In Otkritie International Investment -v- Mr George Urumov [2014] EWCA Civ 1315 decided today the Court of Appeal reviewed the circumstances in which a judge should recuse themselves for bias. It is clear that this cannot be done lightly. In a very unusual case the decision of the judge to recuse himself was overturned and the matter remitted back to him.


After a trial the judge gave a 559 paragraph judgment in which he found that numerous defendants had conspired to defraud the claimants. Mr Urumov was a defendant. The claimants subsequently made an application against Mr Urumov for contempt, including an application for committal. The defendants objected to the trial judge hearing the application on the grounds of bias – given the findings made in the judgment. The trial judge, with considerable reluctance, granted that application. The claimants appealed.



  1. It is a basic principle of English law that a judge should not sit to hear a case in which “the fair-minded and informed observer, having considered the facts, would conclude that there was a real possibility that [he] was biased”, see Porter v Magill [2002] 2 AC 357 para 103 per Lord Hope of Craighead. It is an even more fundamental principle that a judge should not try a case if he is actually biased against one of the parties. The concept of bias includes any personal interest in the case or friendship with the participants, but extends further to any real possibility that a judge would approach a case with a closed mind or, indeed, with anything other than an objective view; a real possibility in other words that he might in some way have “pre-judged” the case.
  2. This can give rise to potential difficulties in long running cases where a judge has been case-managing a case and has then to conduct the trial or in cases where a trial has occurred and the judge has then to consider consequential matters such as, in the present case, proceedings for contempt. It is obviously convenient for a single judge rather than different judges to deal with a complex case but the question can arise whether there comes a point where findings made by a judge pre-trial disqualify a judge from continuing with a case or findings made at trial disqualify a judge from hearing consequential matters. This is the question at the heart of this appeal.


The Law

  1. There is already a certain amount of authority on the question whether a judge hearing an application (or a trial) which relies on his own previous findings should recuse himself. The general rule is that he should not recuse himself, unless he either considers that he genuinely cannot give one or other party a fair hearing or that a fair minded and informed observer would conclude that there was a real possibility that he would not do so. Although it is obviously convenient in a case of any complexity that a single judge should deal with all relevant matters, actual bias or a real possibility of bias must conclude the matter in favour of the applicant; nevertheless there must be substantial evidence of actual or imputed bias before the general rule can be overcome. All the cases, moreover, emphasise, that the issue of recusal is extremely fact-sensitive.
  2. Livesey v New South Wales Bar Association (1983) 151 CLR 288 is a comparatively early case decided in favour of a defendant (barrister) who was facing an application to strike his name off the roll of counsel in New South Wales. Ms Bacon, at the time a law student and later an applicant for admission to the roll as a barrister, had provided $10,000 as cash surety for Sellers, a defendant in criminal proceedings, who was then granted bail but absconded. The cash surety was duly forfeited. When Ms Bacon applied to be admitted as a barrister, the Admission Board rejected her application on the basis that she knew full well that the $10,000 surety was Sellers’ money and not her own and could not therefore be used as surety. She appealed to the New South Wales Court of Appeal which heard evidence and concluded that Ms Bacon had not told the truth. Her case was that she had been lent the money by a Ms Altman and that she had been unaware that Ms Altman had herself obtained the money with the help of Livesey (Sellers’ barrister) who had then transported the cash from Victoria to Sydney and visited Sellers in jail before Ms Altman supposedly visited Ms Bacon and offered to lend her the money. In circumstances where both Livesey and Ms Altman “well knew where the money had come from” (page 296) Moffitt P found it impossible to believe Ms Bacon did not. Reynolds JA agreed so that the judges found that

“the bail money had been lodged by Ms Bacon pursuant to a “corrupt agreement” or a “conspiratorial arrangement”… between a number of persons including [Livesey] aimed at achieving Sellers’ release on bail by depositing $10,000 which was in truth his own money or money which was available to him.”

  1. When a subsequent application was made to disbar Livesey, he found that both Moffitt P and Reynolds JA were members of the tribunal hearing his case. He applied for them to recuse themselves but they refused and he was struck off the roll. The High Court however decided (page 300):-

“a fair-minded observer might entertain a reasonable apprehension of bias by reason of prejudgment if a judge sits to hear a case at first instance after he has, in a previous case, expressed clear views either about a question of fact which constitutes a live and sufficient issue in the subsequent case or about the credit of a witness whose evidence is of significance on such a question of fact.”

