The Meaning of Reasonableness Review after Vavilov

This is an excerpt from my paper, “The Scope and Meaning of Reasonableness Review after Vavilov“

As is well known, the Supreme Court’s guidance on reasonableness review prior to Vavilov was extremely limited, a shortcoming that was emphatically addressed in the majority reasons.

The General Structure of Reasonableness Review

There are three general points to make about reasonableness review under this framework.

General Principles

First, in terms of methodology, the general principles are clearly laid out. As I have observed elsewhere, Vavilov functions almost as a ‘Civil Code’ of judicial review, with principles appearing first and detailed specifications of the principles set out in subsequent paragraphs.[1] In respect of reasonableness review, the majority judges identified the following general principles:

• The burden is on the applicant to show that a challenged decision is unreasonable;[2]
• Reasonableness review is “a robust form of review”;[3]
• Judicial review requires a reasons-first analysis where reasons are required to be given;[4]
• A reviewing court does not conduct a de novo analysis to identify what the decision ought to have been;[5]
• Reasonableness depends on sensitivity to legal and factual constraints;[6]
• Reasonableness requires an appreciation for institutional context and background;[7] and
• “‘[a]dministrative justice’ will not always look like ‘judicial justice’”.[8]

I do not propose to go through these principles in great detail here. The point for present purposes is that these principles can be called into action where the application of the reasonableness standard gives rise to debate.

Consider, for instance, the decision of the Supreme Court in the Mason case. One interesting feature of that case is the majority’s disagreement with the lower court about the approach to take to reasonableness review. This concerned the notion that a judge should conduct a preliminary review of a statutory scheme before looking at the reasons for decision. The Supreme Court criticized the lower court for departing from the guidance in Vavilov.[9] However, in conducting its own reasonableness analysis, commentators noted that the Supreme Court had itself engaged in such an exercise by assessing international law constraints that had not been argued before the decision-maker or addressed by the court below.[10]

My point here is not to seek to arbitrate between the two courts, but simply to highlight that both courts and commentators were able to rely on the general principles in Vavilov to support their arguments. They could do this because the framework is a coherent whole, almost like a civil code for judicial review of administrative action. This is a significant advance on where the law was a decade ago, when the available principles were far too general to be of any assistance in responding to operational difficulties. Indeed, the general principles could sometimes aggravate or contribute to operational difficulties.[11]

The Requirement to Provide Coherent Reasons

Second, the Supreme Court provided guidance on the basic required content of reasons. These must be coherent – describing how the decision-maker reasoned from its premises to its conclusion[12] – and avoid a short list of basic fallacies, “such as circular reasoning, false dilemmas, unfounded generalizations or an absurd premise”.[13] Building on this, courts have subsequently observed that administrative boilerplate that is not responsive to the evidence or arguments before the decision-maker is insufficient to meet the basic required content demanded by Vavilov.[14] That is, reasons that lack “any real analysis” because they are conclusory are, as far as Vavilov is concerned, no reasons at all.[15]

The Requirement to Respect Legal and Factual Constraints

Third, and most importantly, there is a general list of legal and factual constraints that are likely to be relevant to determining reasonableness. Even though the majority judges were careful to observe that their exegesis of reasonableness was not intended to be exhaustive, but rather illustrative of the sort of circumstance in which a decision would be afflicted by a sufficiently serious shortcoming to render it unreasonable,[16] this has been treated as a category of unreasonableness in and of itself. Indeed, it is the dominant category — unless there is a patent (dare I say patently unreasonable) shortcoming in the reasons, the relevant question for the reviewing court will be whether the decision is reasonable given the relevant statutory constraints.

The Tensions within Vavilovian Reasonableness Review

If nothing else, then, Vavilov provides meaningful guidance where, before, none had been present. Of course, it is entirely proper to point out that Vavilovian reasonableness review contains internal tensions that, indeed, have been propagated by post-Vavilov decisions.

