Unappealing Applications for Judicial Review: Best Buy Canada Ltd. v. Canada (Border Services Agency), 2025 FCA 45

The Canadian debate about the relationship between statutory rights of appeal and applications for judicial review continues to rumble on. In Best Buy Canada Ltd. v. Canada (Border Services Agency), 2025 FCA 45, Stratas JA for a unanimous Court of Appeal dismissed an appeal and judicial review application that “adopts the submissions made in the appeal, nothing more” (at para. 17).

The major issue in the case was whether the Canadian International Trade Tribunal had committed an error of law by following a precedent of the Federal Court of Appeal that, in the applicant’s view, was wrongly decided. Stratas JA was not persuaded that there were sufficient reasons to depart from that precedent: “if Danby is to be reversed, the appellant should seek leave to the Supreme Court” (at para. 5). He then went on to make some important observations about the relationship between appeals and applications for judicial review.

Here, the right of appeal to the Federal Court of Appeal from the Tribunal is limited to questions of law (Customs Act, RSC 1985, c 1 (2nd Supp), s. 68(1)). In Yatar v. TD Insurance Meloche Monnex, 2024 SCC 8, the Supreme Court confirmed that a party may make an application for judicial review in respect of matters falling outside the scope of a limited right of appeal. But, as Stratas JA rightly points out, the fact that you may does not necessarily mean that you should (at para. 11).

That then raises the question of the distinction between an appeal on a question of “law” and an application for judicial review. This question has troubled courts around the common law world: I am most familiar with the Irish position, where judges have often wondered whether there is any distinction at all because whether an administrative decision-maker committed any reviewable error is, itself, a question of “law”. Stratas JA doubts the distinction as well. In his view, “just about anything” that can be raised on judicial review can be characterized as a potential error of law for the purposes of a limited right of appeal:

• Alleged legal errors by the administrative decision-maker, whether they be found in the Constitution, legislative provisions, common law principles or administrative law principles. This includes questions of law that are extricable from (i.e., taint or dominate) questions of mixed fact and law: Canadian National Railway Company v. Emerson Milling Inc., 2017 FCA 79, [2018] 2 F.C.R. 573; Canadian National Railway Company v. Canada (Transportation Agency), 2016 FCA 266 at para. 22; Neptune Wellness Solutions v. Canada (Border Services Agency), 2020 FCA 151 at para. 15.
• Procedural fairness concerns: Emerson Milling at paras. 18-19.
• Sufficiency of reasons or inadequate reasons on a key point: Halton (Regional Municipality) v. Canada (Transportation Agency), 2024 FCA 122 at paras. 21-33.
• Errors that seem factual but are really legal errors or failures to follow legal principles governing fact-finding. For example, a decision-maker that wrongly takes judicial notice (R. v. Spence, 2005 SCC 71, [2005] 3 S.C.R. 458), wrongly finds facts without any supporting evidence (Canada (Border Services Agency) v. Danson Décor Inc., 2022 FCA 205 at para. 14), wrongly draws a factual inference or finds facts contrary to the law of evidence (e.g., Pfizer Canada Inc. v. Teva Canada Limited, 2016 FCA 161 and the cases cited therein; Garvey v. Canada (Attorney General), 2018 FCA 118 at para. 6), or wrongly finds facts contrary to a statutory provision (Walls v. Canada (Attorney General), 2022 FCA 47 at para. 41; Page v. Canada (Attorney General), 2023 FCA 169 at para. 79).

I agree entirely with the first two bullet points. On the latter two, there is more to say.

As far as factual errors are concerned, to the list of cases cited by Stratas JA can be added the scholarly analysis of Carnwath LJ (as he then was) in E v. Secretary of State for the Home Department, [2004] QB 1044. For Carnwath LJ, an error of fact can be considered an error of law where four conditions are met:

First, there must have been a mistake as to an existing fact, including a mistake as to the availability of evidence on a particular matter. Secondly, the fact or evidence must have been “established”, in the sense that it was uncontentious and objectively verifiable. Thirdly, the appellant (or his advisers) must not been have been responsible for the mistake. Fourthly, the mistake must have played a material (not necessarily decisive) part in the Tribunal’s reasoning (at para. 66).

My concern here, however, is that treating errors of fact as errors of law is difficult to square with the key objectives of the Vavilov framework: simplicity and clarity. I am not sure it is either simple or clear to say that, sometimes, facts are law.

As far as sufficiency of reasons is concerned, I think there is great merit to Stratas JA’s proposed approach. However, there is another step that would have to be taken to achieve conceptual harmony. The sufficiency of reasons for the purposes of a statutory appeal (whether limited to a question of law or not) is assessed under the framework set out in R. v. Sheppard, 2002 SCC 26, [2002] 1 SCR 869. There may be “some overlap'” between this framework and the Vavilov framework for judicial review (Halton, at para. 22) and, indeed, I have suggested in the past that Vavilov might cause the frameworks to merge. For now, the two frameworks remain distinct (albeit that the practice of identical appeals on questions of “law” and applications for judicial review gives me some pause in this regard).

That said, using Vavilov’s guidance on reasons in administrative appeals makes good sense to me. From an analytical perspective, why should an applicant  be confined to the restrictive Sheppard standard, which is designed for the relationship between first-instance and appellate courts? And from an institutional design perspective, doesn’t granting a right of “appeal” suggest more intrusive judicial oversight, including of reasons?

But I think the Supreme Court would have to make this change — perhaps prompted by similar analyses from other appellate courts. Vavilov reasserted a categorical distinction between appeals and judicial review that had long since fallen by the wayside in the Dunsmuir era. Pre-Vavilov, to almost all intents and purposes, there was no meaningful difference between the grounds that could be considered on appeal or on judicial review. But Vavilov sets out two sets of rules, one for appeals and one for judicial reviews.

Now, an alternative course would have been to say “a right of appeal brings correctness on extricable questions of law” but that otherwise reasonableness review applies. That road was, however, not taken in Vavilov. As I say, I think there are good reasons for the Supreme Court to switch from Sheppard to Vavilov on appeals from administrative decisions, where the adequacy of reasons is in issue. But until it does so, there is going to be a gap between appeals and judicial reviews.

If the Supreme Court did so, the result would be that only pure questions of fact could conceivably fall outside an appeal clause limited to questions of “law”. That would likely drastically reduce the volume of concurrent applications for judicial review and statutory appeals on questions of law, if only because of the difficulty (even on Vavilov) of challenging pure findings of fact on judicial review.

In the end, then, there are sound principled and practical reasons for carefully engaging with the analysis in Best Buy 2025: this debate will continue to rumble on.

This content has been updated on March 6, 2025 at 20:49.