias? Halliburton’s main complaint was that Chubb, through the other appointments, had an opportunity to influence the arbitrator, share knowledge and assess his approach, in a way which was not available to Halliburton. It referred to a few cases where common appointments could mean an arbitrator was not appropriate as an appointee – but the CA said that this involved a different test to disqualification for bias. An arbitrator was, like a Judge, to be assumed to be trustworthy and that they should approach every case with an open mind and appointments in multiple references arising from the same matter was not sufficient to disqualify. That said, the arbitrator should disclose matters – even if borderline – which might lead the objective observer to conclude there was a real possibility of bias. Although the IBA, ICC and LCIA rules talk about the need to disclose matters which in the subjective opinion of a party might raise doubts as to impartiality or independence, this subjective test did not represent the general English law. Non-disclosure can fortify the view that there is bias.

On the facts, the Court decided that the degree of overlap did not give rise to a case of apparent bias but M ought, nonetheless, to have disclosed the other appointments both as a matter of good practice in international commercial arbitrations and English law. Nonetheless, this did not impugn him sufficiently and the challenge was dismissed.


The English Court of Appeal has taken the customary approach of upholding the autonomy and integrity of arbitrations by: (1) encouraging disclosure and (2) finding an absence of disclosure was not, on these facts, sufficient to doubt the arbitrator.

The Court was not particularly impressed by the suggestion that multiple appointments by the same user was evidence of bias. But in the world of investor-state treaty arbitrations (where ICSID publishes the decisions) there is a body of hotly contested research showing that a range of biases can arise from multiple factors, including being regularly appointed by one party. Indeed, in one case the tribunal accepted the real-politick behind appointments when it said:

“The suggestion […] that multiple appointments are explicable on the basis of a party’s perception of the independence and competence of the oft appointed arbitrator is in our view unpersuasive. In a dispute resolution environment, a party’s choice of arbitrator involves a forensic decision that is clearly related to a judgment by the appointing party and its counsel of its prospects of success in the dispute. In our view, multiple appointments of an arbitrator are an objective indication of the view of parties and their counsel that the outcome of the dispute is more likely to be successful with the multiple appointee as a member of the tribunal than would otherwise be the case”

As to practical tips, it is much easier to challenge an arbitrator pre-appointment on unsuitability rather than on bias. When selecting or opposing, ask about his or her other appointments whether on behalf of a party or on a related subject matter. In the US there is a sophisticated practice of pre-appointment research – and it seems that this approach should be taken up internationally.

Imran has a busy commercial, insurance and arbitration practice and in 2017 was in the Commercial Court successfully upholding a coverage declinature by QBE on an accountants’ policy.

Related barristers