United States: Keep Calm And Carry On Arbitrating: The Latest On Apparent Bias In Arbitration

One never tires of considering the impartiality and independence of arbitrators. Thankfully, cases where arbitrators show flagrant or actual bias are few and far between: the brown envelope is seldom offered, and rarely accepted. However, challenges to arbitrators alleging 'apparent' lack of impartiality or bias are by no means uncommon. In short, apparent bias means that there are circumstances which would cause a reasonable observer to have doubts as to the independence of the arbitrator. This is an objective test: the arbitrator may well be able to put the matter entirely out of his or her mind, and be perfectly capable of rendering an unimpeachable, fair and impartial award. But where it looks as if that might not have been the case, 'apparent bias' still requires the arbitrator to step down, or else be disqualified or risk having the award he or she has rendered challenged successfully. A particular, recurring issue relating to apparent bias relates to 'professional relationships': this might be a relationship between the arbitrator personally and a particular client, or a relationship between the arbitrator's law firm and a relevant party. Both these situations have recently been the subject of judgments by the English Courts in challenges to arbitral awards or applications to remove an arbitrator.

The Commercial Court disagrees with the IBA Guidelines

There has been an interesting recent development. In W Ltd v M SDN BHD [2016] EWHC 422, the Commercial Court considered the International Bar Association's Guidelines on Conflicts of Interest in International Arbitration, ultimately concluding that they were (as they stand) not a reliable guide. The IBA Guidelines, though they clearly do not have the force of law, are often referred to in international arbitration proceedings. They include a 'traffic light' system of classifying certain situations as 'red, yellow and green', with some 'red' situations amounting to a non-waivable conflict. Many practitioners would expect an arbitrator who finds himself or herself in such a 'red' situation to resign. The authors of the IBA Guidelines take the view that serious 'red' conflicts of interest are not something the parties should be allowed to simply let go. This is because the arbitrator's position, and hence the award, would be fundamentally undermined or open to questions. As they put it, acceptance does not cure the default when it comes to non-waivable conflicts.

The IBA Guidelines were amended and updated in 2014. An international working group comprising eminent practitioners (a total of 27 members from 19 countries) took just under two years to produce the latest version. In the introduction, the authors note that:

General standards and principles are, accordingly, proposed. As regards apparent bias, the IBA puts the matter in these terms:

It can be seen that as far as the IBA is concerned, a 'red' conflict always amounts to the 'appearance of bias'. The particular conflict that the Commercial Court had to consider appears in paragraph 1.4 of the Non-Waivable Red List:

Before turning to the facts, it is worth noting that paragraph 1.4 is widely expressed. Pursuant to it, apparent bias exists where the arbitrator's firm (not the arbitrator) advises an affiliate of a party to the arbitration (but not the party itself), and this generates 'significant' income.

The arbitrator's law firm carries out significant work for an affiliate of party

In W v M , the arbitrator was a senior Canadian QC, who gave evidence that he had spent the last six years or so almost exclusively sitting as an arbitrator. He was, however, still a partner in a Calgary law firm. The arbitrator explained to the Commercial Court that he was semi-retired from the firm and really only a partner in name. He took no part in partnership business. He described himself as a sole practitioner who used the secretarial and administrative support services of the firm. He had also been given his very own department for the purposes of remuneration: he was the sole member of the firm's international dispute resolution group.

