Bias on Multi-Member Panels: Vento Motorcycles, Inc. v. Mexico, 2025 ONCA 82

A question that has garnered relatively little attention in recent years is whether the bias of one member of an adjudicative body taints the entirety of the decision. The Supreme Court of Canada’s authorities point in two directions.

On the one hand, in Newfoundland Telephone Co. v. Newfoundland (Board of Commissioners of Public Utilities), [1992] 1 SCR 623, the decision of a multi-member regulatory board was set aside on the basis that one member had made public comments that raised a reasonable apprehension of bias (and, indeed, suggested that he had a closed mind). Cory J set aside the entirety of the decision on the basis that it had been voided by the presence of bias: “The damage created by apprehension of bias cannot be remedied” (at p. 645).

On the other hand, in Wewaykum Indian Band v. Canada, 2003 SCC 45, [2003] 2 SCR 259, the Supreme Court observed in obiter that it would be inappropriate to set one of its decisions aside by reason of the bias of one of its members (though there, on the facts, no bias had been made out). This was because of the collegial process of decision-making at the Supreme Court, which meant that the courts’ “reasons express the individual views of each and every judge who signs them, and the collective effort and opinion of them all” (at para. 92):

Here, the nine judges who sat on these appeals shared the same view as to the disposition of the appeals and the reasons for judgment. Cases where the tainted judge casts the deciding vote in a split decision are inapposite in this respect. In the circumstances of the present case, even if it were found that the involvement of a single judge gave rise to a reasonable apprehension of bias, no reasonable person informed of the decision-making process of the Court, and viewing it realistically, could conclude that it was likely that the eight other judges were biased, or somehow tainted, by the apprehended bias affecting the ninth judge (at para. 93).

In Vento Motorcycles, Inc. v. Mexico, 2025 ONCA 82, Huscroft JA engaged in a considered and comprehensive analysis of the question, coming down on the side of Newfoundland Telephone: the bias of one member of a multi-member body taints any decision made by that body.

Huscroft JA began with the proposition that an unfair decision will be set aside:

The common law has historically been strict in response to a breach of procedural rights. A failure to provide a fair hearing has resulted in the quashing of the substantive decision, regardless of the result that might otherwise have obtained. It has never been necessary for an applicant seeking relief to establish that the outcome of the relevant decision would – or even might – have been different but for the unfair hearing procedure (at para. 29, citing Cardinal v. Director of Kent Institution, [1985] 2 SCR 643, at p. 661).

He observed that this proposition is “stricter still” so far as bias is concerned (at para. 31, citing Canadian College of Business and Computers Inc. v. Ontario (Private Career Colleges), 2010 ONCA 856, at para. 64), rationalizing as follows:

This approach reinforces the seriousness of an apparent failure of impartiality. No one whose rights, interests, or privileges are at stake can be required to accept a decision made by an adjudicator whose ability to decide fairly is – for whatever reason – reasonably in doubt. The importance of the rule against bias transcends the interests of the parties to a particular dispute: bias is intolerable in any system that aspires to the rule of law. The finding of a reasonable apprehension of bias requires the disqualification of an adjudicator and the nullification of any decision they have made. Nothing less will do (at para. 32).

There are two distinct points here. One is that solicitude for the interests of the individual whose “rights, interests, or privileges are at stake” requires robust judicial intervention: if an adjudicator is not perceived as being able to decide fairly (as opposed to having made a procedural or substantive error at a hearing), the individual should never be subject to the exercise of the adjudicator’s authority. Here, the risk of arbitrariness is too great, even more so than in respect of other types of procedural or substantive error. It is one thing for the adjudicator to potentially get something wrong; it is quite another to be exposed to the exercise of power by someone who may consciously or unconsciously make a decision based on entirely extraneous considerations. In the Vento Motorcycles case, one member of a multi-member arbitration panel was negotiating a lucrative appointment to a national panel of arbitrators during the hearing of the arbitration between Vento Motorcycles and the same government that was making the promises. In those circumstances, we fear bias because the arbitrator might (even with the best of intentions) favour the interests of one of the parties out of a concern for future preferment even though this should never be a relevant consideration. This is the risk of arbitrariness and it is categorically different from the risk that an adjudicator will get something wrong in the course of performing an adjudicative function.

A second point is that this principle is systemic: bias is “intolerable” because the legitimacy of the system of adjudication depends upon decisions being made without the taint of bias. If adjudicators are perceived to be biased, no one will have confidence in the ability of adjudicators to resolve disputes dispassionately. This saps the legitimacy of the system and is entirely inimical to any notion of good administration. Who would have confidence in an arbitral system that permitted adjudicators to negotiate lucrative appointments with parties appearing before them? There is a good reason that Lord Hewart CJ’s admonition that justice should not only be done but be seen to be done continues to echo down the ages.

