The Role of Appellate Courts in Judicial Review Cases I: The Burden (if any) that the Appellant Bears
In its recent decision in Canadian Coalition for Firearm Rights v. Canada (Attorney General), 2025 FCA 82, the Federal Court of Appeal returned to a theme it has mentioned more than once in recent years. What is the posture that an appellate court should take in relation to a first-instance judgment on a judicial review application?
The general rule is that the appellate court ‘steps into the shoes’ of the first-instance court and performs the analysis the reviewing judge ought to have performed. The exception to this general rule is that where the reviewing judge acts as a decision-maker of first instance, for example by making a finding of fact or exercising discretion, it is necessary to demonstrate a palpable and overriding error (see generally Northern Regional Health Authority v. Horrocks, 2021 SCC 42, [2021] 3 SCR 107, at paras. 10-12).
However, the Federal Court of Appeal has added a gloss to this general rule, in the form of a burden on appellants to demonstrate error on the part of the reviewing judge. In CCRA, at para. 25, de Montigny CJ explained:
as this Court has reiterated on a number of occasions, where the Federal Court’s reasons seem compelling, the appellants bear a tactical burden to show that these reasons in fact are flawed: see Bank of Montreal v. Canada (Attorney General), 2021 FCA 189 at para. 4, leave to appeal to SCC refused, 39899 (7 April 2022); Grewal v. Canada (Attorney General), 2022 FCA 114 at para. 11; Sun v. Canada (Attorney General), 2024 FCA 152 at para. 4; Kandasamy v. Canada (Attorney General), 2024 FCA 181 at para. 7; Power Workers’ Union v. Canada (Attorney General), 2024 FCA 182 at para. 181.
Consider, to begin with, the formulation in this paragraph.
It is interesting to review the statements in the prior cases. In Bank of Montreal (quoted in Power Workers’ Union), de Montigny CJ’s predecessor described matters as follows: “Where, as is the case here, the Federal Court appears to have given a complete answer to all the arguments that it advances, an appellant bears a strong tactical burden to show on appeal that the Federal Court’s reasoning is flawed”. In Grewal, the appellate role was said not to “ignore” the Federal Court’s reasons “when they answer the appellant’s contentions”. In Sun, appeals were said to be “essentially do-overs” but with a “tactical burden” (not a “strong” tactical burden) where the Federal Court’s reasons are “compelling and complete”. “Strong” reappeared in Kandasamy, the burden being triggered by “a thorough analysis [that] appears to have addressed all the arguments advanced by an appellant”.
The underlying idea here is clear enough: if an argument was not addressed by the first-instance court, the Federal Court of Appeal will address the argument de novo; otherwise, the appellant has to convince the Federal Court of Appeal that the first-instance judge erred.
There are two differences in the language used in these cases, however. First, sometimes the burden is described as “strong”, though quite what “strong” might mean in this context is not entirely clear. My view is that “strong” should be jettisoned as it does not appear to do any analytical work. It is much simpler and more straightforward to say that the onus is on the appellant to demonstrate an error in the Federal Court’s analysis.
Second, sometimes the burden is triggered by reasons that “seem” or “appear” to be compelling but has also been triggered by the reasons actually being compelling, the latter being a higher standard that, when invoked, also suggests that the Federal Court of Appeal is in agreement with the Federal Court’s reasons. The latter formulation should, in my view, be preferred. In all events, the Federal Court of Appeal should be telling the Federal Court (as well as lawyers and litigants) whether it was right or wrong in its application of legal principles. The use of “appears” or “seems” muddies the waters. It is similar to the Supreme Court of Canada’s use of the “substantially for the reasons of…” formulation in some criminal appeals. The use of such language means that it is unclear to what extent the decision of the lower court is good law for courts of coordinate jurisdiction (see Bogach, Opolsky and Veel, “The Supreme Court of Canada’s From-the-Bench Decisions” (2022) 106 SCLR (2d) 251). Ideally, then, having heard argument and deliberated, the appellate body should tell us whether the court below was right or wrong. If there is an aspect of the analysis below with which the appellate body is not entirely in agreement, this could be summarily signalled without providing chapter and verse.
In saying this, I have some sympathy for the Federal Court of Appeal, which hears judicial review appeals as of right. Unlike the Ontario or Quebec equivalents, which only hear appeals with leave, the Federal Court of Appeal cannot decline to entertain an appeal. No doubt, this means unmeritorious appeals are brought, perhaps even often brought. It might help to explain why the Federal Court of Appeal has developed the concept of a tactical burden. But I think the language could be tightened up.
