Pepa v Canada (Citizenship and Immigration), 2023 FCA 102
Last week, the Supreme Court granted leave to appeal Pepa v Canada (Citizenship and Immigration), 2023 FCA 102. This came as a surprise, because the FCA decision is so short and punchy. Nonetheless, the Court seemed interested in certain aspects of the case, substantively and methodologically:
Whether Parliament intended for a permanent resident visa holder to lose right to appeal removal order based on visa expiring after entering Canada due to administrative delay outside their control — Whether it is unreasonable for an administrative decision-maker to find itself bound by obiter statements in an earlier court decision when interpreting a statutory provision, rather than ascertaining the meaning of the provision itself in accordance with the modern principles of statutory interpretation that are indicative of Parliament’s intent — Immigration and Refugee Protection Act, S.C. 2001, c. 27, s. 63(2).
Pepa (aged 20) came to Canada with a permanent residence visa as an accompanying dependent child of her father. After the visa was issued, Pepa got married. When she arrived to Canada, she disclosed her marriage to an officer. Later on, the Immigration Division of the Immigration and Refugee Board [ID] issued an exclusion order for misrepresentation, but the preparation and issuing of the order was infected by administrative delay—in this period, Pepa’s visa expired. Pepa appealed to the Immigration Appeal Division [IAD] but the IAD concluded that she lacked the right to appeal because her visa had expired prior to the issuing of the exclusion order. The IAD relied on previous Federal Court jurisprudence (see e.g. Ismail v Canada (Citizenship and Immigration), 2015 FC 338) holding (purportedly) that the expiration of a visa extinguishes a right of appeal to the IAD.
The substantive issue in this case is whether the visa expiry legally extinguishes the right of appeal. Complicating matters was a claim, in the Federal Court, that administrative delay leading to abuse of process was the primary driver of the expiration of her visa. The IAD rejected this argument (see Federal Court Reasons, at para 45). Perhaps building on the principles enunciated in its recent Law Society of Saskatchewan v Abrametz, 2022 SCC 29, the Court may be interested in whether and how principles of administrative delay interact with the IAD’s appeal jurisdiction. And I think this is a live issue for the fairness of Canada’s immigration system, as well as other Canadian institutions, that also suffer from delay & inefficiency.
The Court also appears—worringly, for those of us who think administrative law is best when the Court does not touch it too often—interested in the methodology of reasonableness review. The Federal Court of Appeal concluded that there was nothing wrong with the IAD relying on Ismail and other IAD cases for its interpretation of the provisions, rather than conducting its own review of the text, context, and purpose [17]. This calls to mind Vavilov’s constraints pertaining to prior judicial decisions (see Issue #42). The question appears to be whether obiter statements in a previous judicial decision are binding—or should be at least considered—by a subsequent administrative decision-maker, potentially obviating the need to interpret the provision anew.
Readers will recall that courts in the post-Vavilov world have have occasion to address the relevance of prior judicial precedents: see Canada (Attorney General) v National Police Federation, 2022 FCA 80; Canada (Attorney General) v Hull, 2022 FCA 82, in Issue #42. By my lights, the general framework in these cases is as follows:
(1) A judicial precedent is not dispositive. An administrator can depart from precedent so long as it has regard to that precedent, and so long as they properly distinguish it.
(2) Does the precedent actually apply? Is it squarely on point? If it is, then an administrator will have to offer particularly cogent reasons to overcome it.
(3) If the precedent does apply, and the administrator properly concludes that it does, then the question is whether the administrative reasoning is reasonable under the normal standard of review.
