Vavilov Hits the Road (Updated May 17) (Pinned to the Top of the Blog)

Courts around Canada are beginning to apply the Vavilov framework. Obviously, it is too early to draw any conclusions. Nonetheless, even at this early stage in the life of Vavilovian judicial review there have been some interesting decisions. I have been coming back to this post periodically to update it with significant new decisions. If you think a decision warrants attention, please do get in touch, in the comments or via email. The post is now getting a bit long, so I have added some headings.

Preliminary Issues

An issue which emerged early in 2020 is whether cases argued before Vavilov was handed down should be reargued before being disposed of. Vavilov, as the Supreme Court made clear, is the ‘big bang’ and now provides the starting point for any administrative law analysis (Lamoureux c. Cour du Québec, 2020 QCCS 619, at paras. 5-11). Sometimes, it will be quite obvious that the Vavilov framework changes nothing (see Centre intégré de santé et de services sociaux des Laurentides c. Roy, 2020 QCCS 230, at para. 45; Syndicat de l’enseignement de Champlain c. Commission scolaire Marie-Victorin, 2020 QCCA 135, at para. 31; Beals v. Nova Scotia (Attorney General), 2020 NSSC 60, at paras. 18-19) or changes something so clearly that further argument would be pointless (see Canadian National Railway Company v. Richardson International Limited, 2020 FCA 20, at paras. 41-44). If there is doubt, prudence argues in favour of allowing the parties to make submissions on the consequences of Vavilov. And courts have mostly been prudent.

Several months into the roll-out of the Vavilov framework, however, with some cases argued prior to Vavilov still being decided, it is jarring to see matters decided without further submissions on the effect of Vavilov (see e.g. Li v. Canada (Citizenship and Immigration), 2020 FC 554, at para. 38). Given the increasing body of case law (discussed further below) suggestive of a more demanding standard of reasonableness, parties really should be allowed to make additional submissions on how Vavilovian reasonableness review should be applied in their case (as in British Columbia (Assistant Water Manager) v. Chisholm, 2020 BCSC 545, at para. 29). This point applies with more force to applicants than respondents. Applicants might be prejudiced by a failure to apply an exacting reasonableness standard whereas respondents are unlikely to reframe their arguments in response to Vavilov (and, of course, cannot rewrite their decisions). Accordingly, where a judge is going to strike a decision down anyway, as in Yavari v. Canada (Public Safety and Emergency Preparedness), 2020 FC 469, there is comparatively little unfairness. But my view is that the better course of action is to request submissions on reasonableness review for any outstanding matter where the oral and written submissions pre-dated Vavilov.

Consider, moreover, the decision of the Ontario Divisional Court in Zhou v. Cherishome Living, 2020 ONSC 500. This was an appeal from the Landlord and Tenant Board under s. 210 of the provincial Residential Tenancies Act. Appeals are available on questions of law only. The panel concluded that it was “not appropriate” (at para. 32) to request additional submissions on the consequences of Vavilov. Yet the effect of the panel’s application of Vavilov was that one of the issues raised by the tenants — who were self represented — was one of mixed fact and law and thus “not appealable” (at para. 68). But it was made clear in Vavilov that appeal clauses which do not cover all of the issues in dispute do not preclude unhappy appellants from bringing judicial review proceedings in respect of the other issues (Vavilov at para. 52). Post-Vavilov, the appellants could have commenced judicial review proceedings in parallel to their appeal (indeed, unless and until there is further clarification of this point, I think wise counsel will generally advise clients to do so). That is not to say that a court would ultimately have allowed a judicial review application to proceed (not least because the litigation between the parties has been ongoing for many years), just that it would have been more prudent at least to ask the appellants if they had something to say (see also Van de Sype v Saskatchewan Government Insurance, 2020 SKCA 18, where this problem does not seem to have been identified).

Correctness Review

There have been a couple of divergent opinions on the scope of the correctness categories. The first post-Vavilov decision, Peter v. Public Health Appeal Board of Alberta, 2019 ABQB 989, did embrace correctness review on constitutional issues. By contrast, in Syndicat des employé(e)s de l’école Vanguard ltée (CSN) c. Mercier, 2020 QCCS 95, at paras. 17-19, St-Pierre J applied the reasonableness standard even in the face of an argument based on the quasi-constitutional Quebec Charter. This is, obviously, an area to watch, though I think St-Pierre J’s position is more persuasive given the narrow conceptual basis provided for the correctness categories in Vavilov (see similarly Coldwater First Nation v. Canada (Attorney General), 2020 FCA 34, at para. 27; Borradaile v British Columbia (Superintendent of Motor Vehicles), 2020 BCSC 363, at para. 34; Redmond v British Columbia (Forests, Lands, Natural Resource Operations and Rural Development), 2020 BCSC 561, at paras. 28-31). St-Pierre J applied correctness in Régie de l’assurance maladie du Québec c. Morin, 2020 QCCS 294, but this was an uncontroversial example of overlapping jurisdiction (at paras. 9-11), which is a category of quite limited scope (see e.g. Heffernan v. Saskatchewan Police Commission, 2020 SKQB 65, at paras. 25-27).

Of course, direct challenges to the constitutionality of statutes or similarly general norms will continue to attract correctness review (as in Canada Union of Correctionnel Officers) v. Canada (Attorney General), 2019 FCA 212, at para. 21), as will decisions touching on the borderline between provincial and federal regulation (as in Canada (Attorney General) v. Northern Inter-Tribal Health Authority Inc., 2020 FCA 63, at paras. 12-13), but like it or not, the Doré framework survived Vavilov and has perhaps emerged even stronger. In Redmond v British Columbia (Forests, Lands, Natural Resource Operations and Rural Development), 2020 BCSC 561, at para. 30, Masuhara J relied on the expertise of the decision-maker to justify the application of the Doré framework. Post-Vavilov, expertise is not relevant to the selection of the standard of review, but contrary to one view, this means the argument for deference is a fortiori: Vavilov teaches that deference applies in all scenarios unless the rule of law (or legislative intent) requires otherwise which, given the narrow conceptual basis provided for it, will be exceedingly rare (see also Doré at paras. 51-52).