For example, on the one hand, responsive justification is a central principle in Vavilov and requires that administrative decision-makers demonstrate through their reasons that they have listened to the parties, indicating a citizen-led approach to administrative law; but on the other hand, Vavilov insists that decisions must be consistent with the relevant legal constraints. Hence, in the Mason case, the Supreme Court found an administrative interpretation of law to be unreasonable because it violated a relevant legal constraint (the international law prohibition on refoulement), even though the international law argument had not been made to the administrative decision-maker.

Another example is the tension between the Vavilovian proposition that administrative decision-makers are not required to interpret statutes as judges would and the accompanying proposition that interpretations of law must comply with the principles of statutory interpretation developed by judges. Hence in the Information Commissioner case, an interpretation of access-to-information law was found to be unreasonable because of a failure to consider statutory purpose despite a well-reasoned and thorough decision considering the problem from a multiplicity of angles well-known in world of access-to-information law.

Does this doom us to “still more confusion and instability in the law”?[17] Here, I would plead in favour of Vavilov that its emphasis on principles means that disagreement about the application of the framework in individual cases can be addressed by reference to those principles. In other words, the resources for a critique of Mason or Information Commissioner are found in Vavilov itself. There is no need to resort to higher-level principles relating to administrative law, Canada’s constitutional structure, or the requirements of liberal democracy. Critique can be based on such principles but it need not be. The problem prior to Vavilov was that the principles underpinning substantive review in administrative law were either unclear or pitched at such a high level of abstraction as to practically invite disagreement.[18] The advantage of Vavilov is that it provides the tools for immanent critique: disagreement can never be eliminated from the law, but it can be structured and, if nothing else, Vavilov does that. To anticipate a point I will seek to bludgeon home below, this goes to the sociological legitimacy of the framework. Pulling back to take a broader view of the landscape of legitimacy, it is notable that the meaning given to reasonableness review in Vavilov is much ‘thicker’ than the meaning giving to the rule of law and institutional design: substance is favoured over form; administrative supremacy is favoured over judicial supremacy; and reason is favoured over authority. Sauce for the goose and just as much for the gander.

The Concept of Constraint

And so to the heart of this paper: what is the nature of the legal and factual constraints that are central to the application of the reasonableness standard?

The question is prompted by the charge of the concurring judges in Vavilov that the majority’s approach upends Canadian administrative law by “dramatically expanding the circumstances in which generalist judges will be entitled to substitute their own views for those of specialized decision-makers who apply their mandates on a daily basis”.[19] There is no doubt that this charge is amply made out in respect of the change made in Vavilov to the treatment of statutory rights of appeal. As noted above, this change stripped deference on questions of law from expert regulators.[20] But I am more interested for present purposes in whether the charge has been made out in respect of reasonableness review. Here, the concurring judges feared that Vavilovian reasonableness review would invite reviewing courts to “dissect administrative reasons” by providing “a wide-ranging catalogue of hypothetical errors to justify quashing an administrative decision”.[21]

Assessing whether this fear has been realized requires carefully considering the majority reasons in Vavilov and post-Vavilov case law. I will cover four points: first, explaining how the constraints are objective in nature, and thus subject to judicial determination; second, explaining that a failure to provide reasons on a relevant constraint will render an administrative decision unreasonable; third, explaining that in some situations the relevant constraints will require the provision of more extensive reasons; and fourth, using a series of cases involving the treatment of precedent in the reasonableness analysis to provide an assessment of reasonableness review under the new framework.

The Relevant Legal and Factual Constraints are Objective in Nature

A reasonable decision “must be justified in relation to the constellation of law and facts that are relevant to the decision”.[22] In this constellation of law and facts will be legal and factual “[e]lements” that “operate as constraints on the decision maker”.[23] In general, these are:

the governing statutory scheme; other relevant statutory or common law; the principles of statutory interpretation; the evidence before the decision maker and facts of which the decision maker may take notice; the submissions of the parties; the past practices and decisions of the administrative body; and the potential impact of the decision on the individual to whom it applies.[24]

Each of these constraints is objective in nature. Either there is a governing statutory scheme bearing on a decision, or there is not;[25] either there are relevant statutory or common law principles, or there are not;[26] either there are harsh consequences or not;[27] and so on. These are binary questions and, ultimately, it can only fall to the reviewing court to determine whether or not a particular constraint is relevant.