It was important that Livesey was not a party to Ms Bacon’s original application nor does it appear that he had himself given evidence to the first tribunal.

  1. By contrast many English cases have emphasised that the fact that a judge has made adverse findings against a party or a witness does not preclude him from sitting in judgment in subsequent proceedings and some cases have even emphasised the desirability of his doing so.
  2. In wasted costs proceedings or applications that a witness be liable for costs (the consequences of which can be highly adverse to a defendant, if not as potentially adverse as contempt proceedings) it is usual for the trial judge to be the decision maker. In Bahai v Rashidian [1985] 1 WLR 1337 the claimant’s solicitor had given evidence in support of the claim and the judge had been very critical of that evidence. The defendant sought an order that the solicitor be jointly liable with his client for the defendant’s costs and the solicitor asked for the application to be heard by a different judge. The judge refused to make that order. In this court Sir John Donaldson MR said:-

“I accept that it must always be open to a judge to decline to proceed further with the hearing of any matter on the grounds that he is personally embarrassed by, for example, an appearance of bias. Subject to that, I have no doubt that it was the duty of (the trial judge) having heard and determined the issues in the action, himself to determine all applications as to the costs of the action … the fact that a judge has determined the issues in the action and in doing so has expressed views on the conduct of the parties and of the witnesses, neither constitutes bias nor the appearance of bias in relation to subsequent applications in the action … If the application can only be sustained by proof of serious misconduct or crime, the standard of proof should be higher than would otherwise be the case, but, subject to that, the application should be dealt with the same way as would any other application for costs against a solicitor.”

Parker LJ dissented as to the result on the particular facts of the case, but agreed as to the principle, saying at page 1343G:

“Save in exceptional circumstances, it will be for the judge, who heard the case … to determine the matter on a subsequent hearing … there can be no doubt of this, the judge is dealing with the costs of an action which he has himself heard.”

Balcombe LJ agreed with Sir John Donaldson, saying at page 1346D:-

“I accept that the judge has a discretion to direct that the application be heard by another judge, but the discretion is a judicial one, to be exercised in accordance with settled principles, of which one is undoubtedly that the application should be tried by the judge who heard the action unless there are compelling reasons to the contrary. … A judge properly exercising his judicial function, e.g. by criticising the conduct of a party’s solicitor in the course of his judgment on a matter which he considers relevant to his decision, cannot by that process be said to be biased. Bias is the antithesis of the proper exercise of a judicial function. … If such an application has to be heard by another judge, the procedure will lose its summary character. It will become even more expensive and time consuming than it is already, and the defendants are justified in their contention that the remedy of the party damnified by the solicitor’s misconduct will become illusory”.”

This authority was followed in Re P (A Barrister) [2002] 1 C.A.R. 207.

  1. In long trials where many applications have to be decided in the course of the hearing, a party may persuade himself that a judge is biased against him as a result of his rulings. In Arab Monetary Fund v Hashim (No. 8) (1993) 5 Admin LR 348, it was suggested that Chadwick J should not continue with the case. He refused to recuse himself and, on appeal, Sir Thomas Bingham MR posed the question whether a reasonable and fair-minded person sitting in court and knowing all the relevant facts would have a reasonable suspicion that a fair trial was not possible. He continued (pages 354-355):-

“…Most, if not all, of the cases in which this test has been discussed have been cases of modest dimensions. We know of no case approaching the scale of this where a charge of apparent bias has been made. That makes it the more important to recognise, as we understand to be agreed, that the hypothetical observer is not one who makes his judgment after a brief visit to the court but one who is familiar with the detailed history of the proceedings and with the way in which cases of this kind are tried. We find assistance in observations made in the Supreme Court of New South Wales by Mahoney JA in Vakauta v Kelly (1988) 13 NSWLR 502, 513A: “In considering the content of the apprehended bias principle the court must look to, inter alia, two things: what are the norms or standards relevant to the kind of case before it; and whether, on the facts, the requirements have been fulfilled.”

Sir Thomas Bingham MR also said this, at p 355:-

“In a case such as this, in which interlocutory applications proliferate, it may well be that one side fares more successfully, perhaps much more successfully, than the other. There are a number of possible explanations for this, the most obvious being that the successful party has shown greater judgment, determination and knowledge of the rules than its opponent. Mr Ross-Munro accepted, as we understood, that no inference of apparent bias could be drawn from the fact that most, or all interlocutory applications had been decided against Dr Hashim. We agree. He also disclaimed any attack on the correctness of Chadwick J’s interlocutory decisions. This we find puzzling. It must, we think, be hard to show consistent unfairness in the absence of consistent error.”