The category most apt to be expanded after Vavilov is surely the ‘questions of central importance to the legal system’ category. But the narrow rule of law basis for the correctness categories does not provide a solid foundation for such arguments. In Bank of Montreal v. Li, 2020 FCA 22, for example, the issue was whether an employee who had signed a release on conclusion of her employment could nonetheless make an unjust dismissal claim. An adjudicator held she could and, on judicial review, the company sought to persuade the courts to apply correctness review on the basis that the issue of whether an individual can waive a statutory entitlement is a general one requiring definitive judicial resolution. De Montigny JA was not persuaded, concluding that the waiver issue would not have systemic or constitutional implications and noting that “framing an issue in a general or abstract sense is not sufficient to make it a question of central importance to the legal system as a whole” (at para. 28. See similarly Beach Place Ventures Ltd. v British Columbia (Employment Standards Tribunal), 2020 BCSC 327, at paras. 32-34). The question of privilege in College of Physicians and Surgeons v. SJO, 2020 ONSC 1047, at para. 10 was summarily held to be subject to correctness review but as Mark Mancini notes there is ample precedent in support of this conclusion and it does not suggest that this category has widened post-Vavilov (see also Alberta Health Services v Farkas, 2020 ABQB 281, at para. 37). A more curious decision is Low v. Nova Scotia Police Complaints Commissioner, 2020 NSSC 113, where the common law doctrine of discoverability was held to apply in the context of police complaints, on what looked quite like a correctness basis (at paras. 46-51) but without consideration of paras. 111-114 of Vavilov, where the issue of statutory or common law constraints on administrative decision-makers is authoritatively addressed.

One question left unresolved by Vavilov was the standard of review applicable to regulations. On the one hand, the question of whether a particular regulation is intra vires its parent statute might be said to require a final, definitive answer from the courts, engaging Vavilov’s rule-of-law justification for correctness review. On the other hand, reasonableness is the presumptive standard of review and the most obvious justification for correctness review of regulations — that they relate to jurisdiction — was eliminated by Vavilov (1120732 B.C. Ltd. v. Whistler (Resort Municipality), 2020 BCCA 101, at para. 39; G.S.R. Capital Group Inc. v The City of White Rock, 2020 BCSC 489, at para. 71) and, as Loparco J observed in Morris v Law Society of Alberta (Trust Safety Committee), 2020 ABQB 137, at para. 40, the Supreme Court spoke explicitly to the issue of jurisdiction or vires in Vavilov:

[T]he Supreme Court concluded that the question of whether or not a delegated decision-maker should “be free to determine the scope of its own authority [can] be addressed adequately by applying the framework for conducting reasonableness review.” The Court specifically endorsed the use of reasonableness review standard in cases “where the legislature has delegated broad authority to an administrative decision maker that allows the latter to make regulations in pursuit of the objects of its enabling statute”: Ibid at para 66, citing Green v Law Society of Manitoba, 2017 SCC 20, [2017] 1 SCR 360; West Fraser Mills Ltd v British Columbia (Workers’ Compensation Appeal Tribunal), 2018 SCC 22, [2018] 1 SCR 635.

See similarly Portnov v. Canada (Attorney General), 2019 FC 1648, at para. 23.

Notably, Loparco J applied the reasonableness standard even though the rule at issue touched upon matters relating to solicitor-client privilege (which are subject to correctness review), reasoning that the relevant issue was whether the Law Society had the authority to enact the rule (at para. 45). However, she arguably muddied the waters (at para. 57) by reviewing the reasonableness of the Trust Safety Committee’s determination of the reasonableness of the rule (rather than directly assessing the reasonableness of the rule), creating a double deference problem and not engaging in the robust reasonableness review Vavilov envisages when compliance with a decision-maker’s governing statutory scheme is in issue.

There has been disagreement on the niche issue of whether arbitration decisions are subject to the appellate review framework post Vavilov or, as was the case previously, subject to the judicial review framework (per Sattva Capital Corp. v. Creston Moly Corp., 2014 SCC 53, [2014] 2 SCR 633). In Buffalo Point First Nation et al. v. Cottage Owners Association, 2020 MBQB 20, at para. 56 and Allstate Insurance Company v. Her Majesty the Queen, 2020 ONSC 830, at para. 19 the courts took the view that Sattva has been superseded by Vavilov. But in Cove Contracting Ltd v Condominium Corporation No 012 5598 (Ravine Park), 2020 ABQB 106, at para. 10, it was held that Sattva continues to bind. Subject, obviously, to the details of the statutory provision in a given jurisdiction, I think the better view must be that the use of the word “appeal” in relation to arbitration decisions now carries with it the appellate review framework (correctness on extricable questions of law, palpable and overriding error for the rest). Hainey J doubted this view in Ontario First Nations (2008) Limited Partnership v. Ontario Lottery And Gaming Corporation, 2020 ONSC 1516, at paras. 61-75, reasoning that the legislative intent branch of the Vavilov framework did not apply because the right of appeal was found in an agreement between the parties, not the provincial arbitration statute. To my eye, this distinction is far too fine. It is true that s. 45 of the Arbitration Act does not explicitly provide for appeals, but it does draw distinctions between the treatment of questions of law (s. 45(1) and (2)) and questions of fact and questions of mixed fact and law (s. 45(3)). The natural reading, in light of Vavilov, is that the legislature has proceeded on the basis that the courts will apply the appellate review framework. Whilst no standard of review is specified in the legislation, I do not think this is necessary to engage the legislative intent branch of Vavilov where the statute provides for an appeal (it is necessary where the statute purports to set out grounds of review: see the discussion of patent unreasonableness below).

In Bureau de la sécurité privée c. Tribunal administratif du Québec, 2020 QCCS 571, St-Pierre J was asked to identify a new correctness category but demurred, on the basis that this was a task for the Court of Appeal or Supreme Court (at para. 14); I tend to think, however, that it would be perfectly appropriate for a first-instance judge to do this, albeit that here it seems the argument for the applicant was not particularly strong.