Consider Ontario (Attorney General) v. Ontario (Information and Privacy Commissioner).[28] Here, the Commissioner had ordered that mandate letters issued by the Premier of Ontario to his cabinet ministers should be released under the Freedom of Information and Protection of Privacy Act.[29] The government of the province had resisted disclosure on the basis of s. 12(1) of the Act, which creates an exemption “where the disclosure would reveal the substance of deliberations” of the cabinet for a range of documents “including” (but not limited to) agendas, minutes or lists of policy options presented to cabinet. The basic premise of the Commissioner’s detailed reasons for decision was that mandate letters, which memorialize decisions that have already been taken, would not reveal the substance of deliberations.

Applying reasonableness review, Karakatsanis J concluded that the Commissioner’s decision was unreasonable. The key error was that the Commissioner had failed to have adequate regard to the constitutional context: “Because s. 12(1) was designed to preserve the secrecy of Cabinet’s deliberative process, the constitutional dimension of Cabinet secrecy was crucial context in interpreting s. 12(1)”.[30] This caused a loss of confidence in the outcome.[31] The Commissioner erred in two ways.

First, he gave too narrow a scope to s. 12(1). The Commissioner focused only on two rationales for cabinet confidentiality — promoting ministerial candour and preserving collective solidarity — to the exclusion of a third, efficient government. Failing to take this rationale into account caused him to ascribe too narrow a purpose to s. 12(1) and to fail to respond to one of the government’s submissions.[32] As a result, the Commissioner also “did not acknowledge Cabinet Office’s submission that determining “when and how” to communicate policy priorities to the public and opposition parties is itself an important part of Cabinet’s deliberative process”.[33] In particular, the Commissioner “concluded that “outcomes” of the deliberative process are not encompassed by the opening words of s. 12(1), full stop, without acknowledging that an important part of Cabinet confidentiality is government’s prerogative to decide how and when to announce policy priorities (see para. 104)”.[34] Second, the Commissioner failed to have regard to constitutional conventions and traditions relating to the nature of cabinet decision-making and the premier’s role in the process.[35]

Both of these failures related to constraints that bore on the decision-maker. The Supreme Court necessarily had to form the view that these constraints were relevant.

Something similar can be said about the Supreme Court’s decision in Mason. Relevant here was subparagraph 34(1)(e) of the IRPA, which makes inadmissible anyone “engaging in acts of violence that would or might endanger the lives or safety of persons in Canada”. In concrete terms, a finding of inadmissibility will typically lead to deportation from Canada. M was charged with attempted murder after discharging a firearm and injuring two people at a concert at the Canadian Legion in British Columbia. The charges were stayed, however, and the applicant was thus not convicted. The Minister argued — and the Immigration Appeal Division agreed — that the applicant’s conduct came within s.34(1)(e). M argued on judicial review that it was unreasonable to interpret s.34(1)(e) as encompassing “acts of violence” which do not occur in the context of terrorism, war crimes or organized criminality, these being the concerns underpinning the inadmissibility provisions of the IRPA. There must be, M argued, a national security nexus to s. 34(1)(e).

Jamal J held that the tribunal’s decision was unreasonable.

First, the tribunal had failed to grapple with M’s argument that s. 34(1)(e) requires a security nexus because the availability of discretionary relief (from the responsible minister) was narrower for s. 34(1)(e) than the relief available for serious criminality and criminality offences (which may lead to inadmissibility under s. 36). The logic of M’s argument is that the context of the statute indicates that national security crimes are to be treated more seriously than other crimes, but on the opposing interpretation, less serious crimes (indeed, here, one for which there was not even a conviction) would carry the most serious possible consequences.[36] The tribunal had also failed to grapple with a related argument about the impact of a finding under s. 34(1)(e) on the pre-removal risk assessment to be conducted by the responsible minister before deporting M or a similarly situated person. Again, M’s argument was that the statutory scheme would only make sense if the most serious consequences attached to the most serious inadmissibility ground, namely national security.[37]