Consider, now, the separate question of whether this approach is justifiable. I think it is — especially given the absence of a leave requirement — but that it could be applied variably, depending on the nature of the appeal. Issues in judicial review cases run along a spectrum, with purely legal issues at one end (think, maybe, of questions of general importance certified as such under the Immigration and Refugee Protection Act) and issues of pure fact at the other. On a legally suffused issue, it is not clear to me that any tactical burden should be imposed. Questions of law are treated de novo on appeal as the objective is to obtain the appellate court’s considered, authoritative view of legal issues. As the Supreme Court put it in Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 SCR 235, at para. 9, ” the primary role of appellate courts is to delineate and refine legal rules and ensure their universal application”, which requires “a broad scope of review with respect to matters of law”. Questions of principle relating to the methodology of reasonableness review, or the content of the record on judicial review, would be examples of the types of legally suffused issue upon which, in my view, a tactical burden would be less appropriate. In CCRA, notably, there is fairly extensive analysis of two grounds, one relating to the vires of regulations and another relating to sub-delegation; even though de Montigny CJ agreed with the trial judge, he nonetheless offered detailed analysis of these two important legal issues.
By contrast, appellate intervention on questions of mixed law and fact and questions of fact is much more sparing. In Housen, the Supreme Court approved the observations of the Supreme Court of the United States on this subject:
The trial judge’s major role is the determination of fact, and with experience in fulfilling that role comes expertise. Duplication of the trial judge’s efforts in the court of appeals would very likely contribute only negligibly to the accuracy of fact determination at a huge cost in diversion of judicial resources. In addition, the parties to a case on appeal have already been forced to concentrate their energies and resources on persuading the trial judge that their account of the facts is the correct one; requiring them to persuade three more judges at the appellate level is requiring too much. As the Court has stated in a different context, the trial on the merits should be “the ‘main event’ . . . rather than a ‘tryout on the road.’” . . . For these reasons, review of factual findings under the clearly-erroneous standard — with its deference to the trier of fact — is the rule, not the exception (Anderson v. Bessemer City, 470 U.S. 564 (1985), at pp. 574-75).
Where the application of the reasonableness standard is heavily influenced by the factual setting and surrounding context, the appropriateness of a tactical burden is at its greatest. Here, indeed, where a Federal Court judge has carefully considered all arguments and applied the applicable standard of review to a settled area of law without any obvious deviation from the required methodological approach, the likelihood of error is relatively low and, therefore, it is up to the appellant to demonstrate why the first-instance judge was wrong. In jurisdictions were leave to appeal is a requirement, these are cases in which leave is not likely to be granted.
Here, an analogy might be helpful. The UK Supreme Court has wrestled in a number of recent cases with the standard of appellate review in situations where a first-instance judge has opined on a question of proportionality. There is a useful synthesis, with helpful references, in the reasons of Lord Sales in Re JR123, [2025] UKSC 8 (at para. 41):
The type of measure in relation to which the question of proportionality may arise is very wide, ranging from individual action by a state official to provisions of general law enacted by the legislature. The factual contexts in which a question of proportionality may arise also vary widely. Where a first instance court has made an assessment of proportionality the question for an appellate court is whether that court’s assessment is wrong: R (Z) v Hackney London Borough Council [2020] UKSC 40; [2020] 1 WLR 4327, para 74; Safe Access Zones, para 33. This is a standard which is capable of being applied flexibly, depending on the nature of the measure and the circumstances of the case: Safe Access Zones, para 33, referring to Director of Public Prosecutions v Ziegler [2022] AC 408, paras 102-103 (Lady Arden) and paras 129-140 (Lord Sales). Sometimes – in particular when a one-off decision is in issue which only affects persons involved in the proceedings, there is no controversy about the content and Convention compatibility of the general law which is applicable and the case turns essentially on a factual assessment of the circumstances which the lower court was particularly well placed to make – it will be appropriate for the appellate court to adopt an approach according to which it asks whether the lower court directed itself correctly and has had due regard to relevant matters, without any need to second guess that court’s proportionality assessment if it has. But in other situations – in particular where matters of general principle are in issue or the question concerns the Convention compatibility and proportionality of general rules set out in legislation – it is the proper function of the appellate court to determine the question of proportionality for itself without deferring to the assessment made by the lower court, even if that court has directed itself correctly and its decision cannot be said to be unreasonable. Only by adopting this approach can the appellate court fulfil its function of providing general guidance on the law. For commentary, see P Daly, “Appellate Standard of Review in Public Law Cases” [2021] Public Law 334; P Sales, “Proportionality Review in Appellate Courts” (2021) 26 Judicial Review 40 (For more on the case, see Lewis Graham (here and here)).
Put very simply, legally-suffused issues are for appellate courts, one-off factually suffused matters for first-instance bodies. This is only an analogy, as the UK Supreme Court does not want appellate courts ‘stepping into the shoes’ of first-instance courts in the manner of their Canadian counterparts. In Canada, under the approach confirmed in Horrocks, it is not open to an appellate court to ‘defer’ to the reviewing judge: it must determine whether the reviewing judge’s analysis was correct. But a tactical burden in this regard — on matters that are relatively factual in nature — is justifiable, in my view.
In the end, I think the current approach of the Federal Court of Appeal is justifiable in principle, though I think the language could be tightened up and close attention paid to the relationship between the burden imposed and the nature of the question being appealed.
In a subsequent post, I will address recent debates in Ontario about the threshold for granting leave to appeal from a first-instance judicial review.
This content has been updated on June 4, 2025 at 14:55.