In terms of correctness review on statutory appeals (under the Housen v. Nikolaisen framework), the following comment from Swinton J is notable:

While the Court will ultimately review the interpretation of the Act on a standard of correctness, respect for the specialized function of the Board still remains important.  One of the important messages in Vavilov is the need for the courts to respect the institutional design chosen by the Legislature when it has established an administrative tribunal (at para. 36).   In the present case, the Court would be greatly assisted with its interpretive task if it had the assistance of the Board’s interpretation respecting the words of the Act, the general scheme of the Act and the policy objectives behind the provision.

Planet Energy (Ontario) Corp. v. Ontario Energy Board, 2020 ONSC 598, at para. 31. See also Edmonton (City of) v Ten 201 Jasper Avenue Ltd, 2020 ABCA 60 and Borgel v Paintearth (Subdivision and Development Appeal Board), 2020 ABCA 192, at paras. 21-22 (taking tribunal (and lower court) jurisprudence into account in applying the correctness standard to an extricable question of law).

As I suggested in my paper on the Vavilov framework, if judges continue to consider and give weight to administrative interpretations of law on statutory appeals, deference might not be dead just yet. That said, in Municipal Property Assessment Corporation v. Zarichansky, 2020 ONSC 1124, Favreau J did not consider in detail the Ontario Assessment Board’s rationale for taking a pro-ratepayer view in situations where MPAC (which assesses properties in Ontario for the purposes of calculating municipal property taxes) has failed to discharge its burden of proof. Rather, Favreau J insisted (on correctness review) that the Board was bound by the terms of its governing statute: my cursory review of the Board’s jurisprudence suggests, however, that it had provided a reasoned basis for its approach; there is little consideration by Favreau J of whether his approach will create difficulties for the Board, MPAC and ratepayers in future cases.

One other way in which deference might persist on statutory appeals post-Vavilov is in the classification of matters falling within a decision-maker’s expertise as factual questions or mixed questions of fact and law (see e.g. 1085372 Ontario Limited v. City of Toronto, 2020 ONSC 1136; Donaldson c. Autorité des marchés financiers, 2020 QCCA 401). As Watson JA rightly insisted in Canadian Natural Resources Limited v Elizabeth Métis Settlement, 2020 ABCA 148, at para. 32 a question of law must be extricable to be subject to correctness review: “It must go to the defining elements of the relevant legal test and not merely to how the tribunal assesses the evidence before applying the test”. Relatedly, Slatter JA observed in Yee v Chartered Professional Accountants of Alberta, 2020 ABCA 98, at para. 30 that “i) the standard of practice the profession expects in any particular case, and ii) whether, on the facts, the professional subject to discipline has met that standard” are questions of mixed fact and law calling for deferential review. If so, the scope for appellate oversight of professional disciplinary decisions will be quite limited and the change wrought by Vavilov not especially dramatic (albeit that in Yee the appeal was allowed!). It is also worth mentioning the possibility that a right of appeal might be restricted by legislation to a consideration of the reasonableness of a decision, that is, the legislature might specify a standard of review in respect of particular decisions: see e.g. Ahmadzai (Re), 2020 ONCA 169, at para. 12; Nguyen (Re), 2020 ONCA 247, at para. 28 (appeals from a Criminal Code Review Board to the Court of Appeal).

There has been some discussion of what counts as an “appeal” clause, with cases falling on two sides of the line. In British Columbia, it has unsuccessfully been argued that s. 623 of the Local Government Act, which makes provision as to how an application to set aside a bylaw can be made, constitutes an appeal clause: properly interpreted, s. 623 is simply a “particular procedure that applies to the judicial review of decisions of local governments” (G.S.R. Capital Group Inc. v The City of White Rock, 2020 BCSC 489, at para. 69); it “serves to clarify for whom a right to judicial review exists, the powers the court can exercise on such a review, and what procedural requirements must be met to assert that right”, but does not constitute a legislative institutional choice to have courts review matters on a correctness standard (O’Shea/Oceanmount Community Association v Town of Gibsons, 2020 BCSC 698, at para. 51). On the other side of the line, at the federal level, it is clear (if further clarity were needed) that judicial reviews of trademark decisions in the Federal Court are, where new evidence is admitted, de novo, with a correctness standard thus applying (see e.g. Pentastar Transport Ltd. v. FCA US LLC, 2020 FC 367 at paras. 42-45; The Clorox Company of Canada, Ltd. v. Chloretec S.E.C., 2020 FCA 76, at paras. 21-23; FFAUF S.A. v. Industria di Diseno Textil, S.A., 2020 FC 520, at paras. 26-28 and 39) Meanwhile, in Anderson v Saskatchewan Apprenticeship and Trade Certification Commission, 2020 SKCA 54, at para. 11 Barrington-Foote JA clarified (if any clarification were needed) that a legislature may preclude “an appeal to the Court of Appeal from a decision of the Court of Queen’s Bench”, a proposition which was not qualified in any way by Vavilov.

Lastly, one issue which may crop up from time to time is the scope of procedural fairness, which, of course, is not subject to the Vavilov framework. Although Synott J applied reasonableness review in Ville de Montréal c. Mouscardy, 2020 QCCS 1448, at paras. 31-35, on the basis that the audi alteram partem rule is not a question of central importance to the legal system, it is tolerably clear from paragraphs 23 and 77 of Vavilov that procedural issues are subject to the Baker framework for procedural fairness, with only the “merits” subject to the Vavilov framework. In Hildebrand v Penticton (City), 2020 BCSC 353, at paras. 29-30, Weatherhill J applied the Vavilov framework to a decision not to grant an adjournment. It is debatable whether this was a matter going to the “merits” of the underlying decision, to which Vavilov clearly applies, or related to procedure, in which case Vavilov would not apply. There will be future cases presenting similar difficulties of classification.

