International Law Blog Essay

Discretion to Declare: The Notwithstanding Clause and the Democratic Function of the Canadian Courts

I.  Introduction

The Notwithstanding Clause of the Canadian Charter of Rights and Freedoms is a constitutional innovation (albeit colored by historical statutory analogues) first proposed by Professor Paul C. Weiler in 1980, during a fierce federal-provincial negotiations process that would ultimately produce the Constitution Act, 1982. After the Supreme Court of Canada concluded in 1981 that the federal government would need the support of more than two provinces to impose constitutional patriation, the Notwithstanding Clause became an essential element of interprovincial compromise on the terms of an entrenched Charter.

Section 33 of the Charter, also known as the Notwithstanding Clause, states that the Canadian Parliament or provincial legislatures “may expressly declare” in a statute that “the Act or a provision thereof shall operate notwithstanding” one or more provisions of the Charter. These legislative declarations may override individual rights conferred by Sections 2 and/or 7 through 15 of the Charter, but not other provisions protecting democratic, mobility, and language rights. Critically, Section 33(2) explains that a legislative act involving such an express declaration “shall have such operation as it would have but for the [Charter] provision . . . referred to in the declaration.” Declarations remain in effect for a renewable period of five years.

In the decades following its enactment, the Notwithstanding Clause was rarely invoked by legislatures apart from Quebec’s National Assembly—and to date, it has never been invoked by the federal legislature. But in recent years, its popularity has grown. Since 2020, provincial governments have exercised their powers under the Notwithstanding Clause seven times1 to address legislative matters ranging from election spending to labor disputes with teachers. Critics of the Clause point to its relationships with provincial populist movements, report diminishing political consequences for repeated invocations, and cite attempts by lawmakers to “short-circuit constitutional protections afforded to vulnerable groups.” Recent scholarship points to shifting partisan dynamics between provincial conservatives and the liberal federal government as one factor underlying the Notwithstanding Clause’s recent and ongoing revival.

Two Notwithstanding Clause cases recently percolated from provincial appellate courts to the Supreme Court of Canada, each presenting the same question: How does Section 33(2) affect superior and appellate courts’ authority to issue judgments concerning the constitutionality of statutes invoking the Notwithstanding Clause? The Supreme Court’s decision could affirm that Canadian courts have the power to assess whether any act invoking the Notwithstanding Clause does in fact violate the constitutional provisions it disclaims, or bar those courts from doing so.

This Essay outlines at Part II the first of these appellate decisions, which originated in the provincial courts of Quebec and resulted in a finding against jurisdiction to hear constitutional claims covered by an invocation of the Notwithstanding Clause. At Part III, it outlines the parallel case arising in the provincial courts of Saskatchewan, where the Court of Appeal reached the opposite conclusion. And at Part IV, it argues that the latter decision made the more persuasive argument. By issuing declarations as to whether statutes would violate specific constitutional provisions but for the invocation of the Notwithstanding Clause, Canadian courts facilitate a critical dialogue that is both structurally necessary for the functionality of the Clause and also exemplified in practice by the English approach to weak judicial review of statutes against a national constitution.2

II. The Court of Appeal of Quebec Concludes that Its Hands Are Tied

In English Montreal School Board v. Attorney General of Quebec, the Supreme Court of Canada granted the English Montreal School Board’s motion for leave to appeal from the Court of Appeal of Quebec’s decision that Section 33 prohibited it from deciding whether a Quebec statute complied with various Charter provisions.  

In 2021, several individuals and organizations challenged the constitutionality of Quebec’s Act Respecting the Laicity of the State (the Laicity Act), which prohibited certain public employees from wearing religious symbols, including face coverings like niqabs, while executing their duties. The Quebec legislature had invoked Section 33 to render the Laicity Act operational notwithstanding Sections 2 and 7 through 15 of the Charter and had also invoked an analogous provision in the Quebec Charter to circumvent various rights conferred therein. The challengers argued that the Laicity Act impermissibly violated Sections 2 and 15, as well as various Charter and Constitution Act, 1867 provisions beyond the reach of Section 33.

