Sanis Health Inc v British Columbia: Flexible Federalism Permits National Class Action on Opioids

In Sanis Health Inc v British Columbia, 2024 SCC 40 [Sanis], the Supreme Court of Canada (“SCC”) upheld British Columbia’s power to represent other Canadian governments in a national class action under its Opioid Damages and Health Care Costs Recovery Act, SBC 2018, c 35 [ORA]. Nearly all provincial, territorial, and federal governments supported the law and intended to participate in the proposed class proceeding.
Côté J’s strong dissenting opinion stressed that the other governments could not affirmatively consent to participate under the ORA’s opt-out regime. She believed the law therefore violated the provincial territorial limits imposed by section 92 of the Constitution Act, 1867, 30 & 31 Vict, c 3 [Constitution Act].
However, the six other judges who heard the case strategically overlooked the distinction between opting in and opting out. Given that the provinces had clearly consented to participate in the specific action at issue, the majority found a meaningful connection between the ORA and the external jurisdictions it affects.
Facts
In 2018, British Columbia passed the ORA to support its recently launched class action against pharmaceutical companies with links to Canada’s opioid epidemic (Sanis, paras 20-22). Primarily, the legislation created a direct cause of action and altered some rules of evidence to facilitate the suit (Sanis, para 22). These provisions were modeled on the Tobacco Damages and Health Care Costs Recovery Act, SBC 2000, c 30 [TRA], which the SCC upheld in British Columbia v Imperial Tobacco Canada Ltd, 2005 SCC 49 [Imperial Tobacco].
Uniquely, section 11 of the ORA (“section 11”) allows British Columbia to bring an action under the ORA on behalf of Canada’s other provincial, territorial, and federal governments (ORA, s 11(1)(b)). If a government does not want to participate in the proceeding, it must opt out according to the regular procedure in section 16 of British Columbia’s Class Proceedings Act, RSBC 1996, c 50 [CPA] (ORA, s 11(2)).
Section 11 has no analog in the TRA (Sanis, para 23). The Court of Appeal for British Columbia (“BCCA”) described it as “represent[ing] a bold step, if not an experiment, in bringing government-led class litigation as close as possible to truly ‘national’ proceedings in Canada’s federal structure” (Sandoz Canada Inc v British Columbia, 2023 BCCA 306, para 3).
The provincial, territorial, and federal governments were receptive to British Columbia’s plan. Nearly all the provinces and territories passed their own ORA-like statutes governing the recovery of opioid healthcare costs incurred within their jurisdictions (Sanis, para 25). The British Columbia class action accommodated these provinces by including them in a sub-class that would recover under their respective statutes (Sanis, para 24).
Judicial History
Several pharmaceutical companies applied to strike section 11 on the basis that it was ultra vires British Columbia’s legislature (Sanis, para 4). They asserted that section 11 dealt with “Property and Civil Rights” under section 92(13) of the Constitution Act but violated section 92’s territoriality limits by addressing substantive rights outside British Columbia.
Both the Supreme Court of British Columbia and the BCCA rejected the applicants’ claims (Sanis, paras 27-29). In their view, the ORA was a purely procedural mechanism falling under section 92(14) of the Constitution Act, “The Administration of Justice in the Province” (Sanis, paras 27, 29). Regarding section 92’s territorial limits, any governments subject to the law had a “meaningful connection” to British Columbia due to their choice to participate in the class proceeding (Sanis, paras 28, 30). It was irrelevant that section 11 imposed an opt-out rather than an opt-in procedure for participation (Sanis, para 28).
Issue
The only issue was whether section 11 was ultra vires British Columbia’s legislature under the Constitution Act.
Decision
Writing for a six-judge majority, Karakatsanis J dismissed the appeal and upheld the lower court decisions. Côté J dissented alone.
Broadly, Karakatsanis J stressed the need for the SCC to take “a more flexible approach to interjurisdictional cooperation” and support cooperative federalism (Sanis, para 1). Côté J protested that “[t]he provinces cannot amend the Constitution by mutual consent” (Sanis, para 196).