Reasonableness Review

Vavilov teaches that judicial review should be neither a line-by-line treasure hunt for error (see Radzevicius v. Workplace Safety and Insurance Appeals Tribunal, 2020 ONSC 319, at paras. 35-39; Mudjatik Thyssen Mining Joint Venture v. Billette, 2020 FC 255, at paras. 59-77; Bashir v. Canada (Attorney General), 2020 FC 278, at para. 29) nor an effort in redoing the work of the administrative decision-maker (see Bombardier Aéronautique inc. c. Commission des normes de l’équité, de la santé et de la sécurité au travail, 2020 QCCA 315, at paras. 30-46; Yassin v. Canada (Attorney General), 2020 FC 237, at paras. 42-43; Mohammed v. Canada (Citizenship and Immigration), 2020 FC 234, passim; Huang v. Canada (Citizenship and Immigration), 2020 FC 241, at para. 27; Singh v. Canada (Citizenship and Immigration), 2020 FC 328, passim; Theivendram v. Canada (Public Safety and Emergency Preparedness), 2020 FC 419, at para. 40). Reasons for administrative decisions should be read fairly, with due attention to the decision-making context and the arguments made before the decision-maker (see e.g. Calgary (City) v Sunridge Mall Holdings Inc, 2020 ABQB 148; Saleh v. Canada (Citizenship and Immigration), 2020 FC 457, at para. 15). Departures from prior decisions are entirely possible, as long as adequate justification is provided (see e.g. Canada (Public Safety and Emergency Preparedness) v. Shen, 2020 FC 405; Canada (Attorney General) v. Honey Fashions Ltd., 2020 FCA 64; and Nation Rise Wind Farm Limited Partnership v. Minister of the Environment, Conservation and Parks, 2020 ONSC 2984, at paras. 83-85). And the Quebec Court of Appeal has signalled, in excellent reasons by Moore JA, that it is on the look-out for unfaithful applications of reasonableness review, so-called disguised correctness review: see Syndicat de l’enseignement de Champlain c. Commission scolaire Marie-Victorin, 2020 QCCA 135, at paras. 41, 65; Syndicat des métallos, section locale 9449 c. Glencore Canada Corporation, 2020 QCCA 407, at paras. 33-34).

My view is that the methodology of Vavilovian reasonableness review is inherently deferential. But it is certainly arguable that Vavilov has, in respect of supplementation, responsiveness and justification, set a slightly higher bar for decision-makers than the pre-Vavilov regime. Moreover, judicial re-writing of defective decisions has been definitively ruled out: see Hasani v. Canada (Citizenship and Immigration), 2020 FC 125, at paras. 67-68; and see generally Gleason JA’s even-handed analysis in Canada (Attorney General) v. Zalys, 2020 FCA 81. This was certainly Gauthier JA’s conclusion in the important decision in Farrier c. Canada (Procureur général), 2020 CAF 25. Quashing as unreasonable a one-page decision from the Parole Board which failed to engage with the applicant’s arguments, she commented:

Before Vavilov I would probably have found, as did the Federal Court, that, in light of the presumption that the decision-maker considered all of the arguments and the case law before it and after having read the record, the decision was reasonable. The absence of reasons dealing with the first two issues before the Appeal Division was not at the time sufficient to set aside the decision. It was implicit that the Appeal Division did not accept that the Board’s interpretation of the Act was erroneous, particularly considering subsection 143(1) of the Act. Under the circumstances, the administrative decision-maker was presumed to have rejected Mr. Farrier’s arguments regarding any prejudice caused by the lack of a recording regardless of whether the Act provides for such a recording or whether there was simply a breach of the Manual. Such a finding was one of the possible outcomes given the Supreme Court’s decision in CUPE, even if that decision was not cited by the Appeal Division (at para. 12. See similarly Walker v. Canada (Attorney General), 2020 FCA 44, at para. 10).

In the absence of any internal policies, previous Parole Board jurisprudence or other explanations for not addressing the applicant’s arguments (Vavilov at para. 94; Haddad Pour v. The National Dental Examining Board of Canada, 2020 ONSC 555, at paras. 37-40), the conclusion that the decision was unreasonable was irresistible. Courts post Vavilov might not be able or willing to “infer” that an argument or evidence was considered in the absence of reasons dealing with the argument or evidence (Mattar v. The National Dental Examining Board of Canada, 2020 ONSC 403, at paras. 51-52; Walker v. Canada (Attorney General), 2020 FCA 44, at para. 10). Testing the limits of “coherence and justification” is not a wise strategy, as Rennie JA put it in Langevin v. Air Canada, 2020 FCA 48, at para. 18, where the Canada Industrial Relations Board had reverted to a “conclusory, boiler-plate statement” in respect of a point in dispute: but there, luckily for the Board, a response to the point would have been “of little assistance” (at para. 19) and so the decision was upheld (see also the very relaxed approach — too relaxed, in my view — taken in ImagineAbility Inc v City of Winnipeg, 2020 MBCA 39, at para. 45). In Osun v. Canada (Citizenship and Immigration), 2020 FC 295, at para. 26, a boilerplate comment to the effect that the decision-maker had given a piece of evidence “careful consideration” was insufficient, as the decision lacked an “assessment” of the evidence. Failure to provide reasons at all risks inviting reasonableness review so robust as to resemble (if not transform into) correctness review (see e.g. Alexis v Alberta (Environment and Parks), 2020 ABCA 188, especially at paras. 35-41 and 139-146).

There have already been some robust applications of reasonableness review, justified by reference to the emphasis in Vavilov on responsiveness. For example, in Langlais c. Collège des médecins du Québec, 2020 QCCA 134, at paras. 39-44, it was unreasonable for the Collège to fail to address the regulatory provision which a doctor invoked to support his application for recognition as a specialist in internal medicine (necessary because, in 2012, the Collège had introduced more stringent standards in this regard) (see also Alexander v. Canada (Citizenship and Immigration), 2020 FC 313 (failure to respond to a mass of evidence was unreasonable), Alsaloussi v. Canada (Attorney General), 2020 FC 364 (failure to grapple with contradictory evidence, in a context where the decision (to bar the applicant from passport services for three years) had significant consequences for the individual concerned), Chaffey v. Her Majesty the Queen in Right of Newfoundland and Labrador, 2020 NLSC 56, at para. 50 (failure to grapple with the critical factual issue raised by the applicant), and Pryce v. Canada (Citizenship and Immigration), 2020 FC 377 (where a question has been certified to the Federal Court of Appeal)).