The Superior Court of Quebec concluded that it had jurisdiction to hear the claims covered by the government’s Section 33 invocation. The court cited Section 24(1) of the Charter, which grants “court[s] of competent jurisdiction,” including provincial superior courts, broad discretion to supply remedies for constitutional deprivations as “appropriate and just in the circumstances.”  Declaratory relief is among the remedies available thereunder. Notably, Section 24(1) coexists with Section 52(1) of the Constitution Act, 1982—also known as the Supremacy Clause—which provides that “any law . . . inconsistent with the provisions of the Constitution is, to the extent of the inconsistency, of no force or effect.”  Notably, even where a court does not strike down a law under Section 52(1), “a remedy under [Section 24(1)] may nonetheless be available.” The Supreme Court of Canada has historically engaged Section 24(1) where an unconstitutional state action, as opposed to a law, infringes a litigant’s Charter rights. 

Having held that it had discretion to hear the relevant claims, the Superior Court exercised that discretion by declining to hear them. In doing so, it noted that there was rigorous debate between the parties on the constitutional issues, that a declaratory remedy would have no effect on the operability of the challenged law, and that courts should exercise restraint when addressing purely theoretical questions of law. 

On the government’s appeal from the trial court’s conclusions that the Laicity Act violated other constitutional provisions not excepted under Section 33, the Court of Appeal of Quebec reached a different holding as to the trial court’s jurisdiction under Section 33. The court held that Section 33 not only protected the statute from the application of the Charter provisions cited, but “also thereby limit[ed] the judicial review of the statute’s constitutionality.”  

The Court of Appeal Justices reasoned, as a functionalist matter, that permitting declaratory relief against a statute that nonetheless continued to operate would put the province in the unacceptable position of having to either “paradoxically” defend the constitutionality of its enactment or agree with the plaintiffs and concede unconstitutionality. And as a structural matter, the court argued that the proper route for “determining the correctness of the legislature’s political and legal choice in invoking [Section] 33” was “through the tools of parliamentary democracy” rather than the judiciary. The Court of Appeal also held, in the alternative, that the constitutional claims covered by Section 33 were moot because declaratory relief would have no “practical and tangible consequences for the rights of the parties.” The court determined that the trial court had no ability to hear the case despite its mootness, in part because the government had defeated the adversarial nature of the proceeding by merely “halfway debat[ing]” the constitutional infringement claim.

It is important to note that the Quebec government’s French civil law origins characterize, and potentially distinguish, its approach. The federal Interpretation Act specifies that, “when an enactment contains both civil law and common law terminology, or terminology that has a different meaning in the civil law and the common law,” the province of Quebec will adopt the civil law terminology or meaning and the other provinces will adopt the common law version. But if this interpretive principle is animating the appellate court split explored in this Essay, the Quebec courts involved do not state as much. Nor has the Supreme Court of Canada explored such a divide in prior cases addressing other questions of Section 33’s interpretation.

III. The Court of Appeal for Saskatchewan Finds Authority to Issue Declaratory Relief

Ten months later, in Government of Saskatchewan v. UR Pride Centre for Sexuality and Gender Diversity, the Supreme Court of Canada granted the Saskatchewan provincial government’s motion for leave to appeal from an appellate decision asserting judicial authority to declare its Section 33 statute unconstitutional.

UR Pride Centre for Sexuality and Gender Diversity challenged a provincial policy requiring schools to obtain parental consent if a student less than sixteen years old asked “to use a preferred name, gender identity, or gender expression.” UR Pride argued that the policy violated two Charter provisions: Section 7, which preserves an individual “right to life, liberty and security of the person,” and Section 15(1), which guarantees “equal protection and equal benefit of the law.” During the course of litigation, the Saskatchewan provincial government ratified the school gender identity policy in a statute that invoked Section 33 to shield the act from unconstitutionality under Sections 7 and 15(1) of the federal Charter. In response, UR Pride amended its complaint to supply a Section 12 claim alleging “cruel and unusual treatment.”

The Saskatchewan government asked the Saskatchewan Court of King’s Bench to dismiss, at a minimum, UR Pride’s Section 7 and 15(1) claims. The government argued that its invocation of the Notwithstanding Clause either stripped the court of jurisdiction to hear the claims or rendered those claims moot. The trial court declined, holding that it did have jurisdiction to hear claims concerning the provisions covered by the government’s Notwithstanding Clause invocation, and deferring the question of mootness until later in the proceedings.

