The Implications for Administrative Tribunals of R. v. Sullivan, 2022 SCC 19

In R. v. Sullivan, 2022 SCC 19, the Supreme Court of Canada authoritatively addressed the effect of declarations of unconstitutionality. Along with colleagues from Torys LLP, I intervened on behalf of the British Columbia Civil Liberties Association to argue that declarations of unconstitutionality issued by superior courts mean that unconstitutional laws are of no force and effect nationwide.

The Court unanimously (and roundly) rejected this argument (developed at length in this article). The purpose of this post is not to criticize the careful and comprehensive reasons of Kasirer J, who has provided long-overdue clarity on the effect of declarations of unconstitutionality. Rather, my goal is to highlight the two key premises of his analysis, both of which have important implications for administrative tribunals and require, in my view, a rethinking of the relationship between administrative tribunals and superior courts.

First, Kasirer J locates the normative force of a declaration of unconstitutionality in the reasons a court gives for finding that a law is incompatible with the Constitution, rather than in the judgment the court issues on foot of those reasons (see especially at para. 63). It is a simple matter of stare decisis (at paras. 44, 68).

Whether this reflects the practice of Canadian courts (who routinely first determine whether a law is incompatible with the Constitution and then separately determine the appropriate remedy) I leave to others to debate. From the perspective of administrative tribunals, however, the collapsing of the distinction between reasons and judgment is significant.

The hallmark of the difference between a superior court and an administrative tribunal has been that a superior court can issue an authoritative declaration of unconstitutionality, effective against the world at large, whereas an administrative tribunal can only conclude that an unconstitutional provision cannot be applied in the context of a particular matter (Nova Scotia (Workers’ Compensation Board) v. Martin; Nova Scotia (Workers’ Compensation Board) v. Laseur, 2003 SCC 54, [2003] 2 SCR 504, at para. 31; R. v. Lloyd, 2016 SCC 13, [2016] 1 SCR 130, at para. 15). Put another way, superior courts have general authority to declare laws unconstitutional, but administrative tribunals may only decline to give effect to unconstitutional laws in specific cases.

This hallmark has been eliminated by Sullivan. Both superior courts and administrative tribunals can give reasons for concluding that a law is unconstitutional. The normative force of these reasons is a matter of stare decisis. Whatever is special about a superior court’s reasons relating to the unconstitutionality of a law comes from the doctrine of stare decisis.

The effect, I think, must be to enhance the status of administrative tribunals relative to the superior courts. True, an administrative tribunal’s reasons for finding a law to be unconstitutional will not bind a superior court, but they may be persuasive, especially where a tribunal demonstrates through its legal analysis that it has particular expertise in respect of the operation of the law in question (Nova Scotia (Workers’ Compensation Board) v. Martin; Nova Scotia (Workers’ Compensation Board) v. Laseur, 2003 SCC 54, [2003] 2 SCR 504, at para. 30). Indeed, as I will note again below, a superior court’s reasons have only persuasive force outside the province, further eroding the distinction between superior courts and administrative tribunals.

Moreover, administrative tribunals should now be able to depart from superior court reasons about the constitutionality of laws. In Canada (Minister of Citizenship and Immigration) v. Vavilov, 2019 SCC 65, at para. 112, the Supreme Court held that an administrative tribunal’s decision “may” (my emphasis) be unreasonable if the tribunal fails to “explain or justify” a departure from binding precedent (and, of course, only decisions of the superior court in the same province as the tribunal have this sort of precedential value). Accordingly, as Gleason JA has recently explained, “providing adequate explanations are given, an administrative decision-maker may sometimes decline to follow a decision from the courts, depending on the circumstances” (Canada (Attorney General) v. National Police Federation, 2022 FCA 80, at para. 49). A failure to follow precedent — even Supreme Court of Canada precedent (ibid., at para. 50) — is not necessarily unreasonable.

As such, a declaration of unconstitutionality by a superior court will not, in all cases, bind an administrative tribunal, which is free to chart its own course if it can explain and justify doing so.* This plainly enhances the status of administrative tribunals relative to superior courts.

Second, because the normative force of a declaration of unconstitutionality lies in the reasons for the declaration rather than the judgment issuing the declaration, a declaration has no binding force outside the province in which it is made (Sullivan, at para. 61).

It follows from this premise that an administrative tribunal in one province is not bound by the reasons of a superior court in another province: in such situations, there is no “binding precedent” (Vavilov, at para. 112) and the administrative tribunal is under no explanatory or justificatory burden in refusing to follow the reasons of the extra-provincial superior court.

It also follows from this premise that, as far as a superior court in another province is concerned, the reasons of an administrative tribunal and of an extra-provincial superior court are both entitled to persuasive force only. They stand on the same footing.

Again, the implication is that Sullivan has enhanced the status of administrative tribunals relative to superior courts.

My analysis of Kasirer J’s two premises leads to an observation about the standard of review. In Vavilov, the Court made clear that the standard of review of constitutional questions is correctness, on the basis that the rule of law requires uniformity as to matters of constitutionality (at least as the rule of law was defined in Vavilov, at para. 53: see the broader definition in Sullivan, at para. 64).

But in light of Sullivan it is harder to maintain that a standard of correctness is required to ensure uniformity. One of the key premises of Sullivan, after all, is that declarations of unconstitutionality do not lead to uniformity and, indeed, that a lack of uniformity is a feature of Canada’s federal system (Sullivan, at para. 62).

Moreover, if constitutional questions are all about stare decisis, and Vavilov empowers administrative tribunals to depart from binding precedent in some circumstances, there is reason to question the stability of Vavilov’s rule-of-law basis for correctness review on constitutional questions.

Now, I think that it is unlikely that there will be a post-Sullivan spate of administrative tribunals revisiting settled understandings of the constitutionality of various statutory provisions. The point is that Sullivan requires us to re-think the relationship between superior courts and administrative tribunals, as the Court — perhaps inadvertently — has enhanced the status of administrative tribunals.

* I think, pace Gleason JA, that the suggestions in Service d’administration P.C.R. Ltée v. Reyes, 2020 FC 659 can usefully guide the analysis, as long as they are not applied formalistically. See also Retraite Québec c. Tribunal administratif du Québec, 2020 QCCS 1592. I hope to write a separate blog on this point, as stare decisis may become a post-Vavilov battleground between judges who are more or less inclined to defer to administrative decision-makers.

This content has been updated on May 27, 2022 at 16:59.