Second, the tribunal failed to grapple with M’s argument that not requiring a national security nexus would lead to absurd consequences that Parliament could not possibly have intended, rendering people inadmissible to Canada for everything from domestic altercations to bar fights and schoolyard fights, including young people who would otherwise be protected from the harsh consequences of the law by the youth offender system.[38] According to Jamal J, the tribunal should have considered this point, which was not a “minor aspect” of the interpretive context.[39]

Third, the tribunal had failed to consider Canada’s international law obligations. Jamal J’s analysis is lengthy, but the key point was that the tribunal’s interpretation would allow “a foreign national found inadmissible under s. 34(1)(e) to be subject to refoulement contrary to Article 33(1) of the Refugee Convention”, that is, to deportation to a country where they would face a risk of persecution.[40]

For Jamal J to conclude that these shortcomings in the tribunal’s decision contributed to a finding of unreasonableness, he first had to conclude that each was a relevant constraint. For the first two shortcomings, Jamal J had to form the view that these were key arguments advanced before the tribunal. Once it was established that they were key arguments, the tribunal was then under an obligation to grapple with them, which it had evidently failed to do. For the third shortcoming, Jamal J again had to determine whether the obligations operated as a constraint on the tribunal: having analyzed the statute, he concluded that they were.[41] Given that the tribunal had not provided reasons on any of these constraints, the decision was necessarily unreasonable. For now, my point is that the constraints were authoritatively and definitively identified by the reviewing court.

The decision in Commission scolaire francophone des Territoires du Nord-Ouest v. Northwest Territories (Education, Culture and Employment)[42] is similar in character.  This case involved the review of ministerial discretion. Section 23 of the Charter of Rights and Freedoms creates categories of ‘rights holders’ who, because of their personal circumstances as speakers or learners of a minority language, “have the right to have their children receive primary and secondary school instruction in that language in that province”. Here, the Minister had adopted a policy that expanded the s. 23 categories and made a wider group of children eligible to attend school in French. However, the children at issue in this case did not fall within the scope of the policy. The children were either French-language speakers, otherwise embedded in the French-language community in the Northwest Territories or would contribute to the vitality of the community by attending school in French. In each case, the provincial Francophone school board recommended that they be permitted to attend a French-language school. But the children did not fall within the scope of the policy. Essentially on that basis the Minister refused to permit them to attend a French-language school.

The children successfully invoked the values underlying s. 23 of the Charter. According to the Supreme Court, the underlying values of s. 23 were engaged on the facts, as there was a “clear link” between s. 23 and the exercises of discretion “because the decisions were likely to have an impact on a minority language educational environment”.[43] Preserving and developing the minority-language community (which admission of the children would have contributed to) are s. 23 values[44] which the Minister was required to consider.[45]  The Minister failed to justify her decisions given the evidence of this link: her reasons did not demonstrate that she “meaningfully addressed the values of preservation and development of the Francophone community of the Northwest Territories so as to reflect the significant impact that the decisions might have on it”.[46] Here, again, the relevance of a constraint – Charter values – was determined by the Court.

The Absence of Reasons on a Relevant Constraint is Unreasonable

In the Mandate Letters case, Mason and CSFTNO, the Supreme Court determined whether or not a constraint was relevant. Having determined, in each instance, that a constraint was relevant, the Supreme Court went on to consider whether the decision-maker had provided reasons in relation to the constraint. In the absence of reasons explicitly responsive to the constraints, the decisions were found to be unreasonable (and, of course, had there been reasons that were conclusory or mere boilerplate, the same result would have been reached).

The point is best appreciated a contrario. Where reasons are provided in respect of a constraint, the decision will most likely be held to be reasonable. Consider one of the companion cases to Vavilov, the decision in Canada Post Corp. v. Canadian Union of Postal Workers,[47]  released the following day. At issue here was the scope of s. 125(1)(z.12) of the Canada Labour Code, pursuant to which an employer shall,

…in respect of every work place controlled by the employer and, in respect of every work activity carried out by an employee in a work place that is not controlled by the employer, to the extent that the employer controls the activity…ensure that the work place committee or the health and safety representative inspects each month all or part of the work place, so that every part of the work place is inspected at least once each year.

