THE preservation of confidence in the integrity of official decision-making demands that, as Lord Hewart C.J. famously put it in Ex parte McCarthy [1924] 1 K.B. 256, “justice should not only be done, but should manifestly and undoubtedly be seen to be done”. A rule against apparent, not just actual, bias is therefore necessary. As such, the rule can operate, according to Porter v Magill [2001] UKHL 67, [2002] 2 A.C. 357, on the basis of a real possibility (as opposed to probability) of bias, the existence of which falls to be judged by reference not to the court's perceptions, but to those of the “fair-minded and informed observer”. That formulation, however, begs more questions than it answers. What are the notional observer's characteristics? Exactly how “informed” is she? And (therefore) to what extent may her perception diverge from that of the court? These questions, which have vexed the courts since Porter was decided over a decade ago, were recently confronted by the Judicial Committee of the Privy Council in Belize Bank Ltd. v Attorney General of Belize [2011] UKPC 36.

The case centred upon a US$20m sum transferred by Venezuela to Belize. Belize used half of this to repay a government-guaranteed loan that had been advanced to a third party by Belize Bank Ltd (“BBL”). However, following a general election in 2008, the new Prime Minister of Belize characterised this as an “improper diversion” of funds, on the ground that the whole of the US$20m sum had been intended to support the provision of a housing and sporting complex. He vowed to “bring to account those who have robbed the people of this country”, accusing members of the previous government of “outrageous” and “highly immoral” behaviour that was “the product of a conspiracy”. Following a statutory investigation by the Central Bank of Belize, a directive was issued requiring BBL to transfer the contested US$10m to the government. BBL exercised its statutory right of appeal. In line with the relevant legislation, the board responsible for hearing the appeal was chaired by a judge appointed by the Chief Justice; its other two members, both of whom had worked at the Central Bank, were appointed by the Minister for Finance, which office was held by the Prime Minister. BBL argued that the board was tainted by the appearance of bias owing to the combination of the strong views previously expressed by the Prime Minister and his subsequent involvement, qua Finance Minister, in the appointment of board members. By a majority, the Privy Council rejected that argument. Of particular interest is the fact that the disagreement between Lord Brown, dissenting, and the majority turned specifically upon the characteristics with which the fair-minded and informed observer should—and should not—be invested.

Lord Kerr put his finger on the central difficulty when he acknowledged the danger of characterising the observer as “someone who, by dint of his engagement in the system that has generated the challenge, has acquired something of an insider's status”. Doing so would inevitably frustrate the underlying policy of the law, the aim being to preserve public confidence by disqualifying decision-makers who would be regarded with suspicion by an ordinary person. Yet, Lord Kerr—along with the other members of the majority in Belize Bank, and in common with several other post-Porter decisions, of which Taylor v Lawrence [2002] EWCA Civ 90, [2003] Q.B. 528 is doubtless the most striking example—arguably fell into precisely the trap that he himself had identified.

Giving the leading majority judgment, he held that the observer, once invested with knowledge of pertinent aspects of the appeal process, would perceive no real possibility of bias. It did not matter, said Lord Kerr, that some of that information would not be “immediately in the public domain”, for the observer “must have such information as may be necessary for an informed member of the public without any particular, specialised knowledge or experience to make a dispassionate judgment”. Yet while the dividing line between “particular, specialised knowledge” and matters of which an “informed member of the public” would be aware is crucial to the operation of the test, it is also highly malleable. For his part, Lord Kerr thought that much fell to be placed on the latter side of the line. Thus the observer was to be taken to know: of the “general structure of the system of appeal from the Central Bank's directive”; that statute provided for ministerial involvement in appointments; that the two members in question had in fact been appointed on the recommendation of a senior civil servant; of the limited size of the pool of credible candidates; that appointees swear an oath to act independently and impartially; that one of the three members of the board is appointed by the Chief Justice without ministerial involvement; and that while the board can take decisions by a majority, the majority must always include the judicially-appointed member. That an observer as well-informed as this would be likely to have confidence in the impartiality of the panel is perhaps unsurprising. But to imbue the notional observer with information of this nature and extent is arguably to render her an “insider”, thereby divorcing the operation of the test from the policy that supposedly animates it. The risk—to which the majority was apparently only abstractly alive—is that the observer is cast in the image of the reviewing judge, such that their perspectives converge in a way that marginalises the viewpoint of the ordinary, inevitably less well-informed, member of the public.

Lord Brown acknowledged that risk more concretely in his dissent (with which Lord Phillips had “initially” been “in sympathy”). In doing so, he cited with approval from a lecture given by the late Lord Rodger, who had observed that the “whole point of inventing this fictional character [viz. the observer] is that he or she does not share the viewpoint of a judge”. Lord Rodger went on to caution against judges ascribing artificially extensive knowledge to the observer, thereby distancing the observer from the ordinary person “who does not know these things”. Against this background, Lord Brown concluded that the fair-minded and informed observer—the “ordinary person in Queen Square Market” as the Belizean Chief Justice had put it in the proceedings below—would perceive a real possibility of bias. This conclusion was apparently shaped not only by Lord Brown's willingness to heed Lord Rodger's “salutary warning”—which led Lord Brown to approach matters from the perspective of a person with a “reasonable grasp of how things are usually done”—but also by the attitude that fell to be imputed to the observer. It is evident from their respective judgments that Lord Brown's observer was simply more prone to be suspicious than Lord Kerr's—a divergence that highlights a further problem inherent in applying the test: that a mind-set, as well as a dataset, falls to be ascribed to the notional observer.

Lord Hewart's axiom that justice “must be seen to be done” is as relevant today as ever. The difficulty lies in fashioning a test capable of realising that objective. One option (which the House of Lords was taken to have advocated in R. v Gough [1993] A.C. 646) would be for the reviewing court to determine whether there is in fact a real possibility of bias, and to hope that public trust in the independence and objectivity of the reviewing court would secure confidence in the wider apparatus of bureaucratic and judicial decision-making. On this analysis, reviewing courts would stand as guarantors of the impartiality of the system as a whole. Yet the public's willingness to place such trust in reviewing judges would likely be limited when other judges' impartiality were in question, and would likely be eroded if reviewing courts were to uphold decisions that appeared to the (less well-informed) public to be tainted by bias. It is unsurprising, therefore, that conventional wisdom holds that public confidence is best secured by adopting a test that, in this sphere, synchronises judicial decision-making with public perception—and that this, in turn, is best achieved by requiring the reviewing court to approach matters through the eyes of the ordinary person. To the extent that it gave life in English law to the reviewing-court-as-guarantor model, Gough is now seen as a wrong-turning. Yet the majority's analysis in Belize Bank comes close to reinstating such an approach in practice, since the fair-minded observer assumes a degree of judge-like omniscience that results in substantial misalignment of the court's evaluation and the perception of the “ordinary person in Queen Square Market”. This is not to suggest that courts should construct the observer in the image of an ill-informed, unthinking cynic: but if legal doctrine in this sphere is to reflect the policy that underpins it, reviewing judges should certainly make greater efforts to avoid (as Lord Rodger put it extra-judicially) “holding up a mirror” to themselves.