Similarly, in Patel v. Canada (Citizenship and Immigration), 2020 FC 77, at para. 17, Diner J noted that Vavilov requires “basic responsiveness” to the evidence presented (and found it lacking here), in Samra v. Canada (Citizenship and Immigration), 2020 FC 157, at para. 22, Favel J found a decision unreasonable because it “lacked analysis”: “the officer’s decision is merely a recitation of the evidence before him followed by a conclusion” and in Li v. Canada (Citizenship and Immigration), 2020 FC 279, at para. 13, Fuhrer J struck down a sparsely reasoned study permit decision issued by a line officer who failed to “engage” with the applicant’s evidence. These Federal Court cases all addressed decisions made by line decision-makers processing hundreds of thousands of applications (see also Low v. Nova Scotia Police Complaints Commissioner, 2020 NSSC 113, at para. 58; A.B. v. Canada (Citizenship and Immigration), 2020 FC 203, at para. 53). In Rodriguez Martinez v. Canada (Citizenship and Immigration), 2020 FC 293, McHaffie J explained that while institutional constraints “must inform the assessment of reasonableness” (at para. 13), a decision-maker — even a line decision-maker — must nonetheless respond to the evidence (at paras. 15-17). Given the emphasis on responsiveness in Vavilov, and this line of cases, the analysis in Tarnow v NWT Legal Aid Commission, 2020 NWTSC 13 came as a surprise. Here, a decision not to allocate work to counsel on a legal aid panel who had worked for the Commission in the previous calendar year was considered to be reasonable in view of the “multiple factors” the Commission had to balance (at para. 50). But the fact that work had so recently been assigned to the applicant called, I think, for specific justification (see e.g. Canada (Attorney General) v. Honey Fashions Ltd., 2020 FCA 64, at para. 38, per De Montigny JA).

The decision of the Ontario Court of Appeal in Romania v. Boros, 2020 ONCA 216 warrants a special mention as it was issued in the context of extradition proceedings, where the executive has typically been given a wide margin of appreciation. Although the Minister had provided a lengthy, 20-page letter ordering the surrender of the applicant to Romania, he did not provide an adequate justification for an eight-year delay in seeking the extradition. The applicant had been convicted in absentia in 2000; there was a dispute about the state of knowledge of the Romanian authorities and, in particular, whether they knew in 1998 that the applicant was in, or soon to arrive in, Canada, long before making the extradition request in 2008. That the “combined Canadian delay of nearly 8 years is not addressed beyond an implicit general claim that these matters take a long time” meant the decision was not “adequate” (at para. 29). Strikingly, although the Supreme Court held in Sriskandarajah v. United States of America, 2012 SCC 70, [2012] 3 SCR 609 that procedural fairness does not require extradition authorities to seek out evidence which may be helpful to an applicant, the Ontario Court of Appeal held in light of Vavilov that it was “incumbent upon the Minister to make inquiries” about the point at which the Romanian authorities knew or ought to have known that the applicant was in Canada (at para. 29). As in Vavilov, both procedure and substance were considered together, holistically, to justify the conclusion that the decision should be struck down:

The delay between [1998] and the issuance of the summons on November 15, 2016 – more than 18 years – has not been properly investigated, nor properly explained. In the circumstances, the surrender order cannot stand. On the existing record, we are unable to determine whether the decision to order Ms. Boros’ surrender was reasonable. More information is required before we can properly conduct this analysis (at para. 30).

Sriskandarajah was not mentioned but it is entirely possible that it has simply been superceded by the blurring of the line between procedure and substance effected by the emphasis in Vavilov on responsiveness (compare the more restrained approach in Beaumann v. Canada (Minister of Justice), 2020 BCCA 124 but note that this matter had already been up and down the British Columbia court system on a number of previous occasions, so is perhaps to be limited to its special facts).

Note also Phelan J’s comments in Ennis v. Canada (Attorney General), 2020 FC 43, where he quashed a decision of the Canadian Human Rights Commission not to refer Ennis’s complaint for a tribunal hearing in the face of an investigator’s report recommending a hearing:

The Supreme Court’s analytical framework for reasonableness review reflects much of the work of this Court in this area. Therefore, this Court’s decisions relevant to the issues here are relevant and binding authority and have not been altered by Vavilov except to emphasise that reasonableness review is to be a vigorous review (at para. 20. See also Begum v. Canada (Citizenship and Immigration), 2020 FC 162 (ministerial refusal to approve a provincially nominated visa application); Salmonid Association of Eastern Newfoundland v. Her Majesty the Queen in Right of Newfoundland and Labrador, 2020 NLSC 34, at para. 64 (unjustifiable departure from previous decisions); Gomes v. Canada (Citizenship and Immigration), 2020 FC 506 (Refugee Appeal Division failed to engage with the reasons the Refugee Protection Division provided for refusing the applicant’s claim); Nation Rise Wind Farm Limited Partnership v. Minister of the Environment, Conservation and Parks, 2020 ONSC 2984 (Minister unreasonably raised new issues on an appeal which was supposed to involve simply a review of the record for error)).

The requirement of responsiveness does not necessarily mean that decision-makers will have to produce excessively detailed reasoned to live up to the requirements of Vavilovian reasonableness review. In Mao v. Canada (Citizenship and Immigration), 2020 FC 542, at para. 49, Favel J noted that “there is no need for a decision maker to engage with every argument—it is enough that they are alive and aware of them”; in Vavilov, the requirement to “meaningfully grapple” with an individual’s submissions applies only to those which are “key” (Vavilov, at para. 128). Moreover, as Slatter JA observed in Mohr v Strathcona (County), 2020 ABCA 187, at para. 40 (albeit in dissent and albeit in respect of a statutory duty to give reasons), “the length of the reasons is not determinative”. Thus in Edmonton (City of) v Edmonton Police Association, 2020 ABCA 182, five paragraphs of analysis by the arbitrator satisfied the reasonableness standard. The brevity of this critical section of the decision did “not necessarily mean that the arbitrator did not engage with the issues” (at para. 27).