The situation that was said to amount to a conflict of interest came about as a result of an arbitration that was commenced in April 2012. The dispute related to a construction project in Iraq. The respondent in the arbitration was a subsidiary of company "P". On 18 May 2012, the arbitrator made a declaration of independence. In June 2012, a public announcement was made that "P" would acquire another company, "Q". Following that acquisition, both "Q" and the respondent in the arbitration would be subsidiaries of "P". The arbitrator's law firm had significant links with "Q". The managing partner was "Q"'s company secretary, while the senior partner was both a board member and a shareholder. "Q" was also a major client of the firm. In December 2012, during the arbitration, "P"'s acquisition of "Q" completed. On completion, the senior partner and managing partner both resigned from their positions with "Q", but the law firm continued to advise "Q". The Court concluded that the legal services provided by the firm to "Q", and the fees earned as a result, were 'substantial'. The arbitrator did not disclose to the parties that, as from December 2012, he was sitting in a matter where the respondent was a subsidiary of a substantial, current client of the arbitrator's firm. He explained to the Commercial Court that the firm's system for checking conflicts failed to warn him of this, because the acquisition had not been correctly labelled in the system. He did not otherwise learn of the relationship between the companies, despite there having been some publicity around the acquisition. He was clear that, had he known of the acquisition, he would have made a disclosure. Instead, unaware of the situation, the arbitrator proceeded to issue two awards in October 2014 and March 2015. It is worth noting that throughout the arbitration, the arbitrator did disclose other matters to the parties as to potential conflicts, but these were all immaterial.

The test for apparent bias under English law

The facts brought the matter squarely within paragraph 1.4 of the non-waivable red list under the IBA Guidelines. The claimant proceeded to challenge the awards under Section 68 of the Arbitration Act 1996, on the basis that there was a serious procedural irregularity which caused substantial injustice. English law applied to determine whether there was apparent bias. The leading case in that regard is Porter v Magill [2002] AC 357, in which the House of Lords set out the test at common law, namely whether:

'Bias' in English law means the "absence of demonstrated independence or impartiality" (Yiacoub v The Queen [2014] UKPC 22), which is perhaps a less pejorative meaning than the word may carry in ordinary usage. 

Beyond that, the authorities contain a little further guidance as to what characteristics are ascribed to the 'fair minded and informed observer', who is the arbiter of apparent bias, as a matter of law:

• They are neither complacent nor unduly sensitive or suspicious (Helow v Secretary of State for the Home Department [2008] UKHL 62).
• Although not a lawyer, they are assumed to have all the facts which bear on the question and to be aware of the way in which the legal profession operates in practice (A v B [2011] 2 Lloyds Rep 591).
• They reserve judgment until both sides of the argument have been understood, and do not approach the matter as the complainant does. They do not share the assumptions made by the complainant unless those assumptions can be justified objectively (A v B).
• They take a balanced approach and appreciate that context is important (A v B).

In W v M, Knowles J did not cite all the above authorities, but he did specifically address the relevance of the further explanations provided by the arbitrator to the Court in response to the challenge. As noted by the House of Lords in Helow, apparent bias is an objective matter, to be determined as a matter of law. The Court may take note of what the arbitrator says he or she knew at the time, though there will be no cross-examination. However, any assertions as to what impact such knowledge had on the arbitrator's state of mind will be irrelevant, and of no importance to the Court as they cross the line and become subjective matters.  

Keep calm and carry on arbitrating

Turning to the facts, Knowles J found without hesitation that a fair minded and informed observer would not see any real possibility that the arbitrator was biased, or not impartial. The learned judge noted that the arbitrator had checked for conflicts, and had made disclosures where matters had been drawn to his attention. He concluded that

Knowles J seems to have felt that this was a clear-cut case, and that any uncertainty had only arisen because of the IBA Guidelines. Considering that the IBA Guidelines are the product of considerable collaboration between international arbitration practitioners, and are often referred to, the judge went on to consider them more closely, even though that would not have been necessary to dispose of the matter. Knowles J noted that under paragraph 1.4, the parties appeared unable to waive any conflict where the arbitrator's law firm had advised a subsidiary of a party to the arbitration, and the arbitrator is neither involved, nor even aware of, the advice. It is apparent that Knowles J felt this situation was some distance removed from a case where apparent bias could be assumed or taken for granted, and that each such situation called for "case specific judgment". Yet, tracing the various statements of principle and the 'General Standard' in the IBA Guidelines, in all the examples that appeared in the Non-Waivable Red List, the authors state "... that justifiable doubts "necessarily exist" as to the arbitrator's impartiality or independence "in any of the situations described in the Non-Waivable Red List"." The judge criticised that dogmatic position. The reason given in the IBA Guidelines as to why party consent could not excuse a conflict on the Non-Waivable Red List was that the list included instances "... deriving from the overriding principle that no person can be his or her own judge ...": of course, paragraph 2.4 was not one of those instances. Ultimately, Knowles J concluded that the IBA Guidelines could not (yet) be correct.

