The Return of Context? Society of Composers, Authors and Music Publishers of Canada v. Entertainment Software Association, 2022 SCC 30
Since its decision in Canada (Minister of Citizenship and Immigration) v. Vavilov, 2019 SCC 65,  4 SCR 653 in December 2019 (and a couple of companion decisions released the same month), the Supreme Court of Canada has done nothing in respect of the standard of review of administrative action. And, it must be said, it has done it exceptionally well.
This summer, the Court has ventured back into the territory it covered in Vavilov, in its decision in Society of Composers, Authors and Music Publishers of Canada v. Entertainment Software Association, 2022 SCC 30. Time will tell whether Entertainment Software Association is mostly “noise” about the Copyright Board of Canada or a “signal” about the expansion of correctness review and the return of contextual analysis. But there is little doubt that the majority’s analysis relies on context in order to permit more wide-ranging correctness review.
The underlying issue in this case was whether the Copyright Board had erred in determining that there is a separate ‘making available’ right in Canadian copyright law. The Board had found that a right to payment exists where a recording is ‘made available’ for download in an online repository (think iTunes). On judicial review, the Federal Court of Appeal found that the Board’s decision was unreasonable. On appeal, the Supreme Court agreed, the majority (per Rowe J) concluding that the Board was incorrect, the minority (per Karakatsanis J) applying the reasonableness standard and finding the Board’s decision unreasonable. There is much of interest in the decision (see e.g. Howard Knopf), including a discussion of international law, to which I will return. In this post, I focus on standard of review.
Writing for the majority, Rowe J applied the correctness standard. In a previous era of Canadian administrative law, the Court held that correctness applies to the review of interpretations of law by the Board: Rogers Communications Inc. v. Society of Composers, Authors and Music Publishers of Canada, 2012 SCC 35,  2 SCR 283. In obiter in the Federal Court of Appeal, Stratas JA argued that Rogers is consistent with Vavilov, teeing the issue up for the Court. Notably, however, none of the parties before the Court argued for reasonableness review.
This matter involved the interpretation by the Board of its home statute, the Copyright Act. The question of whether the Act protects a ‘making available’ right did not fall into any of the correctness categories in Vavilov, either on the “institutional design” branch or “rule of law” branch. As this was a judicial review, not an appeal, the appellate standards of review could not apply on the basis of “institutional design”. And as the question involved the interpretation of the Board’s home statute the “rule of law” was not engaged: it was not a constitutional question, a question of overlapping jurisdiction between administrative decision-makers or a question of general importance to the legal system.
But in Vavilov, the majority refused to “definitively foreclose” the recognition of additional correctness categories (Vavilov, at para. 70). Rowe J held that this was a “rare and exceptional” case where it was “appropriate to recognize a new category of correctness review”, namely “when courts and administrative bodies have concurrent first instance jurisdiction over a legal issue in a statute” (at para. 28). Questions of the interpretation of copyright law can arise on judicial review of decisions of the Board (at the Federal Court of Appeal) but also in the context of actions to enforce copyright heard in the Federal Court (with a right of appeal to the Federal Court of Appeal). This concurrent jurisdiction was enough, Rowe J held, to require the application of the appellate standards of review to decisions of the Board. This way, questions of copyright law will invariably be subject to correctness review in the Federal Court of Appeal regardless of whether the question is posed in a Board proceeding or an action in Federal Court in the first instance.
Rowe J provided several reasons why applying the correctness standard in such circumstances “accords with legislative intent and promotes the rule of law” (at para. 28).
In terms of legislative intent, Rowe J commented that Vavilov’s presumption of reasonableness review “no longer applies” (at para. 30) in a situation where “the legislature expressly involves the court in the administrative scheme”:
When the legislature enacts a statute that gives concurrent first instance jurisdiction to courts and administrative bodies, the legislature expressly involves the courts. Like legislated standards of review and statutory appeal mechanisms, concurrent first instance jurisdiction signals a legislative intent for judicial involvement. By conferring such jurisdiction, it should be inferred that the legislature wanted to subject those decisions to appellate standards of review (at para. 31).
Rowe J does not quote Vavilov at length and, as it happens, there are some inconvenient details in the relevant passages. Paragraph 24 of Vavilov is the one in which the presumption of reasonableness review is established:
Where a legislature has created an administrative decision maker for the specific purpose of administering a statutory scheme, it must be presumed that the legislature also intended that decision maker to be able to fulfill its mandate and interpret the law as applicable to all issues that come before it. Where a legislature has not explicitly prescribed that a court is to have a role in reviewing the decisions of that decision maker, it can safely be assumed that the legislature intended the administrative decision maker to function with a minimum of judicial interference.
