Clarity on the Challenges for Systemic Challenges: Canadian Council for Refugees v. Canada (Citizenship and Immigration), 2023 SCC 17

The Supreme Court of Canada’s decision in Canadian Council for Refugees v. Canada (Citizenship and Immigration), 2023 SCC 17 sets out some important principles about challenges to government action at the intersection of constitutional and administrative law. On balance, I think the decision limits the potential for systemic challenges to complex statutory and regulatory regimes (i.e. government schemes with multiple moving parts contained in statute and regulations) and creates incentives to seek individualized remedies for violations of the law rather than declarations that a regime in its entirety is unlawful.

The decision is unlikely to be significant in the daily lives of courts, counsel and citizens, as systemic challenges are comparatively rare (though encouraged by Canada’s liberal approach to standing: British Columbia (Attorney General) v. Council of Canadians with Disabilities, 2022 SCC 27). That said, when such challenges are made they tend to involve high-profile matters. This case had a high profile because it involved international relations between Canada and the United States relating to immigration — the Safe Third Country Agreement — which was hugely politically salient during the presidency of Donald Trump. Going forward, this decision will be the touchstone for any individual or (as is more common) group considering a challenge to a complex regulatory scheme.

Four aspects of the judgment are significant: the presumption of validity of regulations; the identification of the source (in legislation or administrative action) of a violation of a constitutional right; the possibility that statutory provisions can cure or prevent violations that would otherwise arise due to the operation of a statutory scheme; and the requirement of demonstrating a causal link between government action and harm caused by the operation of a statutory scheme. The analysis required by the Supreme Court will invariably be highly fact-dependent and context-specific. Those considering challenges to complex regulatory systems will therefore have to think long and hard about whether to mount such challenges at all, given the inevitable uncertainty about whether a particular set of facts will justify a conclusion that a system is unconstitutional.

The STCA is around 20 years old but there have been few points in its existence when it has not been the subject of a constitutional challenge. At base the STCA is quite simple. The agreement provides that some refugee claimants who arrive in Canada via the United States should be handed over to the American authorities to have their claims addressed by them. They are ineligible to make refugee claims in Canada: Immigration and Refugee Protection Act, S.C. 2001, c. 27, s. 101(e)). The legislative framework is as follows: s. 102 of the IRPA permits Canada to designate a third country as safe where Canada has considered a series of factors: whether the country is a party to the Refugee Convention and to the Convention Against Torture; its policies and practices with respect to claims under the Refugee Convention and with respect to obligations under the Convention Against Torture; its human rights record; and whether it is party to an agreement with the Government of Canada for the purpose of sharing responsibility with respect to claims for refugee protection. Canada reached an agreement with the United States (the STCA) and the federal cabinet designated it as a safe third country (Immigration and Refugee Protection Regulations, SOR/2002‑227, s. 159.3). This designation is subject to ongoing oversight: the federal cabinet “must ensure the continuing review of factors set out in subsection (2) with respect to each designated country” (IRPA, s. 102(3)).

However, refugee advocates have long complained that the American system is not fit for purpose and that many of those turned away at the Canadian border suffer serious maltreatment in the United States. This case was based on evidence of just such treatment. At first instance (2020 FC 770) McDonald J held that the STCA was unconstitutional as its operation violated the constitutional right to security of the person by causing refugee claimants turned away from Canada to be detained in dreadful conditions and to be subject to an asylum adjudication system so deficient in its operation that it caused genuine refugees to be returned (refouled) to their country of origin notwithstanding the persecution they would suffer there. Notably, the challenge was directed at the STCA in its entirety: the applicants alleged that the designation of the United States created an ongoing violation of constitutional rights; they did not impugn the federal cabinet’s failure to effectively review the designation. They sought and were granted declarations that the designation of the United States (s. 159.3 of the Regulations) and the consequent ineligibility of refugee claimants to seek asylum in Canada (IRPA, s. 101(e)) were unconstitutional.

The Federal Court of Appeal (2021 FCA 72) allowed Canada’s appeal, essentially because the challenge was misdirected (as the applicants had not attacked the federal cabinet’s review process) and lacked the demonstration of a causal link between the harm suffered and Canadian government action.