Did the Commercial Court really drive through a red light?

Taking a step back, is this an instance where English law and the consensus of international arbitration practitioners, assuming that is indeed as set out in the IBA Guidelines, differ? Is English law out of step by being slower to hold that arbitrators are conflicted? It is suggested that the answer is 'no'. Instead, the careful analysis of the IBA Guidelines set out in the judgment of the Commercial Court rather suggests that authors made a simple mistake when drafting paragraph 1.4 of the Non-Waivable Red List. When one reads the relevant passages of the IBA Guidelines as a whole, one is left wondering how paragraph 2.4 ever made it into the list.

General Standard 2(c) in the IBA Guidelines offers a definition of 'apparent bias', or when a reasonable person would have justifiable doubts as to the arbitrator's independence. It provides that:

It can be seen that the IBA Guidelines are concerned with an apparent 'likelihood' that the tribunal is not impartial, whereas English law requires a 'real possibility' of bias (Porter v Magill). General Standard 2(d), in Part I of the IBA Guidelines, is then put in mandatory language:

The IBA Guidelines go on to address circumstances very similar to those that came before the Commercial in W v S in General Standard 6, headed Relationships:

The last sentence in particular is worth noting: it specifically says that where one of the parties is the member of a group with which the arbitrator's law firm has a relationship, then that calls for a consideration of the individual case – as opposed to a foregone conclusion that there is apparent bias. That is a direct contradiction of the inclusion of paragraph 2.4 in the Non-Waivable Red List.

The IBA Guidelines also provide further explanatory notes in relation to General Standard 6. Those notes make a number of common sense points. They recognise that it is a fact of life (or at least modern legal practice) that arbitrators may be members of large, international firms. While an arbitrator ought generally be identified with his or her law firm for the purposes of considering conflicts of interest or apparent bias, it is specifically said:

With the above in mind, it is difficult to understand why the IBA Guidelines also say that apparent bias must necessarily exist in all cases identified on the Non-Waivable Red List, which of course includes paragraph 1.4:

"1.4 The arbitrator or his or her firm regularly advises the party, or an affiliate of the party, and the arbitrator or his or her firm derives significant financial income therefrom."

The underlined words were added in the 2014 revision to the IBA Guidelines. That looks to have been an infelicitous amendment, when the rest of the document is considered. It leads to a direct contradiction within the IBA Guidelines. It cannot be squared with General Standard 6, as further explained in the notes. Knowles J was plainly aware of this, although (perhaps out of "diffidence") he did not quite say that something must have gone wrong with the addition of those five words to paragraph 1.4, when one reads the IBA Guidelines as a whole. In its old guise, paragraph 1.4 made a lot more sense: if it is the arbitrator who advises a party, or even an affiliate, then one can see why that creates a problem. As a practicing lawyer, the arbitrator will owe duties to an existing client, and it would be inappropriate to insist on strict corporate personality and say that the lawyer-client relationship for these purposes excludes an affiliate of the entity that the arbitrator is personally acting for or advising. 

It will be recalled that the Non-Waivable Red List appears in Part II of the IBA Guidelines, entitled "Practical Application of the General Standards", which specifically note that the General Standards "... should control the outcome". If one were construing the IBA Guidelines like a contract, one might conclude that the intention of the parties was that General Standard 6 should take precedence over the amendment to paragraph 1.4, so that the matter remained case-specific. After all, if the intention had been to the contrary, so that automatic disqualification really was to follow, then one would have also expected a substantive rewording of General Standard 6 and the explanatory notes.