A complaint was filed about Canada Post’s compliance with this duty in relation to employees in Burlington, Ontario. Canada Post took the view that the duty under the Code extended only to its local depot, not to all of the carrier routes its employees traipse along to reach letter boxes all around Burlington. The Union favoured a more liberal interpretation, which would encompass the routes and the letter boxes. The stakes here were high, as a determination in favour of the Union would, in principle, have nationwide repercussions. A Health and Safety Officer agreed with the Union, but an Appeals Officer took Canada Post’s side at the Occupational Health and Safety Tribunal Canada.

By majority, the Supreme Court upheld the Appeals Officer’s decision. Here, Rowe J. noted, the Appeals Officer had provided detailed reasons that “contended with the submissions of the parties throughout his analysis”.[48] Crucially, these reasons “amply” demonstrated that the Appeals Officer “considered the text, context, purpose, as well as the practical implications of his interpretation”.[49] The Appeals Officer had not considered, because he had not been referred to it, another provision of the Code which deals with the concept of control. This, Rowe J. held, was not fatal, not least because the Appeals Officer had not been referred to it but also because the provision in question would, if anything, have bolstered the Appeals Officer’s conclusion.[50]  Put another way, statutory text, context and purpose were relevant legal constraints under Vavilov but once the decision-maker had considered them, there was no basis for a finding of unreasonableness. This is the key distinction between Canada Post and the Supreme Court’s other post-Vavilov reasonableness review cases.[51]

Situations in Which Reasons Must Not Only be Present but Adequate

In some situations, however, it will not be sufficient for reasons to be present. They will also have to be adequate. Determining adequacy has aptly been described as “challenging”.[52] In my view, there are several points in Vavilov at which the majority makes clear that simply having reasons on a particular point is insufficient, or put another way, “the requirement of a reasoned explanation is higher”:[53]

• The decision-maker must “meaningfully grapple with key issues or central arguments”;[54]
• The decision-maker must discharge a “justificatory burden” where it departs from longstanding practices or established internal authority;[55]
• The decision-maker must explain why a decision that “has particularly harsh consequences for the affected individual…best reflects the legislature’s intention”;[56] and
• The decision-maker must “properly justify” an interpretation that expands the scope of its authority.[57]

In respect of these constraints, reasons tout court are not enough: they must be reasons that grapple with arguments or provide adequate justification given a departure from practice, harsh consequences or an apparent expansion of authority. For instance,

• in Party A v. The Law Society of British Columbia, a brief decision about whether or not to anonymize the name of a disciplined lawyer did not grapple with the lawyer’s submissions about the unique circumstances of his case;[58]
• in Canada (Attorney General) v. Honey Fashions Ltd., an importer was owed an “explanation” from the Canada Border Services Agency as to why a past practice had not been followed and how continuing to follow it “would undermine the customs scheme when such agreements had been accepted without question in the past”;[59]
• in Alsaloussi v. Canada (Attorney General), the harsh consequences of a three-year suspension of a passport required a “sufficiently individualized” assessment that explained and justified the result given the relatively minor underlying misconduct;[60] and
• in Galderma Canada Inc. v. Canada (Attorney General), the Patented Medicines Prices Review Board could not properly justify an interpretation of its home statute that would permit it to regulate the price of an unpatented medicine.[61]

I would also add that in a situation where the governing statutory scheme tightly constrains the authority of the decision-maker, the standard of adequacy employed will be more exacting. A good example is Canadian Frontline Nurses v. Canada (Attorney General).[62] Here, Mosley J held that the invocation of the Emergencies Act[63] by the federal government in 2022 to respond to a series of occupations and blockades at sensitive points around the country was unlawful because the reasons given were inadequate given the constraints.