There have also been a number of Federal Court cases in which decisions were struck down for unreasonableness because the decision-maker failed to grapple with relevant factors as established by prior judicial jurisprudence: Mora Alcca v. Canada (Citizenship and Immigration), 2020 FC 236, at para. 18; Demirtas v. Canada (Citizenship and Immigration), 2020 FC 302, at para. 30; Chikadze v. Canada (Citizenship and Immigration), 2020 FC 306, at para. 22; Lopez Bidart v. Canada (Citizenship and Immigration), 2020 FC 307, at para. 30 and Haile v. Canada (Citizenship and Immigration), 2020 FC 375, at paras. 25-26. Note that the unreasonableness here resulted from failures to seriously consider the factors at all: it might, in principle, be permissible for decision-makers to deviate from judicial decisions but obviously they bear a justificatory burden when they do so (Canada (Public Safety and Emergency Preparedness) v. Taino, 2020 FC 427, at para. 80). Consistency with prior judicial jurisprudence will, by contrast, indicate that a decision is reasonable: Cousineau c. Commission de protection du territoire agricole du Québec, 2020 QCCS 900.

Moreover, as Diner J sagely noted in Ortiz v. Canada (Citizenship and Immigration), 2020 FC 188, at para. 22, whereas under Dunsmuir reviewing courts began with the outcome and then looked back at the reasons, Vavilov instructs them “to start with the reasons, and assess whether they justify the outcome”. And in light of the decisions emphasizing the importance of responsiveness, I think it is too early to say categorically that “Vavilov does not constitute a significant change in the law of judicial review with respect to the review of the reasons of administrative tribunals” (Radzevicius, at para. 57, per Swinton J. See also Hildebrand v Penticton (City), 2020 BCSC 353, at para. 26).

It is worth mentioning, lastly, cases in which the decision was held to be unreasonable for want of internal coherence: see e.g. Canadian Pacific Railway Company c. Flynn, 2020 QCCS 983, at paras. 76-83.

Different considerations might apply, however, in the context of judicial review of legislative-type decisions for which contemporaneous reasons were not provided. In Vavilov the Supreme Court left the door open to focusing on the outcome of a decision-making process in situations where reasons are not provided. Accordingly, in 1120732 B.C. Ltd. v. Whistler (Resort Municipality), 2020 BCCA 101, at para. 88 Tysoe JA held that the enactment of a municipal by-law was reasonable on the basis there were “at least three ways in which the Municipality’s council could have reasonably concluded” it had the necessary statutory authority. This seems to me to be an accurate application of Vavilov but one which is jarring to read given the stark contrast it creates with the post-Vavilov jurisprudence emphasizing the importance of responsiveness and justification. Moreover, the application of reasonableness review in a context where municipalities already benefit from a broad and purposive approach to the interpretation of their jurisdiction has the potential to give these bodies a significant degree of regulatory authority, exercisable without detailed reasons (see also O’Shea/Oceanmount Community Association v Town of Gibsons, 2020 BCSC 698, at para. 156, commenting that “significant deference” was due to the municipality on the compatibility of a by-law with an official community plan). Contrast, however, the relatively intensive review undertaken in Minster Enterprises Ltd. v City of Richmond, 2020 BCSC 455, especially at paras. 114-118, where Crerar J rejected the suggestion that the City’s policies could expand the meaning of a bylaw relating to building construction; and in Canadian Natural Resources Limited v Elizabeth Métis Settlement, 2020 ABQB 210. Perhaps the best way to understand these contrasting decisions is that, in some instances, the governing statutory scheme (see Vavilov at paras. 108-110) will give municipalities a large margin of appreciation but in others municipalities will be more tightly constrained by prescriptive statutory language (see also Champag inc. c. Municipalité de Saint-Roch-de-Richelieu, 2020 QCCA 613, at para. 30, not a case about the authority to promulgate a bylaw but nonetheless an example of a municipality’s discretion being constrained by statutory language). Finally, it will not invariably be the case that reasons or reasoning are entirely absent in cases involving municipal by-laws; if so, the judicial review will look quite conventional (see e.g. G.S.R. Capital Group Inc. v The City of White Rock, 2020 BCSC 489, at paras. 107-114 and 139).

On the general issue of whether Vavilovian reasonableness review is more or less robust than the palpable and overriding error standard applicable to mixed questions of fact and law in statutory appeals, see Al-Ghamdi v College of Physicians and Surgeons of Alberta, 2020 ABCA 71, where the Court explained, at the outset, the wide scope of the issues subject to the palpable and overriding error standard:

In professional disciplinary appeals, interpretation of the governing statute is reviewed for correctness. Important questions of mixed fact and law calling for deference by a reviewing court will often include i) the standard of practice the profession expects in any particular case, and ii) whether, on the facts, the professional subjected to discipline has met that standard.

(at para. 11).

Although the College had “improperly relied” on some evidence and overemphasized the importance of a human rights complaint made by the appellant, “those errors were not sufficient to undermine the overall finding of professional misconduct” (at para. 51). Query whether these would have been considered to be problematic on reasonableness review (and, if so, whether the appellant would find it odd that a statutory appeal mechanism would be less favourable to him than an application for judicial review). In Mayer v The Superintendent Of Motor Vehicles, 2020 BCSC 474, at para. 47, Forth J was quite clear that “sufficiency of reasons is not a stand-alone basis for interfering with a decision on appeal” and observed, moreover, that a right of appeal does not necessarily require an “exacting review”: “Instead, it may be that the legislature wished to prescribe the parameters of review, to circumscribe the role of the reviewing court and to ensure that the decision is not unduly interfered with” (at para. 50). She held that there was sufficient information in the record to justify the impugned decision (at para. 53).

The Ontario Divisional Court has warned against conflating Vavilovian reasonableness review and the palpable and overriding error standard. I appreciate the conceptual distinction between the two, but I fear that these judges are rather like the Dutch youngster holding a finger in the dyke, as comparisons are inevitable given that the statutory appeal and judicial review streams run so close together (Miller v. College of Optometrists of Ontario, 2020 ONSC 2573, at para. 79; Houghton v. Association of Ontario Land Surveyors, 2020 ONSC 863 at para 15). In particular, as I observed in my paper on the Vavilov framework, if the palpable and overriding error standard on appeal is less generous to appellants than reasonableness review would be (and that is certainly my impression of the matter), there will inevitably be pressure to expand the scope of the palpable and overriding error standard. I certainly expect comparison to continue, with conflation a distinct possibility.