A red card for another arbitrator

W v S does not tell us anything further about when a case-specific review of a relationship between an arbitrator and a party might lead to a finding of apparent bias. Another recent decision of the English Courts, Cofely Ltd v Bingham and Knowles Limited [2016] EWHC 240 (Comm), however, does, although the facts of that case were out of the ordinary, to say the least. In Cofely, Mr Justice Hamblen removed an arbitrator following a challenge under the Arbitration Act 1996. Section 24(1)(a) of the Arbitration Act 1996 gives the Court the power to remove an arbitrator if "circumstances exist that give rise to justifiable doubts as to his impartiality." This reflects the common law test of apparent bias. 

At the heart of the case was the fact that this particular arbitrator had received 18% of his appointments, and 25% of his earnings, over the preceding three years through appointments (either as arbitrator or adjudicator) by a particular firm of claims consultants, who were a party to the arbitration from which he was unceremoniously removed. The case also shows how playing tactical games in the hope of securing the appointment of arbitrators who are thought to be favoured candidates can seriously backfire. 

The claimant, Cofely, had employed a firm of construction consultants to prepare a claim against the employer under a construction contract for the design, construction and subsequent operation of energy service to the Olympic Park and the Westfield Shopping Centre in London. The claim was primarily concerned with an entitlement to an extension of time and prolongation costs incurred by Cofely because completion of the works was delayed. Preparing it proved a costly undertaking: the costs of the claims consultants eventually exceeded £1 million, yet proceedings had not been launched. Cofely then entered into a success fee agreement, under which the claims consultants would become entitled to certain payments upon the occurrence of a series of events relating to the progress of the proceedings. With that new arrangement in place, an adjudication was commenced. Soon enough though Cofely became concerned by the advice it received from, and the strategy adopted by, the claims consultants during the adjudication. This prompted Cofely to enter into negotiations directly with the employer, which led to a settlement. The claims consultants, who had been unaware of the discussions and had not approved the settlement, alleged that this was a breach of the success fee agreement, and commenced an arbitration under that contract against Cofely. The claim appears to have been for the loss of a chance to recover further payments that would have been due to the claims consultants if the claim in the adjudication had been successful. It was put at more than £3.5 million.

In their arbitration against Cofely, the claims consultants approached an appointing authority to secure the appointment of a sole arbitrator. In their request to the appointing body, the claims consultants stated that it was desirable for the arbitrator to have experience of delay analysis and quantity surveying. They requested the arbitrator who was ultimately appointed by name: this produced the desired result, although Cofely had proposed a different candidate. The arbitrator duly provided a signed statement of acceptance. In that statement, he gave no further details or information in the space provided below the following question:

The arbitration proceeded. Cofely sought a determination of a preliminary issue, as to whether the actual value of the prolongation claim needed to be established in order to work out any entitlement that the claims consultants might have under the success fee agreement. The arbitrator suggested that investigating the claim might very well be necessary, but did not decide the point. Progress in the arbitration seemed sluggish, especially (Cofely thought) when Cofely wanted something addressed. Then matters took an unexpected turn. 

Playing games with the appointment process

In late 2014, Ramsey J, sitting in the Technology and Construction Court, decided Eurocom Ltd v Siemens Plc [2014] EWHC 3710. Eurocom had sought to enforce an adjudication decision against Siemens, by applying for summary judgment (the usual procedure). The adjudicator who had found in favour of Eurocom was the arbitrator in Cofely's arbitration, and he had been appointed in the Eurocom adjudication by the same appointing authority as in Cofely's arbitration. Eurocom had been represented by the same claims consultants in the adjudication. One of the grounds on which Siemens resisted enforcement was that the appointment of the adjudicator had been invalid. Ramsey J agreed. He found that there was a very strong prima facie case that the claims consultants had made a fraudulent misrepresentation when requesting the appointment of the arbitrator. The claims consultants had filled in a box on the appointment form asking whether there were any adjudicators who would have a conflict of interest. There, they had stated that a fairly long list of potential candidates 'should not be appointed'. The judge found that this might very well be a fraudulent misrepresentation: the implication given to a reader of the form was that the listed candidates all had conflicts of interest, hence they 'should not be appointed' for that reason. In reality, the list consisted of candidates who, as the judge found, the claims consultants simply did not want to be appointed, for whatever reasons, and not because of any known conflicts. Ramsey J found a prima facie case that the claims consultants had been reckless in what they had said to the appointing authority. They had seriously overstepped the mark.