  1. A few years later in Locabail (UK) Ltd v Bayfield [2000] QB 451, Lord Bingham of Cornhill (now CJ) giving the judgment of the court which comprised himself, Lord Woolf MR and Sir Richard Scott V-C said (para 25):-

“… a real danger of bias might well be thought to arise if there were personal friendship or animosity between the judge and any member of the public involved in the case; or if the judge were closely acquainted with any member of the public involved in the case, particularly if the credibility of that individual could be significant in the decision of the case; or if, in a case where the credibility of any individual were an issue to be decided by the judge, he had in a previous case rejected the evidence of that person in such outspoken terms as to throw doubt on his ability to approach such person’s evidence with an open mind on any later occasion; or if on any question at issue in the proceedings before him the judge had expressed views, particularly in the course of the hearing, in such extreme and unbalanced terms as to throw doubt on his ability to try the issue with an objective judicial mind (see Vakauta v Kelly (1989) 167 C.L.R. 569); or if, for any reason, there were real grounds for doubting the ability of the judge to ignore extraneous considerations, prejudices and predilections and bring an objective judgment to bear on the issues before him. The mere fact that a judge, earlier in the same case or in a previous case, had commented adversely on a party or witness, or found the evidence of a party or witness to be unreliable, would not without more found a sustainable objection. In most cases, we think, the answer, one way or the other, will be obvious. But if in any case there is real ground for doubt, that doubt should be resolved in favour of recusal. We repeat: every application must be decided on the facts and circumstances of the individual case.”

  1. In Summers v Fairclough Homes Ltd [2012] 1 WLR 2004, the claimant had an accident at work which injured his heel. He obtained judgment on liability and claimed £800,000 on the basis that he was grossly disabled, largely dependent on crutches and in constant pain. The defendants secured surveillance evidence showing that he was not dependent on crutches, was in fact working, and was able to play football. The defendants sought to strike out the claim in its entirety as supported by false evidence. The Supreme Court decided that it would have to be an exceptional case to justify striking out the claim rather than giving judgment for the proper sum, one effective sanction for exaggeration of a claim being an application to commit for contempt. The defendants had indicated some reluctance to proceed for contempt but Lord Clarke of Stone-cum-Ebony giving the judgment of the court said (para 59):-

“… We, however, see no difficulty in proceedings by way of contempt in such cases, provided of course that the relevant facts can be proved. It was submitted in the course of argument that there might be difficulties in inviting the trial judge to hear applications for permission to bring proceedings for contempt. However, in the absence of special circumstances, we cannot see any difficulty in the trial judge hearing both the application for permission and, if permission is granted, the proceedings themselves. On the contrary, it seems to us that the trial judge is likely to be best placed to hear both. Such an approach is likely to be both the most economical and the most just way to proceed. The only circumstances in which that would not be the case would be where there was apparent bias on the part of the judge.”

  1. Lastly there is the similar (though converse) case of JSC BTA Bank v Ablyazov [2013] 1 WLR 1845 in which Teare J had made findings adverse to Mr Ablyazov in interlocutory proceedings holding him to be in contempt by failing to comply with a freezing order, dealing in frozen assets and lying while being cross-examined about his assets. At a later stage Mr Ablyazov applied to the judge to recuse himself from continuing with the case and trying the actions. The judge refused to recuse himself and was upheld by this court. Rix LJ (with whom Toulson and Maurice Kay LJJ agreed) asked whether there could be any difference between the judge who bears in mind his own findings and observations and another (second) judge who reads what the first judge has written, as he must be entitled to do. He answered his own question by saying (para 69):-

“… unless the first judge has shown by some judicial error, such as the use of intemperate, let me say unjudicial, language, or some misjudgement which might set up a complaint of the appearance of bias, the fair-minded and informed observer is unlikely to think that the first judge is in any different position from the second judge – other than that he is more experienced in the litigation.