As a legal matter, it is abundantly clear from the text and structure of the Act (and the legislative history) that a high threshold must be met before declaring a public order emergency. A “reasonable grounds” requirement was inserted precisely to limit the emergency powers of the federal government (as were the various review mechanisms);[64]  there must be a threat to the security of Canada;[65] there must be demonstrable lack of provincial capacity;[66] and the provisions of the Act relating to public order emergencies impose procedural and substantive requirements relating to provincial consultation and (in the case of an emergency in one province only) consent.[67] As a factual matter, again it is clear that by the time the Act was invoked, the only remaining traces of the emergency situation were in downtown Ottawa, other protests having been dismantled by use of existing tools in cooperation with provincial authorities.

The federal government’s reasons therefore had to justify the invocation of a public order emergency in light of these objective legal and factual constraints but fell down in various ways. Justice Mosley’s central conclusion was that the evidence did not “support the conclusion that [the purported emergency] could not have been effectively dealt with under other laws of Canada, as it was in Alberta, or that it exceeded the capacity or authority of a province to deal with it”.[68] Justice Mosley also concluded that there was no “threat to the security of Canada” within the meaning of the Act, because “the record does not support a conclusion that the Convoy had created a critical, urgent and temporary situation that was national in scope and could not effectively be dealt with under any other law of Canada”.[69]

By contrast, in situations where a decision-maker is empowered by relatively broad statutory language, the legal and factual constraints will be less tight and the justification requirements correspondingly somewhat slacker. Thus in Canadian National Railway Company v. Halton (Regional Municipality), it was inappropriate to insist that the federal cabinet provide “more extensive reasons” at the end of a lengthy process involving an expert panel and the exercise of ministerial discretion before, finally, cabinet itself weighed in.[70] Cabinet was here exercising broadly worded powers, and sitting in a supervisory capacity, leading to a less demanding standard of adequacy.

Now, in each of these instances, a judge will ultimately have to make a call about adequacy. This is not a call that can be reduced to a mathematical formula: the exercise of judicial judgement is essential. I would not say that this transforms the reasonableness analysis into a subjective value judgement based on a judges’ personal and political predilections. Reasonableness (and adequacy) remains an objective concept, even if there is inevitably some subjectivity in applying it to the facts of a particular case.

Precedent

In the case law on constraints there is some tension between, on the one hand, the Vavilov principle of reasons-first review and, on the other hand, the prohibition on de novo analysis by reviewing courts. The fact that the constraints are objective in nature suggests that a reviewing court will indeed have to make a determination, if not de novo then at the very least on its own initiative, of whether a particular constraint is indeed relevant in a particular case.

The tension is partly resolved by recognizing that any determination of the relevance of a constraint will have to begin with the administrative decision-maker’s reasons for decision. Indeed, if the decision-maker’s view is that a particular constraint is not relevant, it is difficult to see how a reviewing court could legitimately second-guess the view. Where the decision-maker has not explicitly taken a view on the relevance of a constraint, however, it is difficult to see why a court should be precluded from forming its own view. Reasonableness review requires sensitivity to legal and factual constraints, after all, and a reviewing court can hardly be accused of failing to afford respect to an administrative decision-maker’s institutional setting if it is not second-guessing any of the decision-maker’s reasons.

The case law on judicial precedent as constraint helps to illustrate the point. Whether a precedent is relevant to a particular case is a matter that requires objective determination. It is a binary: either the precedent is relevant, or it is not.

Where a decision-maker is correct about the applicability of a precedent, there can be no ground of unreasonableness. Indeed, a reviewing court can make this determination itself without trespassing on the terrain of the decision-maker.[71]

Where, by contrast, a decision-maker is wrong about the relevance of a precedent, further analysis is required. There are two possibilities.