The government appealed—unsuccessfully—the trial court’s finding that it had jurisdiction to hear the case. On appeal, the Court of Appeal for Saskatchewan affirmed, holding that the Court of King’s Bench had discretion, if it concluded that the government’s law did not comply with the constitutional provisions it purported to override, to declare as much. While patriation debate records evinced an intention to give legislators the “last” or “final” word on whether a law could operate, the Court of Appeal found no evidence within legislative history or other interpretive sources of “any intention to . . . provide[] legislators the only word on issues of Charter application.”

Both the lower and appellate courts relied in part on the English and French text of Section 33, which does not explicitly strip courts of their authority to adjudicate constitutional challenges to Notwithstanding Clause legislation. In fact—and as the Court of Appeal argued—Section 33 makes the statute at issue, rather than judges or other Charter provisions, the subject of its text. Section 33 specifies that the statute will continue to operate notwithstanding the specified provisions, without stating that those provisions have no effect or that courts have no power to comment on their effect. The Saskatchewan courts were likewise united in asserting a valuable democratic purpose to be served by declaratory relief in the Notwithstanding Clause context—a declaratory remedy, the Justices argued, serves an essential “oversight” function and enables the citizenry to make informed democratic choices about how they’ll respond to a legislature’s invocation of the Notwithstanding Clause.

The Court of Appeal also determined that the case was not moot, on the grounds that: (1) as a matter of constitutional law, the Section 33 shield would fall away after five years; (2) the decision would have precedential value for other cases not involving Section 33; and (3) declaratory relief served an informational function for both the legislature and its citizenry.  In the alternative, the Court of Appeal held that the lower court was entitled to hear the case despite its mootness, under the Supreme Court’s three-prong test in Borowski.

IV. The Supreme Court of Canada Should Preserve the Judiciary’s Declaratory Role

Both of these Section 33 cases await Supreme Court adjudication. While the Court has commented in dicta on the application of Section 33, it has never reached the question of whether Section 33 forecloses declaratory relief as to the merits of a claim arising under the Charter provisions it excepts. The parallel resolution of English Montreal School Board and UR Pride stands to answer a novel question of constitutional interpretation that could lend substantial color to the Court’s perspective on judicial review for constitutionality as well as the separation of federal legislative, executive, and judicial powers. 

Amid provincial courts’ competing attempts to answer this pivotal question, the majority opinion from the Court of Appeal for Saskatchewan gets the better of the jurisdictional debate: The constitutionality of legislation invoking the Notwithstanding Clause should be justiciable at the discretion of courts competent to enforce Section 24(1) of the Charter. This is true for six reasons.

First, and as both the Saskatchewan and Quebec courts acknowledged, the text of Section 33 doesn’t explicitly strip superior courts’ existing jurisdiction to hear constitutional challenges. Even if it did so, arguments against discretionary judicial review elide the Supreme Court of Canada’s longstanding instruction that “[t]here is no hierarchy amongst constitutional provisions,” and therefore all of those provisions “must be read together.” In reading Section 33 to subvert discretionary jurisdiction to supply declaratory relief under Section 24(1), the Court of Appeal of Quebec did not express any attempt to achieve a “balance . . . that fully respects the importance of both [provisions],” as the Supreme Court prescribes. Instead, the Court of Appeal held that Section 33 forbids declaratory relief in all circumstances, nullifying not only Section 52(1) but also Section 24(1) anytime a legislature invokes the Notwithstanding Clause.

Nullifying Section 24(1) was far from the only analytical path available to the Court of Appeal.  Where Charter provisions are in competition, the Supreme Court has endeavored to “reconcile[]” those provisions by “circumscrib[ing]” their interpretive “scope.” Indeed, Sections 24(1) and 33 are reconcilable, as the Court of Appeal for Saskatchewan suggested: Where the statute facing judicial review contains a successful invocation of Section 33, the reviewing court’s discretion under Section 24(1) can be limited to a declaration that the statute, while operable notwithstanding the Charter provisions it specifies, violates those specified provisions. The scopes of Sections 24(1) and 52(1) are thus circumscribed in that the range of remedies available to courts is narrowed to ensure the law at issue enjoys full force and effect, while the scope of Section 33 is circumscribed in that the invoking legislature obtains full “operation,” but not the guaranteed silence of the judicial branch.   