In my perfect world, I’d prefer if judicial precedents on point completely constrained administrative choice. But this is not our law, especially under a presumptive reasonableness standard. Instead, and despite its imperfections, I believe this framework can generally encompass the situation raised in Pepa. It should not be the case that administrators must evaluate statutory provisions anew in light of a judicial precedent. Rather, the first question is whether the judicial precedent is sufficiently on point, in which case the burden of justification will be quite high: “The party could argue, first, that the precedent is simply inapplicable because while it appears to deal with an issue relevant in a current case, it does not, perhaps because it did not deal with a specific issue at hand in the current case….” (see Issue #44 discussing Keurig Canada Inc v Canada (Border Services Agency), 2022 FCA 100). This appears to be what the SCC is getting at when it asks: Whether it is unreasonable for an administrative decision-maker to find itself bound by obiter statements in an earlier court decision when interpreting a statutory provision. The question then becomes whether the aspects of Ismail relied on by the IAD (and by extension the FCA) were sufficiently binding to “oust” the need to interpret the statute anew. In her leave materials, Pepa offers several reasons why Ismail was inappropriate relied upon, which goes to question #2 in the framework.
While I cannot venture an opinion quite yet on the substance, it seems to me that we should avoid situations where a decision-maker must parse a judicial precedent, separating out the portions that are strictly binding and the parts that are “obiter” and therefore irrelevant. The framework I have extracted from the cases, above, obviates the need to do this. The administrator only needs to ask whether the decision is directly on point: does it interpret the same statutory provision, for example? Does it deal with the same issue under that provision?
This does create an imbalance. A judicial decision on point rendered before an administrative decision constrains the administrator; but if an administrator interprets a provision first, then the court must defer to it after-the-fact. This imbalance is arguably called for by Vavilov. At any rate, perhaps it can be fudged: we might say that, in either case, the judicial/administrative precedent is constraining but not binding the subsequent actor. Again, I think this should simply be a situation where judicial precedents are given pride of place. But this is not our law, and so we should try to make what we have workable.
Karl Wilson v Attorney General of New Brunswick, 2024 NBKB 27 (February 13, 2023)
Context: This is a referral to the Court under the Right to Information and Protection of Privacy Act. Wilson requested certain information from a Crown Attorney’s file respecting his case [4]. The request was denied by the Attorney General.
Issue: What is the standard of review?
Holding: As a result, I reject the suggestion that this Court’s role is to review the Attorney General’s refusal to disclose the requested information on a reasonableness standard. This Court will apply the law, which requires the Attorney General to prove on the balance of probabilities that the requested information and documents may be withheld, either because they fall within a category of documents which may not be disclosed, or within a category over which the Attorney General has a discretion to disclose or not disclose [22].
Analysis: Readers may get whiplash—just a few weeks ago, the Supreme Court held in Ontario (Attorney General) v Ontario (Information and Privacy Commissioner), 2024 SCC 4 (Issue #123), that the adjudication of freedom of information request implicating Cabinet confidence would be reviewed on a reasonableness standard. In this context, the Attorney General relied on this case to suggest that the reasonableness standard applies. But the Court, perhaps taking quite seriously Vavilov’s insistence on “institutional design choices,” rejects the invitation to apply the reasonableness standard. It distinguishes between a “referral” to the Court under the statute and a complaint to the Ombudsman [16-17]. The referral procedure defines the court’s role, clearly calling “on this Court to determine whether the head of the public body…has properly refused Mr. Wilson’s request…” [21]. Relying implicitly on Vavilov’s holding on rights of appeal—sort of—the Court says that “[t]he scheme adopted by the Legislature is totally inconsistent with the application of a reasonableness standard” [21].
I actually think this is right in this case, in part because the referral process is similar to a right of appeal in character. But it should be noted that the Court should have more clearly recognized that it was recognizing a legislative signal that rebutted the presumption of reasonableness. I take the point that the referral procedure is not a judicial review, but the presumption of reasonableness is broad in nature, and the Court should probably have done more to root its conclusion in Vavilov itself.
Disclaimer: Nothing in the SEAR is legal advice. The SEAR is not designed to be comprehensive. It is designed to collect some cases that I believe are interesting to practitioners and students, and to provide my quick thoughts. Not all of the cases presented are fully analyzed ie) I’ve selected the most relevant issues in some of the cases. Please complement the SEAR with your own research and study. Any mistakes are my own.