There is, albeit in the context of a judicial review application, a neat example of a palpable and overriding error in Syndicat des métallos, section locale 9449 c. Glencore Canada Corporation, 2020 QCCA 407, at paras. 43-44, namely, a failure to consider the proper comparators on a salary scale: this failure completely undermined the labour arbitrator’s decision (see also Sipekne’katik v. Alton Natural Gas Storage LP, 2020 NSSC 111, especially at para. 152, an appeal where failure to engage with a duty to consult issue amounted to a palpable and overriding error).

I emphasized in my commentary on Vavilov that there were important tensions in the majority reasons, for instance, in respect of statutory interpretation. Consider Canadian National Railway Company v. Richardson International Limited, 2020 FCA 20. The standard of review here was correctness, as the matter came before the Federal Court of Appeal as a statutory appeal from a decision of the Canadian Transportation Agency relating to railways. But Nadon JA also commented, in obiter, that he would have struck the decision down for unreasonableness in any event, “because it failed to consider both context and the legislative scheme as a whole” (at para. 46). Citing paragraph 118 of Vavilov — but not the more equivocal language of paragraphs 119 and 122 — Nadon JA commented that Agency’s failure to “observe the fundamental principles of statutory interpretation” (at para. 48) was “fatal to its decision” (at para. 49). This might be thought to betray a favouritism for an interventionist standard of reasonableness review on issues of statutory interpretation (although, to be fair, Nadon JA remitted the matter to the Agency and took pains not to “rule out the possibility that the Agency might come to an interpretation that differs from the one it arrived at in the present matter” (at para. 54)). A more moderate approach was taken by Boone J in Salmonid Association of Eastern Newfoundland v. Her Majesty the Queen in Right of Newfoundland and Labrador, 2020 NLSC 34, at para. 74, where the key flaw was that the Minister “did not explain his reasons for his adoption of an interpretation that he was aware was one of two valid but opposite readings” and by Barnes J in Glaxosmithkline Biologicals S.A. v. Canada (Health), 2020 FC 397, at paras. 34-35, noting that the Minister had failed to have regard to the obligation to interpret Canadian law implementing the Canada Europe Trade Agreement in conformity with the Agreement.

Nadon JA’s obiter comments certainly underscore how some portions of Vavilov are liable to become battlegrounds between different factions of judges, those who favour more intrusive review on questions of law in one camp, their more deferential colleagues in the other (for a similar approach, almost demanding panoptic qualities on the part of an administrative decision-maker, see Beals v. Nova Scotia (Attorney General), 2020 NSSC 60, at para. 32: “the legislature and applicants…are entitled to presume that the person making a decision about an application under [legislation] knows the occasion and necessity for the enactment, the circumstances existing at the time it was passed, the mischief to be remedied, and the object to be attained, without that information necessarily appearing in the record”).

Finally, out in British Columbia it looks like the Court of Appeal will soon have to weigh in on the implications of Vavilov for the province’s patent unreasonableness standard. In College of New Caledonia v Faculty Association of the College of New Caledonia, 2020 BCSC 384 the petitioner argued that the articulation of “robust” reasonableness review in Vavilov should inform the application of patent unreasonableness in British Columbia. Francis J was unimpressed (and went on to uphold the decision):

I agree with counsel for the Board that Vavilov has not changed the law with respect to the meaning of patent unreasonableness under s. 58 of the ATA, just as Dunsmuir v. New Brunswick, 2008 SCC 65 did not affect the meaning of the statutory standard of review (see Pacific Newspaper Group Inc. v. Communications, Energy and Paperworkers Union of Canada, Local2000, 2014 BCCA 496 at para. 44). There is simply nothing in the Vavilov decision that would lead me to conclude that the decision modifies the patent unreasonableness standard in any way (at para. 33).

But in Guevara v Louie, 2020 BCSC 380, Sewell J took the opposite view (and went on to strike down the decision):

In Canada (Minister of Citizenship and Immigration) v. Vavilov 2019 SCC 65 [Vavilov], the Court emphasized that it is the duty of a reviewing court to determine whether the decision maker’s reasons meaningfully account for the central issue and concerns raised by the parties. While these comments were made in the context of a review on a reasonableness standard, it is my view that they also apply to a review of reasons on the standard of patent unreasonableness. What constitutes a patently unreasonable decision may be “understood in the context of the common law jurisprudence” regarding judicial review generally “and will necessarily continue to be calibrated according to general principles of administrative law:” Canada (Minister of Citizenship and Immigration) v. Khosa, 2009 SCC 12 (S.C.C.) at para 19 (at para. 48).

I am sympathetic to Sewell J’s view. The patent unreasonableness standard in British Columbia essentially sets out grounds of review for abuse of discretion. These grounds can be supplemented — indeed, on the logic of Khosa, must be supplemented — by the common law of judicial review. The distinction between grounds and standards was reasserted in Vavilov, leaving the door very much open to Vavilovian reasonableness review informing the meaning of patent unreasonableness in British Columbia. Over to you, Court of Appeal.

In Ontario, meanwhile, the reference to patent unreasonableness in the Human Rights Code, RSO 1990, c h-19, s. 45.8 continues to be taken to require the application of the reasonableness standard: Intercounty Tennis Association v. Human Rights Tribunal of Ontario, 2020 ONSC 1632, at paras. 30-38. Inasmuch as the statute makes reference to a common law concept (patent unreasonableness) which has been assimilated by the courts to another common law concept (reasonableness) this conclusion seems to me to be irresistible. Unlike in British Columbia, the legislature has not attempted to specify the content of patent unreasonableness but rather signalled its willingness to rely on judicial development of patent unreasonableness. The judges have considered since Dunsmuir that the goals of patent unreasonableness can be achieved through the application of the reasonableness standard and, as such, the assimilation of patent unreasonableness to reasonableness does no violence to legislative intent or, to put the point in Vavilovian terms, the legislature’s institutional design choices.