Second, a judicial declaration that a statute violates the federal constitution enhances democratic processes, even if that statute remains in force. If the judiciary falls silent on the constitutionality of Notwithstanding Clause legislation, democratic checks on usage of the clause will be disrupted in two ways. First, the citizenry will lack information about whether a given measure complies with specific provisions in the federal constitution, diminishing its capacity to oppose controversial legislative measures. Second, legislatures will have an incentive to apply the Notwithstanding Clause generously, shielding volumes of statutory law from the judicial review that defines superior courts’ core jurisdiction under Section 96 of the Constitution Act, 1867. Both disruptions would foreclose critical pathways for combatting legislation that burdens minority interests and rights.

Professor Eric Adams and Erin Bower, after summarizing the fulsome history of the Notwithstanding Clause’s generation and enactment, argue convincingly that judicial review satisfies the provincial compromise underlying Section 33 by generating benefits for the invoking legislature as well as for the citizenry to which it is accountable. If a competent court finds that a law would have been invalid but for the invocation of Section 33, a declaration to that effect “provides crucial information for both voters and governments alike as they contemplate their democratic choices during the five-year span that the notwithstanding clause operates.” And if the court declares that the law is constitutionally valid, such that it was unnecessary to invoke Section 33, the legislature is free to “let the sun set” on the invocation “without having to pay the ongoing political cost for a deliberate infringement of Charter rights.”

Third, and on a related note, the Court of Appeal of Quebec doesn’t address an unacceptable consequence of its position: If Section 33 forecloses judicial review, dialogue between the judicial and legislative branches collapses unidirectionally, as an arbitrary matter of timing. The court cited, among other Supreme Court dicta concerning Section 33, the Supreme Court of Canada’s statement in Vriend v. Alberta that the interbranch “dialogue” created by judicial review (whereby the court speaks through its opinions and the legislative branch optionally responds by invoking Section 33) “ha[s] the effect of enhancing the democratic process, not denying it.” And Professor Paul C. Weiler, the architect of the Notwithstanding Clause, argued that its existencewas “justified if one believe[d] . . . that on those exceptional occasions when the court has struck down a law as contravening the Charter and Parliament re-enacts it, confident of general public support for this action, it is more likely the legislators are right on the merits than were the judges.” Weiler therefore conceived of the Clause as facilitating a legislative reply to a judicial ruling, such that both branches had an opportunity to speak.

Under the Quebec approach, if courts reach a constitutional question first, then dialogue between the branches can exist because the legislature is free to respond. But if the legislature reaches the question first by invoking Section 33 (perhaps even days before the judiciary speaks), there is no opportunity for dialogue because the legislature has secured not only the “final word,” but the first and “only word”—a distinction also highlighted by the Court of Appeal for Saskatchewan. Thus, the existence of interbranch constitutional dialogue assumes and requires the possibility of judicial review after invocation of Section 33.

Fourth, looking to English law for comparative insight reveals that declarations of constitutional incompatibility—as a weak form of judicial review—are a workable alternative to invalidating statutes by judicial decree. Dissenting from the Court of Appeal for Saskatchewan’s holding, Justice Caldwell argued that the Notwithstanding Clause “reaches back to and restores the historic Diceyan-hierarchy of a parliamentary democracy in specific circumstances for a limited time.” He was referring to British constitutional theorist Albert Venn Dicey, who in the 19th century “identified parliamentary sovereignty as the fundamental norm of the British Constitution.”

But declaratory relief is entirely consistent with the British model of government, characterized by parliamentary sovereignty and weak-form judicial review, that Justice Caldwell argues is the proper structural reference for the function of the Notwithstanding Clause. The U.K. Human Rights Act provides that courts must, “[s]o far as it is possible to do so,” construe legislation “in a way which is compatible with the [European Convention on Human Rights].” But if an avoidant construction is impossible and “the court is satisfied that [a] provision [of primary legislation] is incompatible with a Convention right,” the court “may make a declaration of that incompatibility.” Professor Mark Tushnet explains that this provision was implemented to dilute the “strong medicine” of judicial review “in a nation where the tradition of parliamentary supremacy had deep roots.” The “declaration of incompatibility” model provides a persuasive framework for understanding how Sections 33 and 24(1) can be read together, and “harmoniously with the scheme of the [Charter], the object of the [Charter], and the intention of Parliament.”