Either (a) the decision-maker assumed the relevance of the precedent and did not give any reasons on the point, or (b) the decision-maker gave reasons to explain why it found that the precedent was applicable or not applicable.[72]

In the former case (a), the decision-maker assumed the relevance of the precedent without providing explanation. In such a circumstance, there cannot be any complaint that a reviewing court that takes a different view has intervened to quash a decision as unreasonable. This is because the decision-maker did not give any reasons on the point: there is nothing to defer to. In Canada (Attorney General) v. Hull, for example, the Social Security Tribunal had treated an obiter comment by a Federal Court judge as binding on the Tribunal, without acknowledging or explaining why an obiter had binding force.[73] This was unreasonable. In a subsequent case with similar facts, decided by the Tribunal before Hull had been handed down, the Tribunal had assumed the obiter applied and, again, its decision was unreasonable.[74]

In the latter case (b), the question again will be the adequacy of the reasons given. In Canada (Attorney General) v. National Police Federation, the Federal Public Sector Labour Relations and Employment Board had engaged in a detailed treatment of a relevant Supreme Court precedent in the labour law area and the Federal Court of Appeal accordingly rejected the invitation “to substitute our interpretation of Wal-Mart for that of the Board to measure the reasonableness of its interpretation against our own”.[75]

The point I take from these cases is that there is a close link between deference and the autonomy of the decision-maker. Where a decision-maker has taken responsibility and positively decided that a particular precedent is applicable, or not applicable, in its particular domain, judicial intervention threatens the autonomy of the decision-maker. It would potentially represent the eulogy for deference feared by Abella and Karakatsanis JJ. By contrast, where a decision-maker has not given reasons on a relevant point, it leaves itself at the mercy of the reviewing court to a significant extent. It is up to the court, by default, to determine whether a particular precedent is relevant or not. If there are no reasons given, there is nothing to defer to, and no interference with any decision-making autonomy.

What is true of “precedent” in the previous paragraph is true, I would say, of “constraint” generally in the Vavilov framework. Deference is, ultimately, earned rather than given, but once it is earned it is difficult to lose. If my conception of constraint is correct, then Vavilov provides a workable means of managing the tensions within its own framework. To link back to the major theme of this paper, the Vavilov framework is itself capable of generating the sociological legitimacy necessary for its own survival.

[1] Paul Daly, « Le droit civil et le droit administratif canadien » dans Élise Charpentier dir., Le code civil – un incontournable? (Éditions Thémis, Montreal, 2025).

[2] Vavilov, at para. 100. See also the concurring reasons, at para. 312.

[3] Vavilov, at para. 13. See also the concurring reasons, at para. 294.

[4] Vavilov, at para. 81. See also the concurring reasons, at paras 291, 296.

[5] Vavilov, at para. 83-84. See also the concurring reasons, at para. 306, 313.

[6] Vavilov, at para. 90. See also the concurring reasons, at para. 292-293.

[7] Vavilov, at para. 93. See also the concurring reasons, at para. 297.

[8] Vavilov, at para. 92. See also the concurring reasons, at para. 299.

[9] Mason, at para. 79.

[10] Paul Daly, “2023 Developments in Administrative Law of Interest to Energy Lawyers” (2024) 12 Energy Regulation Quarterly 7; Mark Mancini, “Issue #108: October 1, 2023”, Sunday Evening Administrative Review (blog), 1 October 2023, online: <https://sear.substack.com/p/issue-108-october-1-2023>

[11] See the discussion of the statutory appeals “flash point” above.

[12] Vavilov, at para. 103.

[13] Vavilov, at para. 104.

[14] See e.g. Safarian v Canada (Citizenship and Immigration), 2023 FC 775, at para. 3; Khosravi v Canada (Citizenship and Immigration), 2023 FC 805, at para. 7; Kashefi v. Canada (Citizenship and Immigration), 2024 FC 856, at para. 8. See generally Nesarzadeh v. Canada (Citizenship and Immigration), 2023 FC 568, at paras. 5-9.

[15] Canada (Citizenship and Immigration) v. Galindo Camayo, 2022 FCA 50, [2022] 4 FCR 220, at paras. 58-60.

[16] Vavilov, at para. 101.

[17] Megan Pfiffer, “What’s the Problem With Substantive Review?” (2024) 69 McGill Law Journal 325, at p. 356.

[18] Dunsmuir v. New Brunswick, 2008 SCC 9, [2008] 1 SCR 190, at para. 27, referring to the rule of law and legislative intent. Again, in Vavilov, the majority wisely chose thinner versions of those potentially capacious concepts.