In short, the “stricter still” approach to bias has individual and systemic foundations: it ensures individuals are not subject to arbitrary exercises of authority and upholds public confidence in the system.

Note that in the High Court of Australia’s reformulation of its test of ‘materiality’, the judges agreed that a reasonable apprehension of bias will always be material, as it is “inherent in the nature of the error” that the resulting decision must be quashed (LPDT v Minister for Immigration, Citizenship, Migrant Services and Multicultural Affairs, [2024] HCA 12, at para. 6 (noted here)).

It is useful to consider the ‘materiality’ issue further. As I have noted, materiality has three different senses. First, an error can be material in the sense that it is sufficiently serious to justify judicial intervention. A reasonable apprehension of bias is certainly material in that sense. Second, an error can be material in the sense that there is a causal link between the error and the decision. Huscroft JA (and the High Court of Australia) has explained that a reasonable apprehension of bias will always be treated as material in this sense. Third, an error can be material in the sense that a judge, in the exercise of remedial discretion, might decide to decline to grant a remedy, perhaps on the basis that the error complained of was immaterial to the outcome.

This third type of materiality appears to be what the Supreme Court had in mind in the Wewayakum case. But the Supreme Court’s obiter was based on the unique circumstances of the Supreme Court’s decision-making processes and the Supreme Court’s own knowledge of those decision-making processes:

Ultimately, whilst this sort of remedial discretion can be invoked in principle, the circumstances in which it is likely to be appropriate to do so are extremely limited. Huscroft JA correctly noted that refusing to grant a remedy for bias is the exception rather than the rule:

It is important to reiterate that a reasonable apprehension of bias is no minor procedural breach. It is a finding that the integrity and legitimacy of an adjudicative process have been compromised irreparably. It cannot be balanced away on the basis that it is not serious; that it is thought to have had little impact on the result; or that it would be inconvenient and costly to rehear the arbitration if the award were set aside. A finding of a reasonable apprehension of bias means that it is objectively reasonable to conclude it is more likely than not that a dispute would not be decided fairly. It disqualifies an adjudicator in public law proceedings, and I see no reason why a different result should follow in the context of commercial arbitration (at para. 42).

The only qualification to make here is that an applicant will ‘waive’ a claim of bias if they do not raise it at the first available opportunity (at para. 33). This is, again, best understood as materiality in the third sense, as relating to judicial discretion to refuse a remedy, really treating a reasonable apprehension of bias as immaterial on the basis that it could not have been material to the outcome if the applicant was content to let the process unfold in the hope of getting a positive outcome. Even here, however, note that the systemic considerations invoked by Huscroft JA will sometimes lead a court to overlook a waiver: if the systemic implications are serious enough, a court will nonetheless entertain a bias claim that was not raised in a timely manner (see Fundy Linen Service Inc. v. Workplace Health, Safety and Compensation Commission, 2009 NBCA 13).

Lastly, it is worth quoting Huscroft JA’s discussion of the authorities underpinning the Supreme Court’s approach in Newfoundland Telephone, making the point with great force that this is a well-known, widely-accepted principle:

This principle, sometimes described as “poisoning the well”, was endorsed by Esson J.A. in Haight-Smith v. Kamloops School District No. 34 (1988), 1988 CanLII 2900 (BC CA), 51 D.L.R. (4th) 608 (B.C. C.A.), at p. 614, and by Rothstein J. (as he then was) in Sparvier v. Cowesses Indian Band (T.D.), 1993 CanLII 2958 (FC), [1993] 3 F.C. 142, at p. 166. Writing in 2001, David J. Mullan summarized the law as follows: “[a] reasonable apprehension of bias in one member of a tribunal is sufficient to disqualify the whole tribunal, even though that member merely sat at the hearing without taking an active role in either it or subsequent deliberations. Mere presence is generally enough”: Administrative Law (Toronto: Irwin Law, 2001), at p. 131.

This principle is also well established in English law, even where the finding of bias concerns a member of a judicial, as opposed to an arbitral, panel. See In re Medicaments and Related Classes of Goods (No 2), [2001] EWCA Civ 1217, [2001] 1 W.L.R. 700, at para. 99, endorsed by the Judicial Committee of the Privy Council in Stubbs v. The Queen, [2018] UKPC 30, [2019] A.C. 868, at para. 33. As that court explained, the bias of a single member necessarily vitiated a panel’s decision: “the whole point of the appeal was that three judges should consider the issues”, and “[t]he mutual influence of each member of the court over the others necessarily means that if any of them was affected by apparent bias the whole decision would have to be set aside” (at paras. 47-49).

Ultimately, then, the bias of one is the bias of all.

This content has been updated on July 9, 2025 at 16:34.