In Entertainment Software Association, Rowe J adds two glosses to paragraph 24 of Vavilov: the grant of jurisdiction must be “exclusive” (a term which does not appear in Vavilov); and the legislature does not need to “explicitly prescribe that a court is to have a role in reviewing the decisions of that decision maker” (my emphasis), rather, it is enough to “expressly involve” the courts in the “administrative scheme” (at paras. 30, 31).
Furthermore, in Vavilov, the majority held that departures from the presumption of reasonableness review could be justified only where there is “clear” statutory language to that effect. But the term “clear” does not appear in the standard of review portion of Entertainment Software Association. Had Rowe J applied a “clarity” test, it is doubtful that it would have been satisfied here, as Parliament did not express any view on standard of review in the Act or elsewhere, still less a “clear” view. As Karakatsanis J rightly pointed out in her minority reasons:
[T]he Copyright Act,R.S.C. 1985, c. C-42, gives no clear signal of legislative intent regarding the applicable standard of review. Vavilov states that the legislature can provide this signal in two ways: it can explicitly prescribe the standard, or it can provide a statutory appeal mechanism (para. 33). Concurrent jurisdiction between a tribunal and a court was before the Court in Vavilov but was not identified as an analogous signal of legislative intent requiring correctness review. Here, there is no legislated standard of review, statutory appeal mechanism, nor any other analogous provisions. Legislative choice “pulls in two directions” — the absence of these “clear signals” is telling (Vavilov,at paras. 39 and 49).It follows that a sixth category cannot be established on the basis of Parliament’s intent (at para. 131).
I would go further. In Vavilov, the majority made clear that statutory provisions which “address procedural or other similar aspects of judicial review in a particular context…do not authorize the application of appellate standards” (at para. 51). In other words, oblique references to the role of the courts are not enough to require the use of appellate standards: “appeal” is the magic word. If this is so, it is jarring that the creation of concurrent jurisdiction, which has nothing to do with judicial review, can be construed as legislative intent to require the application of the appellate standards of review.
The strength of Vavilov, as far as the selection of the standard of review was concerned, lay in the establishment of simple rules for identifying the applicable standard without reference to context: “appeal” or a legislated standard of review were the only means of signalling the application of appellate standards of review. Entertainment Software Association departs from Vavilov in several subtle respects and undermines the simplicity of the Vavilov framework by inviting courts to take a much more holistic view of legislative intent.
The Rule of Law
Rowe J was concerned that deferential review of the Board’s interpretations of law could create legal inconsistencies which would be “antithetical” to the rule of law: it would potentially subject the same issue to a different standard of review depending on the forum deciding the issue in the first instance, which could lead to “conflicting statutory interpretations” (at paras. 33-34) (e.g. reasonableness in the Federal Court of Appeal on an interpretive issue arising before the Board but correctness in the same court on any legal issue decided by the Federal Court).
Moreover, Rowe J held, this scenario is different in rule of law terms from discord within an administrative agency. In Vavilov, the majority refused to recognize “persistent discord” as justifying correctness review. But for Rowe J the three rationales offered for this reticence in Vavilov lose their gravitational force where the “discord” is between different decision-makers.
First, “when there is concurrent first instance jurisdiction, the legislature has expressly involved the courts in the interpretation of a statute” (at para. 37). This is a curious response to include under the heading of the “rule of law”, as it makes an observation about legislative intent. The objections I offered above apply again here. I would note in addition that it is telling that Rowe J here refers to paragraph 37 of Vavilov, which is explicitly about statutory appeals, not concurrent jurisdiction or oblique indicators of legislative intent.
Second, this correctness category can be “defined with precision” and will apply in “rare” circumstances (at para. 39). Here I recall the objection I made to the Rogers decision a decade ago:
[C]lever counsel will undoubtedly argue before provincial and federal courts that the existence of a private right of action negates the deference owed to interpretations of a home statute. After all, in an action, a provincial superior court or the Trial Division of the Federal Court will be called upon to answer any questions of law arising, potentially the same questions addressed by the tribunal in its regulatory functions. As with the Copyright Act, the Competition Act creates a private right of action which may overlap with the jurisdiction of the Competition Tribunal. Securities legislation in the provinces often does the same. Doubtless, counsel for the applicant in judicial review cases will closely parse the statute book in order to find “shared jurisdiction”. Really creative counsel might even argue that public law duties which can ground actions against public bodies in negligence also create “shared jurisdiction”. Working all of this out will be complicated.