[19] Vavilov, at para. 201.

[20] See also Vavilov, at para. 251.

[21] Vavilov, at para. 284.

[22] Vavilov, at para. 105.

[23] Vavilov, at para. 105.

[24] Vavilov, at para. 106.

[25] Generally, there will be, though see MacKinnon v. Canada (Attorney General), 2025 FC 422 (prorogation of Parliament).

[26] See e.g. Conifex Timber Inc. v. British Columbia (Lieutenant Governor in Council), 2025 BCCA 62, at para. 122 (suggesting that principles of utilities law are not relevant constraints where a provincial cabinet makes a decision in a utilities matter).

[27] Sticky Nuggz Inc. v. Alcohol and Gaming Commission of Ontario, 2020 ONSC 5916, at para. 68.

[28] 2024 SCC 4.

[29] RSO 1990, c F.31.

[30] Information Commissioner, at para. 27.

[31] Information Commissioner, at para. 23.

[32] Information Commissioner, at paras. 34-35.

[33] Information Commissioner, at para. 37, emphasis original.

[34] Information Commissioner, at para. 39.

[35] at paras. 53-55.

[36] Mason, at para. 91.

[37] Mason, at paras. 94-95.

[38] Mason, at paras. 99, 102.

[39] Mason, at para. 103.

[40] Mason, at para. 109.

[41] Mason, at para. 117.

[42] 2023 SCC 31. I appeared for the intervenor Commission scolaire francophone du Yukon.

[43] CSFTNO, at para. 78.

[44] CSFTNO, at paras. 80-82.

[45] CSFTNO, at para. 83.

[46] CSFTNO, at para. 102.

[47] 2019 SCC 67, [2019] 4 SCR 90.

[48] Canada Post, at para. 60.

[49] Canada Post, at para. 43.

[50] Canada Post, at para. 53.

[51] See also Yatar v. TD Insurance Meloche Monnex, 2024 SCC 8, at para. 74 (failure to have regard to relevant constraints rendered a benefits decision unreasonable; I appeared for the intervener Canadian Telecommunications Association).

[52] Alexion Pharmaceuticals Inc. v. Canada (Attorney General), 2021 FCA 157, [2022] 1 FCR 153, at para. 14.

[53] Alexion, at para. 21.

[54] Vavilov, at para. 128.

[55] Vavilov, at para. 131.

[56] Vavilov, at para. 133.

[57] Vavilov, at para. 109.

[58] 2021 BCCA 130, at para. 64.

[59] 2020 FCA 64, at para. 40.

[60] 2020 FC 364, at paras. 65, 67. See also Canada (Citizenship and Immigration) v. Galindo Camayo, 2022 FCA 50, [2022] 4 FCR 220; Lorne Sossin, “The Impact of Vavilov: Reasonableness and Vulnerability” (2021) 100 SCLR (2d) 265.

[61] 2024 FCA 208, at paras. 13, 15.

[62] 2024 FC 42.

[63] RSC 1985, c 22 (4th Supp).

[64] Commons Debates, 33rd Parliament, 2nd Session, Volume 12 (April 25, 1988), at p. 14765 (Hon. Bud Beatty).

[65] RSC 1985, c 22 (4th Supp), s. 16.

[66] RSC 1985, c 22 (4th Supp), s. 3.

[67] RSC 1985, c 22 (4th Supp), s. 25.

[68] 2024 FC 42, at para. 254.

[69] 2024 FC 42, at para. 294.

[70] 2024 FCA 160, at para. 102.

[71] See e.g. Pepa v. Canada (Citizenship and Immigration), 2023 FCA 102, at paras. 12-17.

[72] This is a reformulation of the second part of the framework suggested by Grammond J in Service d’administration P.C.R. Ltée v. Reyes, 2020 FC 659, at para. 24.

[73] 2022 FCA 82, at para. 32.

[74] Canada (Attorney General) v. Johnson, 2023 FCA 49, at para. 16.

[75] 2022 FCA 80, [2022] 4 FCR 263, at para. 89.

This content has been updated on June 24, 2025 at 12:14.