In Ponoka Right to Farm Society v Ponoka (County), 2020 ABQB 273, Neilson J considered s. 539 of the Municipal Government Act, RSA 2000, c M-26, which provides: “No bylaw or resolution may be challenged on the ground that it is unreasonable”. Following the analysis in Koebisch v Rocky View County, 2019 ABQB 508, Neilson J held that this amounted to a legislative direction to apply the patent unreasonableness standard (at para. 13). Section 539 raises quite the conundrum. The idea that it requires the application of the common law patent unreasonableness standard runs into the objection that patent unreasonableness has now been assimilated to reasonableness as a matter of common law. Inasmuch as there is any legislative intent, it points as much to the reasonableness standard as it does to the patent unreasonableness standard. Of course, this suggests that s. 539’s bar to reasonableness review is meaningless. But I would not be too quick to jump to this conclusion. As the majority explained in Vavilov (at para. 110), some statutory language “contemplates that the decision maker is to have greater flexibility in interpreting the meaning of such language”. Section 539 can, accordingly, be understood to provide “greater flexibility” to municipalities. It is not necessary to breath new life into the patent unreasonableness standard. My approach to s. 539 also avoids a potential constitutional objection (noted by David Mullan in his Advocate’s Quarterly analysis of Vavilov): robust reasonableness review might be part of the constitutional core minimum of judicial review; if so, s. 539 might be unconstitutional. Of course, the majority in Vavilov is not at all clear on what’s entrenched and what’s not entrenched, and s. 539 is not a complete ouster (as it leaves correctness review in place), but allowing legislatures to oust reasonableness review would surely be constitutionally dubious at best in view of the importance accorded to “robust” reasonableness review in Vavilov.


Initially, judges did not seem overly anxious to jump on the suggestion that they might refuse to remit a matter consequent on a finding of unlawfulness where it is “evident” that a “particular outcome is inevitable” (Vavilov at para. 142). In Canadian Broadcasting Corporation v. Ferrier, 2019 ONCA 1025, for example, Sharpe JA remitted to the decision-maker the question of the applicability of the open court principle to police board hearings (here, a preliminary hearing on whether the time period for reporting alleged police misconduct should be extended). Sharpe JA remitted the question even though much of his analysis was conducted on a standard of correctness (at para. 37), he had little doubt that a recent Ontario Court of Appeal decision (Langenfeld) on the application of the open court principle to police board hearings was dispositive (at para. 58), and the judicial review proceedings had already slowed down a process which was moving quite slowly (at para. 79). On balance, Sharpe JA concluded, the decision-maker “should be permitted to take another look at the matter with the benefit” of the recent decision in Langenfeld (at para. 80). I think Sharpe JA was quite right to remit the matter, especially because the issues in Ferrier and Langenfeld arose in different contexts (see also Edmonton Police Service v Alberta (Information and Privacy Commissioner), 2020 ABQB 10, at para. 661; Alberta Union of Provincial Employees v Alberta Health Services, 2020 ABCA 4 , at para. 59; Langlais c. Collège des médecins du Québec, 2020 QCCA 134, at paras. 64-65; Mbula-Kolela v. Canada (Citizenship and Immigration), 2020 FC 260, at paras. 18-20; Rodriguez Martinez v. Canada (Citizenship and Immigration), 2020 FC 293, at paras. 18-20; McKenzie c. Ambroise, 2020 CF 340, at paras. 31-37; Salmonid Association of Eastern Newfoundland v. Her Majesty the Queen in Right of Newfoundland and Labrador, 2020 NLSC 34, at paras. 112-114; Romania v. Boros, 2020 ONCA 216, at paras. 31-32; United Steel v. Georgia-Pacific LP, 2020 ONSC 1560, at paras. 70-74). But see Guevara v Louie, 2020 BCSC 380, at paras. 85-87; Yu v City of Richmond, 2020 BCSC 454, at para. 38; and see also Coquitlam (City) v British Columbia (Assessor of Area #10 – North Fraser Region), 2020 BCSC 440 (though this was a statutory appeal so arguably the Vavilov principles on discretion do not apply).

More recently, however, the examples of judges taking a muscular approach to refusal to remit unreasonable decisions have multiplied. In Alexis v Alberta (Environment and Parks), 2020 ABCA 188, the majority of the Alberta Court of Appeal was convinced (in spite of a spirited dissent from one of their colleagues) on a thin factual record from which reasons were absent that the outcome of any remittal was preordained. Pentelechuk JA’s partially concurring and partially dissenting observations about the thinness of the record and inadequacy of the decision-making process should have given the majority pause (at paras. 188-192).

The issue in Nation Rise Wind Farm Limited Partnership v. Minister of the Environment, Conservation and Parks, 2020 ONSC 2984 was the Minister’s decision to cancel a major wind farm project on appeal from the Environmental Review Tribunal. The sole reason for the cancellation decision was the effect the project would have on the maternity colonies of bats. But the effect on bats had not been raised by any of the parties to the Tribunal decision or the appeal. It was unreasonable for the Minister to raise the effects sua sponte (at para. 91). The implication was that there would be “no utility” in sending the matter back to the Minister (at para. 159) because the Minister had made clear that the effect on bats was the “only basis” for revoking the permission for the project (at para. 161); and the Minister’s findings on this point were, moreover, unreasonable (at para. 160). There was also “evidence of urgency”, in the form of a “real risk” that the project would be cancelled if the matter were sent back for further redetermination by the Minister (at para. 163). For my part, I wish I could share the Divisional Court’s confidence in the inevitability of the outcome of a complex regulatory process. If the Minister were to do it all again, with the effect on bats out of the picture, would the appeal process have unfolded as it did? Even on the understanding of remedial discretion laid out in Vavilov, inevitability is a high bar and I am not sure it was reached here.

Finally, in Oberg v Saskatchewan (Board of Education of the South East Cornerstone School Division No. 209), 2020 SKQB 96, McCreary J quashed a decision removing the applicant as the principal of a high school and ordered that the applicant be reinstated: it was unreasonable for the Board to have demoted the applicant and the only other option open to the Board was the status quo. This is, I suppose, logical enough so far as it goes but when one considers how reluctant common law courts traditionally have been to make mandatory orders (especially in employment-related matters: Chief Constable of the North Wales Police v Evans [1982] 1 WLR 1155) it is a striking example of the potentially radical results Vavilov’s discussion of remedies might lead to.

This content has been updated on May 17, 2020 at 14:45.