Back to the 'apparent bias' decision: Cofely seems to have been prompted to take action because of evidence given by the claims consultants in Eurocom v Siemens, to the effect that listing 'undesirable candidates' in the box reserved for conflicts on the (same) appointment form was a general practice when requesting appointments. Cofely also suspected that the arbitrator had, in fact, been appointed on many prior occasions (either in that capacity, or as adjudicator) in matters where the claims consultants were engaged, though he had not disclosed any prior relationship in his statement of acceptance. Cofely therefore wrote to the arbitrator asking a number of questions, all aimed at establishing precisely how extensive the prior relationship between the arbitrator and the claims consultants really was. After some considerable toing and froing, with the arbitrator refusing to provide the answers, it eventually transpired that over the preceding years, he had been appointed 25 times (18% of his total appointments) by the claims consultants, and had received 25% of his total income over the last three years out of those appointments. Cofely also ascertained that in the 25 appointments, the arbitrator had found in favour of the clients of the claims consultants in 72% of proceedings.

The arbitration then became rather acrimonious, with (as Hamblen J put it) the arbitrator descending into the arena. The arbitrator convened a hearing, of his own initiative. He informed the parties that he would issue a ruling determining whether he was properly appointed, something neither party had asked for. During that hearing, Cofely repeatedly told the arbitrator that it wanted him to confirm the details of his prior relationship with the claims consultants. The arbitrator refused to do so, and pressed counsel for Cofely to say why any of this mattered. Hamblen J found that the arbitrator was effectively cross-examining Cofely's counsel. Perhaps unsurprisingly, after the hearing the arbitrator issued a ruling declaring himself to have been properly appointed. 

A cumulative approach: apparent bias found to exist

Hamblen J found that there was apparent bias, if a number of matters were taken together. Firstly, there was a prior relationship between the arbitrator and the claims consultants - this generated a quarter of the arbitrator's income over the last few years. Secondly, the claims consultants had a habit of trying to blacklist candidates, which had implications in and of itself:

Thirdly, the arbitrator had three prior appointments where the claims consultants were a party. He should have disclosed that, but had not done so. In this regard, the Court agreed with Cofely that the IBA Guidelines, here the Orange List, helped to show that there was a duty to disclose. The Orange List sets out "a non-exhaustive list of specific situation that, depending on the facts of a given case, may, in the eyes of the parties, give rise to doubts as to the arbitrator's impartiality or independence ... with the consequence that the arbitrator has a duty to disclose such situations ...". Both having been appointed as arbitrator on two or more occasions by a party, or an affiliate, and serving or having served as arbitrator within the last three years in proceedings involving a party, or an affiliate, require disclosure under the Orange List (under paragraphs 3.1.3 and 3.1.5 respectively). Fourthly, the arbitrator had responded in an aggressive and hostile manner to Cofely's reasonable requests for information. In his witness statement before the Court, the arbitrator stated that Cofely had engaged in "... assertive, challenging, perhaps even bullying behaviour ...". Objectively, though, this was not the case. By taking the view that attack was the best form of defence, the arbitrator had displayed a lack of impartiality - added to which the arbitrator continued to question the relevance of his relationship with the claims consultants, and did not see that his conduct in the hearing had been inappropriate, even in his evidence before the Commercial Court. Taken together, these matters gave rise to a real possibility of apparent bias.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.