Substantively, both Karakatsanis J and Côté J applied the two-part test from Imperial Tobacco for whether matters are properly “in the province” under section 92 of the Constitution Act (Sanis, paras 34, 121-22). First, the court conducts a division of powers analysis: it determines the pith and substance of the impugned provision and classifies it under a constitutional head of power. Second, the court determines whether the provision respects the purposes of the Constitution Act’s territorial limitations. The provision must have a “meaningful connection” to the enacting province and respect other jurisdictions’ territorial boundaries (Imperial Tobacco, para 36).
Pith and Substance
The majority found that section 11’s pith and substance is to create a procedural mechanism for implementing British Columbia’s national class action (Sanis, para 76). Because its purpose and effects are purely procedural, it must be classified under section 92(14) of the Constitution Act, which covers the administration of justice in the province (Sanis, paras 78-79).
Karakatsanis J affirmed that section 11 does not create or alter any substantive rights. Class proceedings legislation is procedural and does not affect substantive rights (Sanis, para 72). While access to class proceedings may affect the enforcement of substantive rights, that does not make it a substantive right in itself (Sanis, para 72).
Côté J, dissenting, found that while section 11’s purpose may be procedural, its effects are substantive (Sanis, para 128). It should therefore be classified under section 92(13) of the Constitution Act, which covers provincial property and civil rights (Sanis, para 125).
Côté J focused on section 11’s use of an opt-out regime to govern participation in the class proceeding. Under section 11, a province will be deemed a prospective class member unless it takes positive steps to opt out (Sanis, para 144). This leads to the loss of significant substantive rights, namely the province’s litigation autonomy and its freedom to commence proceedings elsewhere (Sanis, 146-48).
Karakatsanis J and Côté J also disagreed on whether section 11 created a new substantive right for the Crown in right of British Columbia to represent foreign Crowns in a class proceeding. Karakatsanis J found that it did not (Sanis, para 56). Because foreign Crowns who act in British Columbia are “natural persons” in the province, they were already able to participate in proceedings under the CPA (Sanis, paras 46-60). However, Côté J found that foreign Crowns cannot be “persons” under the CPA due to the principle of Crown immunity, which presumes that general language in a statute does not refer to the Crown (Sanis, para 134). In her view, the CPA by itself does not authorize multi-Crown class proceedings, therefore the ORA gave British Columbia a new substantive right to do something it could not do before (Sanis, para 135).
Territoriality
The majority held that section 11 has a “meaningful connection” to British Columbia through the other provinces’ choice to participate in the class proceeding, meeting the Imperial Tobacco requirement (Sanis, para 88). For the class action to proceed, the British Columbia court would have to find common issues shared by British Columbia and the other provinces, thus establishing a connection (Sanis, para 89). Further, the other provinces would have to consent to litigating under British Columbia’s procedural rules (Sanis, para 91).
Côté J held that there is no meaningful connection (Sanis, para 165). In the first place, this is because “Property and Civil Rights” in another province (Constitution Act, s 92(13)) cannot be connected to British Columbia (Sanis, para 184).
However, Côté J concluded that there would also be no meaningful connection if the pith and substance of section 11 were classified under section 92(14) of the Constitution Act. This is because section 11’s opt-out regime gives the other provinces “no true choice” of whether to participate or not, since they are deemed participants by default (Sanis, para 175). Additionally, the “meaningful connection” test is a higher threshold than the test used by courts to determine whether foreign residents can be included in a class action (Sanis, para 171). Certification of a class action based on common issues is therefore not enough to meet the Imperial Tobacco requirement (Sanis, para 172).
Analysis
Karakatsanis J and Côté J fundamentally disagreed on whether other provincial, territorial, and federal governments are consenting parties under section 11. It is clear from the combination of sections 11(1)(b) and 11(2) that, once British Columbia brought an action on other provinces’ behalf, those provinces were participants unless they took active steps to opt out under section 16 of the CPA.
While many of Côté J’s arguments flow from her identification of section 11 as an opt-out regime, the majority put virtually no weight on it. In her reasons, Côté J suggested that the majority “does not seem to make any sort of distinction between the effects of opting in or opting out” (Sanis, para 142). The charge is accurate. Karakatsanis J consistently implied that they amount to the same thing: “choosing to participate by opting in or not opting out” (Sanis, para 68).
