The law on the two kinds of disqualification for judicial bias is recovering from a wrong turn taken in the Pinochet case. On 23 March 2005 the Judicial Committee of the Privy Council, in Meerabux v AG of Belize, had to decide on an application for judicial review by George Meerabux of the decision to remove him as a justice of the Supreme Court. Judicial bias is often alleged, at least, suspected, by inexperienced litigants who feel things are not going way. It is also often suspected of planning authorities, where arcane workings, and social networks linking council officers, councillors and developers trouble objectors keen to protect their neighbourhoods. It seems to be human nature to believe that a decision-maker who decided against one's interests cannot have acted fairly. Practitioners will head off the great majority of suggestions of bias made by litigants before they see the light of day. However, in a few cases, decision-makers and parties need to know whether there is case for the decision-maker to recuse himself. The Meerabux is one of a line of cases which provide guidance on the problem.

The press in Belize saw the sacking of Justice George Meerabux as part of a clean-up of an intolerably slow and corrupt judiciary. If a judge who is removed from office is likely to raise allegations of bias, this is particularly so in place like Belize. The judge was removed by the Governor General after an investigation. There are three justices of the Supreme Court, which handles all criminal and civil trial work. There is one Belizean judge on the Court of Appeal (and four possible visiting judges). First, Mr Meerabux sought declarations on various points amounting to a declaration that he had been deprived of a fair hearing. He objected that the chair of tribunal which made findings against him, and recommended his dismissal to the Governor General, was a member of the same Bar Association which had initiated the complaint against him. (He also objected that the tribunal sat in camera: I do not discuss this complaint further; it did not succeed at any stage). One of his two brothers on the Supreme Court bench declined to makes the declarations he sought (judgment). Then he sought a stay of the Governor General’s action against him from another justice of the Supreme Court (which was refused). Then The Governor General removed him from office. Then he applied again for declarations that he been treated unfairly, and a third judge dismissed that application. That matter went to the Court of Appeal (judgment), and up to the Privy Council. Mr. Meerabux failed in every court.

The law was thrown into mild confusion by the Pinochet drama in the House of Lords. Readers will remember the unfortunate way in which the case developed. Pinochet applied for judicial review of an arrest warrant issued in support of an extradition request. He failed in the House of Lords in a judgment now usually known as Pinochet (No. 1); the House of Lords set aside its first decision in what usually known as Pinochet (No. 2) on the ground that Lord Hoffman had held office in a charity associated with Amnesty International. It is not every day that English lawyers have to argue that where a judge has been formally associated with an organization which opposes torture, their client cannot expect a fair trial. The Senator, of course, failed again in when his appeal was reheard in Pinochet (No. 3).

The confusion in Pinochet (No. 2) arose when their Lordships founded their decision on a proposition which counsel had not argued or advanced (to judge by the report at [2000] 1 AC 119). Counsel had made the correct arguments – which could then be formulated in two ways: either there was a real possibility of bias according to a fair-minded observer, or there was a possible appearance of bias. For what is worth, in R v Gough [1993] A.C. 646, a case about a possibly biased juror, the House said that the test was whether there was a real danger the defendant had not had a fair trial; after Pinochet (No. 2) the Court of Appeal said, in Re Medicaments, to quote from the headnote at [2001] 1 W.L.R. 700, ‘taking into account the approach of the European Court of Human Rights in accordance with the Human Rights Act 1998, the court had first to ascertain all the circumstances which had a bearing on the suggestion that the judge was biased and then ask whether those circumstances would lead a fair-minded and informed observer to conclude that there was a real possibility, or a real danger, the two being the same, that the judge was biased.’

However, their Lordships in Pinochet understandably did not want to accept that Lord Hoffman might have fallen foul of the Gough test. They said that there are two routes to disqualifying for possible bias. One is where there is a real danger of unfairness, which of course could not apply to Lord Hoffman. Then there is the automatic disqualification where the judge has a pecuniary interest. This kind of bias may be labelled nemo iudex in sua causa, or let no one be judge in their own cause. Their Lordships, at any rate Lord Browne Wilkinson expressly, and Lord Hutton almost expressly, and Lord Goff by clear implication, decided that Lord Hoffman should be subject to the automatic disqualification which applies to a judge sitting in his own cause, regardless of the absence of any appearance or danger of bias. They said that because of Lord Hoffman’s links with Amnesty, he had judged in his own cause. Thus they extended the concept of what is a judge's 'own cause' beyond pecuniary interests.