Now that statutory appeals (e.g. in competition and securities matters) require correctness review on questions of law, the potential for some of this mischief has been reduced. Nonetheless, this issue was litigated before Vavilov and in the few weeks since Entertainment Software Association, has already been raised, in AbbVie Corporation v. Canada (Health), 2022 FC 1209, as potentially reversing settled law; see also the ‘seductive’ argument made in Ville de Brossard c. Ville de Longueuil, 2022 QCCA 1139. Just because the Court says the new category applies in “rare” circumstances does not mean counsel and lower courts will agree.
Third, reasonableness review cannot safeguard the rule of law. The majority in “Vavilov offered guidance as to how to manage persistent discord within administrative bodies” but aside from a reference in paragraph 112 of Vavilov gave “little guidance on managing differing decisions between courts and tribunals” (at para. 38). I agree that more could have been said in Vavilov about stare decisis as between courts and tribunals, but the lower courts have attempted to establish some general principles about the proposition that “a binding and material precedent will place a limit on what is reasonable in the circumstances” (at para. 128, per Karakatsanis J; see e.g. Canada (Attorney General) v. National Police Federation, 2022 FCA 80). And as a jurist with considerable expertise in relation to stare decisis, Rowe J could have offered some thoughts in Entertainment Software Association! In any event, I offer the same objection to this point as I offered to Stratas JA’s discussion in the court below of the standard of review of decisions of the Board:
In terms of consistency and certainty, the mere possibility of divergent approaches within an administrative decision-making structure to a particular question does not, on its own, require correctness review. Rather, where there is a departure from a previous decision or prior practice, the administrative decision-maker must justify such a departure. Translated into the Copyright Board-Federal Court-Federal Court of Appeal conundrum, the Vavilovian answer is that the Copyright Board may indeed depart from Federal Court jurisprudence, but would bear a heavy justificatory burden in doing so (with the Federal Court of Appeal charged with determining if the burden has been discharged); and, meanwhile, the Federal Court (and Federal Court of Appeal) would have to be mindful in its jurisprudence of the Copyright Board’s preferred approach.
Whereas Vavilov was concerned with inconsistencies which would undermine the integrity of the legal system, Entertainment Software Association is concerned with the potential for inconsistency. There is no actual inconsistency (or any example of which I’m aware) between the courts and the Board on the interpretation of copyright law. If potential inconsistency and the resultant risk of discord is the trigger for rule of law concerns and, thus, the application of the correctness standard, the door to correctness review is more open than I thought it was. Admittedly, however, my emphasis since Vavilov (e.g. here) on the need for judicially imposed uniformity to be required in order to trigger correctness review might have been misplaced (see also my post on Sullivan).
Did Vavilov Foreclose the Recognition of this Correctness Category?
There is one other aspect of the debate between the majority and minority which is worth noting. Karakatsanis J argued that concurrent jurisdiction could not be recognized as a correctness category:
The establishment of new categories was explicitly reserved for possible circumstances the Court could not realistically foresee. It is “unrealistic” to suggest that the Court’s own precedent represents a “possible set of circumstances” that it could not “have contemplated” (at para. 124).
Indeed, the majority in Vavilov was able to be quite clear when it was excluding issues from the new framework: note the treatment at paragraph 55 of Doré v. Barreau du Québec, 2012 SCC 12,  1 SCR 395, for example.
That said, it is also clear from Vavilov that the compatibility of any precedent with the Vavilov framework falls to be assessed by reference to, well, the Vavilov framework (Vavilov, at paras. 143-145). Hence Rowe J’s rejoinder:
When this Court wanted to reject the possibility of a certain correctness category, it did so expressly: see Vavilov, at paras. 71-72. Concurrent first instance jurisdiction was not discussed in Vavilov. Accordingly, I seek to give effect to Vavilov by considering whether treating concurrent first instance jurisdiction as a new correctness category furthers the framework and principles in Vavilov (at para. 42).
I do agree with Karakatsanis J that it is a little strange to recognize a new correctness category so soon after the “exhaustive” analysis of the standard of review in Vavilov (at para. 122). Nonetheless, the analysis in Entertainment Software Association should be judged, as Vavilov itself suggests, by reference to the majority reasons in Vavilov and its likely future consequences. I do not think that the recognition of correctness categories should be influenced, one way or another, by what the judges in the majority in Vavilov might have had in mind when they signed off on the reasons.