70. In this connection, it seems to me that the critical consideration is that what the first judge does he does as part and parcel of his judicial assessment of the litigation before him: he is not “pre-judging” by reference to extraneous matters or predilections or preferences. He is not even bringing to this litigation matters from another case (as may properly occur in the situation discussed in Ex Parte Lewin; In re Ward [1964] NSWR 446, approved in Livesey v New South Wales Bar Association 151 CLR 288). He is judging the matter before him, as he is required by his office to do. If he does so fairly and judicially, I do not see that the fair-minded and informed observer would consider that there was any possibility of bias. I refer to the helpful concept of a judge being “influenced for or against one or other party for reasons extraneous to the legal or factual merits of the case”: see Secretary of State for the Home Department v AF (No. 2) [2008] 1 WLR 2528, para 53. I have also found assistance in this context in Lord Bingham’s concept of the “objective judgment”. The judge has been at all times bringing his objective judgment to bear on the material in this case, and he will continue to do so. Any other judge would have to do so, on the same material, which would necessarily include this judge’s own judgments.”

  1. There is thus a consistent body of authority to the effect that bias is not to be imputed to a judge by reason of his previous rulings or decisions in the same case (in which a party has participated and been heard) unless it can be shown he is likely to reach his decision “by reference to extraneous matters or predilections or preferences”. There can be no suggestion that Eder J would proceed in the present case by reference to such matters.


The Court of Appeal had no difficulty in overturning the decision of the judge.

  1. The judge appears to have thought that the charges of “actual bias” by Mr Urumov made all the difference because the allegations were “so serious” (para 17) that he ought to recuse himself. But can the mere elevation of the allegation from imputed bias to actual bias make a critical difference? I cannot think that it does. Of course such an allegation is an extremely serious one; it should not be lightly made. But the mere fact that a litigant decides to raise the stakes in that way cannot give rise to any difference of legal principle.
  2. Eder J gave 3 reasons for his decision. The first reason was that it was consistent with previous authority. I cannot agree; the above analysis shows that bias is not to be held to exist unless there is some reason to suppose that the judge would not bring an objective mind to bear on the case. If Mr Urumov wishes to rely on different arguments or different evidence from evidence previously adduced, no doubt the judge will take such arguments or evidence into account. There is no reason to suppose he will be influenced by extraneous matters or predilections or preferences. In one sense the present case is even less promising for Mr Urumov than it was for Mr Ablyazov because Mr Urumov can rely on the principle that a charge of contempt must be proved to a criminal standard. Mr Ablyazov had already been found to be in contempt to the criminal standard of proof and yet the judge was held to have rightly not recused himself from a trial where only the civil standard of proof applied.
  3. Secondly Eder J applied the observation in Locabail that, if there is any real ground for doubt, that doubt should be resolved in favour of recusal. But he does not explain what the real ground for doubt is in this case. The judge specifically said (in para 17 and also in para 13 of the judgment giving permission to appeal) that the allegations of bias are “groundless” and “spurious”.
  4. The third reason given by the judge is that the matter could be dealt with by another judge of the Commercial Court. No doubt it could be but that cannot in itself be a good reason for recusal any more than it could be a good reason not to recuse himself (in a proper case) that another Commercial judge could not be made available.
  5. The judge appears not to have been referred to the remarks of Chadwick LJ in this court in Triodos Bank N.V. v Dobbs [2001] EWCA Civ 468; [2006] C.P. Rep 1 in which Mr Dobbs invited the court to recuse itself and (more particularly) Chadwick LJ to recuse himself, as a result of his conduct in relation to a permission to appeal application in related proceedings. Chadwick LJ, giving the judgment of the court of which Neuberger LJ and I were members, said this:-

“7. It is always tempting for a judge against whom criticisms are made to say that he would prefer not to hear further proceedings in which the critic is involved. It is tempting to take that course because the judge will know that the critic is likely to go away with a sense of grievance if the decision goes against him. Rightly or wrongly, a litigant who does not have confidence in the judge who hears his case will feel that, if he loses, he has in some way been discriminated against. But it is important for a judge to resist the temptation to recuse himself simply because it would be more comfortable to do so. The reason is this. If the judges were to recuse themselves whenever a litigant – whether it be a represented litigant or a litigant in person – criticised them (which sometimes happens not infrequently) we would soon reach the position in which litigants were able to select judges to hear their cases simply by criticising all the judges that they did not want to hear their cases. It would be easy for a litigant to produce a situation in which a judge felt obliged to recuse himself simply because he had been criticised – whether that criticism was justified or not. That would apply, not only to the individual judge, but to all judges in this court; if the criticism is indeed that there is no judge of this court who can give Mr Dobbs a fair hearing because he is criticising the system generally, Mr Dobbs’ appeal could never be heard.