This idea - that there at two types of bias – is good law. However, it has become clear in retrospect that what the House of Lords did in Pinochet (No. 2) has tended to confuse the law. Matters would have been far simpler if they could had found a way to say that Lord Hoffman was disqualified for possible bias in the mind of some observer, or the appearance of bias, or something of that kind. The reason for this is that they opened a loophole. In every case where the danger of bias or appearance of bias test fails, clever counsel will argue, ‘Aha – so – no possible bias, but the decision maker has enough links with the cause for it to be his own cause. Ergo – automatic disqualification, even where bias is impossible. Why, the House of Lords have said so in Pinochet (No. 2)!’ And probably the most important point about the Privy Council’s judgment in Meerabux is the spectacle of the senior judiciary retreating from this extension of the concept of automatic disqualification. As noted, the argument for the appellant (Mr. Meerabux) was that he had been deprived of various constitutional rights to fair treatment because the tribunal which made findings against him was chaired by a member of the very same Bar Association which had brought the original complaint. He argued that the tribunal chair was sitting as a judge in his own cause. As well as insisting that the connection was not close enough, unlike Lord Hoffman’s (perhaps not terribly close) links with Amnesty, for this argument to succeed, the Committee took the opportunity to say, at paragraph 22, ‘As Lord Steyn said in Lawal v Northern Spirit Ltd [2003] UKHL 35, [2003] ICR 856, para 14, public perception of the possibility of unconscious bias is the key. If the House of Lords had felt able to apply this test in the Pinochet case, it is unlikely that it would have found it necessary to find a solution to the problem that it was presented with by applying the automatic disqualification rule.’

To sum up so far, there are two kinds of disqualification. Automatic disqualification, where the judge has a pecuniary interest (whether this distinct kind of disqualification applies to non-judicial decision makers is unclear). Then the disqualification where there may be a perception of possible, if unconscious bias. In Pinochet (No.2) the House of Lords extended the automatic kind of disqualification to include non-pecuniary links. That extended kind of automatic disqualification could probably now still happen in some cases, but I have not found any English or Commonwealth case where it has been applied, and Meerabux show the judiciary seeking to confine it as far as possible.

Does it matter in practice which kind of disqualification is at stake? Surely the unconscious bias test covers everything? Yes, it does matter. In the United States, the tests for recusal are set down in statute (at 28 USC 455 for the federal courts). There, in cases of automatic disqualification, unlike cases of possible bias, it is not possible for the parties to waive the requirement for the judge to stand down. (As an aside, recusal from the final court of appeal should be easier to obtain in most common law jurisdictions, because there is no possibility of further appeal; recusal is more difficult to obtain in the United States Supreme Court because recusal of a single justice has the disadvantage of creating a tribunal with an even number of members.)

However, in English law too the distinction between automatic disqualification and disqualification for possible bias has significant consequences. The distinction is discussed a length by a unanimous Court of Appeal in Locabail (UK) Ltd v Bayfield Properties Ltd and conjoined appeals ([2000] QB 451: the case does not appear to be on www.bailii.org). In cases of automatic disqualification, it is no good the judge saying he was unaware of his interest, or he forgot about it. His decision will be void, nevertheless. Furthermore, in an automatic disqualification case, the judge has a duty to recuse himself straight away, without any party mentioning the issue. If he fails to do so, and costs are wasted as a consequence, the judge will clearly be open to criticism. Further, it must at least be arguable that the Common law, like the statute law in the United States, is or should be that since an irrebutable presumption of bias arises in cases of automatic disqualification, no waiver is possible. Does a party have the right to elect a trial which the law deems irrebutably to be unfair? Maybe not.

Dr John Birchall - Professional English, Legal English, and Common Law Training