In my view, the majority’s oversight was strategic. By declining to comment on the difference between opting in and opting out, Karakatsanis J established that consent is required for section 11 to be constitutional, yet didn’t define what consent in this context means. That omission accomplishes an important goal. By allowing the opioid class action to proceed without creating a blueprint for real erosions of sovereignty, the majority maintained its commitment to “a more flexible approach” to cooperative federalism (Sanis, para 1).
Flexible Consent
The majority was careful to rely on the provinces’ consent to establish a “meaningful connection” between British Columbia, section 11, and the provinces subject to the provision (Sanis, para 88). Karakatsanis J noted that “[n]o governments are compelled to participate against their will, and the opt-out mechanism lets them make informed and voluntary decisions about whether to subject themselves to B.C.’s courts and procedural rules” (Sanis, para 91). These criteria ensure that future laws modelled on the ORA will not be constitutional if they seek to exploit non-consenting governments that are genuinely unable to opt out.
By avoiding a formalistic definition of consent, the majority left room to distinguish section 11 from more factually problematic opt-out regimes that may arise in the future. Côté J’s finding that the provinces had “no true choice” relies on a formalistic interpretation of section 11’s opt-out mechanism as one that in principle cannot give rise to affirmative consent (Sanis, para 175). While the majority declined to draw the same conclusion, it did not suggest that any opt-out regime would automatically be sufficient. In different circumstances, an opt-out regime may not give rise to consensual participation.
The majority also invoked a strengthened presumption of constitutionality to ensure that its reasoning will not be applied where actual consent is absent. Courts normally presume that a legislature does not intend to exceed its territorial limits (Sanis, para 45). Importantly, however, the presumption is “especially strong” when the governments affected by the impugned law support its validity (Sanis, para 45). For section 11, the strengthened presumption justified setting a low bar for consent. But the bar could be set higher in cases where other provinces dispute a law’s validity.
As a result of these strategic considerations, it is hard to see exactly how Sanis could open the door to any real erosion of jurisdictional sovereignty. The case provides no blueprint for abuse. In the meantime, the majority’s decision allowed the provinces’ opioid class action to proceed, arguably in the interest of all Canadians.
Soft and Hard Reasons
The majority ruling in Sanis could be described as a “soft” one. Karakatsanis J relied heavily on the policy objectives of cooperative federalism. She all but ignored a distinction between opt-in and opt-out regimes that would likely be paramount in other kinds of cases, particularly those affecting individuals as opposed to governments. Finally, she largely eschewed the formalistic rigour of Côté J’s spirited dissent.
In the circumstances, however, these were far from shortcomings. Given the good faith and unanimity of the intervening attorneys general, it was appropriate for the SCC to interpret section 11 in a manner that respected their wishes. Meanwhile, by making consent necessary to satisfy the “meaningful connection” test for national class actions, the majority ensured that the court can take a harder line in cases where good faith is missing.
For this reason, Côté J’s dissent plays an important institutional role in the SCC’s overall messaging. Since the majority accepted that the provinces had consented to participate in British Columbia’s opioid class action, it could safely disregard many of Côté J’s arguments. In the absence of consent, however, those arguments would remain compelling. Thus, their inclusion in the reported decision puts the provinces on notice: Canadian courts have the tools to uphold jurisdictional sovereignty if and when cooperative federalism turns uncooperative.
Conclusion
Nearly all Canada’s provincial, territorial, and federal governments supported and sought to participate in the opioid class action launched under the ORA. Their good faith and cooperation motivated a majority of the SCC to look past the formalistic difficulties with section 11’s opt-out mechanism and uphold its constitutional validity. Section 11 is restricted to procedural rights within British Columbia and has a meaningful connection to other provinces through their consensual participation in the class action.
British Columbia’s national class action against opioid makers was certified on January 22, 2025. It remains to be seen what other national class actions could result from the SCC’s relatively permissive approach to legislation like the ORA.
This article was edited by Alexandre Cachon.
Join the conversation