8. In the circumstances of this case, I have considered carefully whether I should recuse myself. Mr Dobbs has not advanced this morning any reason why I should approach his appeal with a disposition to decide against him; other than that he tells me that he is criticising me in relation to past conduct. That, I am afraid, is not a good reason for me to recuse myself. I do not do so. The other members of the court, who are within the rather wider ambit of Mr Dobbs’ application take the same view.”

If the judge had been referred to these remarks (reiterated by this court in Ansar v Lloyds TSB Bank Plc [2007] IRLR 211, para 17) he might very well have decided he ought not to recuse himself.

  1. Another matter which concerned the judge was that a few days before he gave judgment Andrew Smith J decided to recuse himself in the complex and long-running case of Dar Al Arkan Real Estate Development Company v Al Refai [2014] EWHC 1055 (Comm) from hearing a committal application after he had ruled in favour of the applicant/defendant that without notice orders made against them should be discharged because the claimants had misled the court and failed to comply with undertakings given to the court in the court’s orders. Once the defendants had obtained this ruling, they applied for an order that the managing director of the first claimant be committed to prison. The judge first decided that that application should be heard before any trial and then turned to an application by the claimants that he should recuse himself from hearing the contempt application. On the discharge application the judge had rejected the managing director’s account of how documents came to be deleted from certain hard drives which were to be preserved and delivered to the claimant’s solicitors as to which no full and honest explanation had been given. He accepted (para 36) that the views he had formed had been formed on issues (quite possibly on all the issues) likely to be crucial on the committal application and that the evidence on the committal application was likely to be essentially the same as the evidence he had heard on the discharge application. In the light of these considerations he considered the claimants were entitled to have another judge to hear the contempt application.
  2. It is thus clear that in Dar’s case the judge felt that the informed observer could not have the necessary confidence in the proceedings when the judge had already considered the essential evidence that would be deployed on the committal application and had come to the conclusion that the witnesses giving it were lying to him. A recusal application is a very personal matter for the judge to decide and this court will seldom interfere with this delicate jurisdiction. The overall feeling I have from reading Andrew Smith J’s judgment is that he himself felt uncomfortable about reconsidering essentially the same evidence on the very same issue which he would have to decide in the contempt application.
  3. The present case is different for two quite separate reasons. First as trial judge Eder J has considered an enormous number of issues and sub-issues between the parties. He has not focussed solely or mainly on the very issue that the judge will have to decide on the contempt application; the canvas of his judgment is infinitely broader. The focus of the contempt application will be much narrower and it is (in theory) quite possible that that narrow focus, coupled with the higher standard of proof required for the application, could produce a different result even if the evidence is essentially the same. In any event, the evidence may not be essentially the same.
  4. Secondly, it is clear that Eder J feels no personal embarrassment or discomfort in considering the contempt application. Not only has he not said anything to indicate such embarrassment or discomfort; he has positively said that the fears expressed by Mr Urumov are groundless and that he would welcome his decision (which he reached with “extreme reluctance”) being overturned. Since the reasons he gave for his decision are, in my view, defective, overturned it should be.
  5. Usually this court will be astute to support judges exercising what I have called “this delicate jurisdiction” of recusal. But it is also important that judges do not recuse themselves too readily in long and complex cases otherwise the convenience of having a single judge in charge of both the procedural and substantial parts of the case will be seriously undermined. Of course, if the judge himself feels embarrassed to continue, he should not do so; if he does not so feel, he should.
  6. This principle may be less important in less complex cases. Our attention was drawn to Re K [2014] EWCA Civ 905 handed down a fortnight or so before the hearing of this appeal. In that case the trial judge had made it plain to a recalcitrant father that, if he did not take action against the child’s grandparents in Singapore for the return of the child to the UK, he would be likely to be imprisoned for a lengthy term. He took no such action and she declined to recuse herself from the subsequent committal hearing. Kitchin LJ (with whom McFarlane and Maurice Kay LJJ agreed) criticised the judge’s very short judgment on the basis that she did not make clear that, despite her earlier observations and comments, she had not pre-judged the question whether the father was in deliberate breach of her orders and should be sentenced to a substantial term of imprisonment. This court held that there was therefore an appearance of bias or, at any rate, pre-judgment. By contrast, in the present case Eder J has given no indication of pre-judgment let alone that he has in mind any (let alone a lengthy) term of imprisonment.
  7. For these reasons I do not consider the judge was right to have recused himself, even if he was right to hold that the allegations made by Mr Urumov were allegations of actual bias. I would therefore allow this appeal.