The Problem with the Modern Law of Bias

This is an extract from my recent paper on the law of bias (available here):

There is a problem with the common law of bias as applied to administrative decision-makers. Comprised of a set of heavily context-sensitive principles, it is difficult to predict the application of the law of bias in advance. The Privy Council noted how “[t]his is a corner of the law in which the context, and the particular circumstances, are of supreme importance”, which makes generalization hazardous.[1]  As Professor Groves observes in a magisterial study of the area, the “potential level of difference in judicial assessment of bias claims is inevitable”.[2]

The historical evolution of the common law explains why. Previously, the no-bias principle was one of the two rules of natural justice. These were nemo iudex in sua causa (no one shall be a judge in their own cause) and audi alteram partem (hear both sides).[3]

When the prerogative writs were used to control judicial decision-making, the courts originally required strict compliance by decision-makers with these two rules of natural justice. Gradually and incrementally, over a period of many years, the courts extended the scope of the prerogative writs to cover a broad range of administrative decision-making. With the growth of the administrative state in the wake of the Industrial Revolution, the extension of the franchise and crises of war and depression, it became necessary to adapt the strict rules designed to regulate judicial decision-making to an administrative context.

One consequence was that it became entirely unrealistic to require all decision-makers to respect the rules of natural justice to the letter. This was particularly obvious in respect of the hearing principle. Given that the modern administrative state was making decisions about matters ranging from housing benefits and immigration status to tariffs payable to providers of public services, to require full-on hearings for all manner of administrative decisions would be entirely impractical.

It also became increasingly obvious in respect of the no-bias principle. For example, administrative tribunals were, from their earliest days, peopled by members with expertise in a particular area and significant prior knowledge of issues likely to arise.[4] They were not detached, impassive arbiters wearing “the inscrutable face of a sphinx”.[5]

When states later created regulatory agencies to manage complex sectors of the economy, like stock markets and utilities the problem became more acute: such agencies have rule-making, rule-interpreting and rule-enforcing functions that are often exercised by the same people; those doing the interpretation and enforcement will often have been participants in the industry prior to becoming regulators; and regulators are expected to engage in dialogue with the regulated community and broader public. These roles, individually and in combination, are anything but judicial.

Administrative decision-makers are very far removed from the Sphinx-like judge, detached from society and applying law to fact dispassionately, for whom the nemo iudex rule was developed. Indeed, it should be said that even today’s judges are not expected to imitate their forebears by remaining aloof on the dais, rather the no-bias principle is relaxed even in its application to members of the judiciary in recognition of the desirability of judges engaging with civil society.[6]  Accordingly the rules developed for previous eras have morphed into context-sensitive standards appropriate for this one, generating however a lack of predictability in the application of the standards in any given case.

[1] Man O’War Station Ltd v Auckland City Council (No 1) [2002] UKPC 28; [2002] 3 NZLR 577 [11] (PC).

[2] “Clarity and Complexity in the Bias Rule” (2020) 44 Melbourne University Law Review 565, at p. 568.

[3] Contemporary understandings of the latter rule are now addressed under the label of the hearing principle, which is not the subject of this paper.

[4] Chantal Stebbings, “Comment: A Victorian Legal Legacy – the Bespoke Tribunal” (Council on Tribunals, Adjust, April 2007), p. 3.

[5]  R. v. Nat Bell Liquors, [1922] 2 AC 128, at p. 159. I appreciate that Lord Sumner was writing there of judicial review for error on the face of the record, but the ‘sphinx’ point has been picked up on in the bias context: Brouillard Also Known As Chatel v. The Queen, [1985] 1 SCR 39, at para. 17.

[6] Yukon Francophone School Board, Education Area #23 v. Yukon (Attorney General), 2015 SCC 25, [2015] 2 SCR 282, at para. 59.

This content has been updated on April 22, 2025 at 15:57.