Fifth, the Court of Appeal of Quebec’s holding rests in part upon the Supreme Court of Canada’s decision in Ford v. Quebec. But Ford did not explore the question of declaratory relief under Section 24(1). The Court held that it had no power to engage in “substantive review of the legislative policy in exercising the override authority in a particular case”—once a legislature had satisfied the procedural requirements of Section 33 by invoking it within a statute and specifying the Charter provision(s) the legislature wished to override, courts weren’t at liberty to invalidate that invocation as ultra vires

Ford, a case explaining how legislatures can properly invoke Section 33, therefore does not stand for the proposition that a proper invocation then forecloses judicial review under Section 24(1). And to the extent that Ford carries any implications beyond its procedural holding, it cuts against the Court of Appeal of Quebec’s decision. This is because the Court held in Ford that the Quebec legislature could not apply Section 33 retroactively. In that case, the legislature drafted a statute purporting to apply Section 33 from a date before the effective date of the invoking statute. The Court found Section 33 to be ambiguous on the question of whether legislatures could invoke it retroactively and chose the interpretation barring this usage, citing a common law interpretive preference for avoiding retroactivity. In doing so, the Court clarified that its interpretive voice is not entirely silenced at the moment of Section 33 invocation—while courts cannot chastise legislatures for overriding too many Charter provisions at once, they can comment on the relationship between the invoking statute and other relevant principles.

Sixth, and on the subject of mootness, cross-border precedent undermines the Court of Appeal of Quebec’s argument that the government’s complicated positionality on a Notwithstanding Clause challenge is a reason to deny jurisdiction in such cases. In United States v. Windsor, the U.S. Supreme Court considered and rejected an argument that the federal government’s failure to defend the constitutionality of the Defense of Marriage Act defeated the justiciability of a challenge to that statute. The Court noted that “adversarial presentation of the issues [wa]s assured by the participation of amici curiae [and an intervening party] prepared to defend with vigor the constitutionality of the legislative act.” The breadth of advocate intervention in the Section 33 cases at issue here encourages a similar conclusion.

Furthermore, the political importance of the interbranch dialogue facilitated by Sections 24(1) and 33 should not be underestimated, for the purposes of the mootness analysis. The Court of Appeal of Quebec cited a lack of “practical and tangible consequences for the rights of the parties” stemming from declaratory relief without legislative invalidation. But weak judicial review should not be confused for ineffective judicial review. The Supreme Court of Canada noted in Association des parents de l’école Rose‑des‑vents v. British Columbia that “there is a tradition in Canada of state actors taking Charter declarations seriously.” A declarative judgment on the merits of an underlying constitutional dispute produces a number of practical and tangible consequences for the government and its challengers: It facilitates political accountability by clarifying the nature of the legislature’s decision to invoke Section 33, avoids unnecessarily delays that would be imposed by requiring the parties to postpone litigation until the advent of the five-year sunset clause, and develops useful constitutional precedents.

V. Conclusion

There are reasons to be skeptical of the idea that the Notwithstanding Clause gives provincial legislatures and the federal Parliament power not only to override constitutional rights, but also to silence the judiciary as they do so—in other words, that the Notwithstanding Clause prescribes not weak judicial review, but no judicial review at all. There are also reasons to conclude that the structure of Canada’s Constitution not only permits declaratory review of legislation that operates notwithstanding specific Charter provisions, but cannot function properly without it. The Supreme Court of Canada has a distinct opportunity, in resolving the two appeals outlined here, to secure ongoing dialogue between the legislative and judicial branches of government as well as between legislatures and the electorates to whom they’re accountable. 

Footnotes
  1. ^ The seven most recent invocations of the Notwithstanding Clause were as follows: (1) Quebec’s An Act respecting French, the Official and Common Language of Québec, (2) Saskatchewan’s The Education (Parents' Bill of Rights) Amendment Act; (3) Ontario’s Keeping Students in Class Act; (4) Quebec’s An Act to, in Particular, Reinforce Laicity in the Education Network and to Amend Various Legislative Provisions; (5) Alberta’s Protecting Alberta’s Children Statutes Amendment Act; (6) Ontario’s Protecting Elections and Defending Democracy Act; and (7) Alberta’s Back to School Act.

    Return to citation ^
  2. ^ With very special thanks to Professor Vicki C. Jackson for her valuable conversation and guidance, and to my fellow editors at the Harvard Law Review for their feedback and editing expertise.

    Return to citation ^