As the reader will readily surmise, I do not agree with the application of the correctness standard to decisions of the Board. Yet, truth be told, I do not find the outcome especially surprising. Many copyright lawyers, on and off the record, are sceptical of the Board’s ability to engage in statutory interpretation. Outside the Board, therefore, this decision is likely to be welcomed. For my part, I have been banging this particular deference drum for a decade, to no avail, so perhaps my remarks on Rowe J’s analysis can safely be dismissed as the latest cavillings of a inveterate crank.
Whatever about the Board, however, I am less sanguine about the approach taken to reach the correctness review conclusion in this case. As I have outlined above, the modifications to the Vavilov framework are subtle, but important. In terms of institutional design, a wider range of factors than those set out in Vavilov may be taken into account in determining the role “legislative intent” indicates for the courts in judicial review cases. As far as the rule of law is concerned, a more hands-on judicial approach is justifiable whenever there is potential for inconsistent application of legal norms: correctness review is not limited to situations where a uniform answer is required to safeguard the integrity of the legal system.
Moreover, although Entertainment Software Association is a case about the recognition of an additional correctness category, the subtle but important modifications to “institutional design” and the “rule of law” apply with just as much force to the scope of the existing correctness categories. These principles underpinned the Vavilov framework for the selection of the standard of review. It follows that the scope of the correctness categories should be consistent with the principles underlying them. For example, the category of “overlapping jurisdiction” could be expanded to cover situations where “one administrative body has interpreted the scope of its authority in a manner that is [potentially] incompatible with the jurisdiction of another” (Vavilov, at para. 64, my addition; see also here). References to the “proper functioning of the justice system” (Vavilov, at para. 59) in respect of the category of questions of law of general importance to the legal system might have to be understood to encompass the deleterious effects of potential inconsistency: this could particularly arise in respect of the interpretation and application of common law norms by administrative decision-makers. And I am much less confident than I was before that Doré v. Barreau du Québec, 2012 SCC 12,  1 SCR 395 remains good law: a more holistic view of legislative intent and a broader view of the need for judicial intervention to guard against potential inconsistency in the treatment of legal norms could conceivably justify an expansion of the “constitutional questions” category to include the compatibility of an administrative decision with the Charter of Rights and Freedoms. Hence the force of Karakatsanis J’s charge that the majority’s approach will “open the door to endless litigation concerning possible exceptions to the reasonableness presumption” and “erode the presumption of reasonableness in all standard of review cases going forward”, which is “precisely what Vavilov aimed to avoid” (at para. 135).
These thoughts are tentative, of course. As I mentioned at the outset, only time will tell whether Entertainment Software Association is “signal” or “noise”. Here is another tentative thought. Taken with Law Society of Saskatchewan v. Abrametz, 2022 SCC 29, Entertainment Software Association indicates that legislative intent has become the meta-principle of the Vavilov framework: the use of the magic word “appeal” in Abrametz justified the application of the Vavilov framework even though in Vavilov itself the majority explicitly excluded procedural fairness issues from its purview; and Rowe J’s justification for correctness review in Entertainment Software Association relied heavily on what Parliament might have intended by creating concurrent jurisdiction. As a matter of first principles, I would be happy for legislative intent to play a key role in the selection of the standard of review: I laid out my argument to that effect in A Theory of Deference in Administrative Law: Basis, Application and Scope (2012). But as a matter of pragmatically implementing the Vavilov framework, which is a compromise designed to achieve an “overlapping consensus” amongst lawyers who have different views of the applicable first principles, I am much less convinced. As I argued in A Theory of Deference, respect for legislative intent involves a consideration not just of the extent of the authority delegated by a given statute but also reference to the “practical justifications” for delegation, including expertise. In Vavilov, expertise and other contextual factors were shunted out of the framework in order to promote simplicity. But if the turn to legislative intent proves decisive, the pressure to take account of context once more will mount significantly. Again, time will tell. Perhaps Entertainment Software Association will prove to be a ticket good for the Copyright Board of Canada only and not provide a general licence to consider context and expand Vavilov’s correctness categories.
For more on the Supreme Court’s standard of review summer, see Mark Mancini here (and the newsletter links therein).
This content has been updated on September 9, 2022 at 17:28.