The Supreme Court largely came to the same conclusion. It was less dismissive of the applicants’ case — as we will see, it did not consider that the challenge was misdirected — but nonetheless ended up in the same place as the court below: given the presence of preventative and curative provisions in the statutory scheme, and the fact that the state action at issue was mostly American, there was no basis to strike down the STCA. 

First, Kasirer J addressed the proposition that the regulation designating the United States as a safe third country was ultra vires. The difficulty here for the challengers was that they couched their challenge in the present rather than the past tense: they argued, based on the evidence in the record, that it was unreasonable to maintain the designation of the United States. But as Kasirer J pointed out, regulations benefit from a presumption of validity (at para. 54) and, in this instance at least, the conditions were to be satisfied at the time the designation was made: “The statute requires that these conditions be met before, not after, a country is designated” (at para. 52). Any challenge to contemporary failures in the American system would have to be directed at deficiencies in the federal cabinet’s ongoing oversight.

Now, the presumption of validity has recently been questioned, inasmuch as it forms part of a package of principles that contribute to a hyperdeferential approach to judicial review of regulations (Portnov v. Canada (Attorney General), 2021 FCA 171, at paras. 19-28). I do not expect this restatement to weigh heavily in future judicial consideration of this issue: it is well well established that administrative action is presumptively valid; and the real question is whether the Vavilov framework has displaced the hyperdeferential framework for conducting judicial review of regulations. As for the time at which the designation is to be assessed, I do not think STCA stands for a general proposition that vires must always be assessed historically, especially because there was a separate statutory mechanism for keeping the designation under review. Rather, here, both statutory text and context made clear that the focus had to be on conditions at the moment of designation (at para. 53). For contemporary conditions, the focus would instead be on the federal cabinet’s “continuing review” rather than on the original designation.

Second, the Supreme Court had to determine whether the challenge was properly constituted. Here the question of the federal cabinet’s review role had been front and centre in the court below: for the Federal Court of Appeal, the source of any constitutional violation attributable to Canada would have to lie in the failure of the federal cabinet to adequately conduct its ongoing review of the STCA. Kasirer J rejected this approach:

Section 102(3) of the IRPA requires the Governor in Council to ensure the continuing review of the s. 102(2) factors, such as the designated country’s human rights record. The reviews conducted to date have left the designation in place, which remains the legislative basis for the ineligibility of claims advanced by the individual appellants. While the administrative conduct that led to the designation being maintained may also be susceptible to constitutional challenge in its own right, the existence of this alternative does not insulate s. 159.3 of the IRPR from Charter scrutiny. Similarly, while the appellants might also have challenged administrative decisions pertaining to the applicability of exceptions or the availability of exemptions, this does not preclude challenges to s. 159.3 (at para. 82).

There were, nonetheless, strong hints throughout the judgment that it would be easier to seek individualized remedies for harm suffered. For instance, in explaining that the scheme was constitutionally sound, Kasirer J noted:

It … remains open to refugee claimants arriving from the United States to advance constitutional challenges against determinations that their claims are ineligible to be referred to the Refugee Protection Division or against decisions that exceptional relief is unavailable. As this Court has recognized, it is possible to consider challenges to both legislation and administrative conduct within the same proceeding…If claimants in future cases challenge administrative conduct, courts may conclude that returning them would shock the conscience of Canadians … In this way, vulnerable claimants whom Canada improperly seeks to return can be protected. However, at oral argument, the appellants expressly disclaimed any intention to challenge the constitutionality of administrative decisions to deny them exemptions (at paras. 121-122).

Of course, there is nothing to prevent an individual or group seeking both individualized remedies and making a systemic challenge to the constitutionality of a statutory and regulatory regime. But an elusive question remains without an answer: in what conditions is a systemic challenge more appropriate than an individualized remedy? No test has yet been laid down by the Supreme Court.  I suggested more than a decade ago that where there is an administrative actor in a position to grant relief from the breach of a constitutional right, any challenge should be directed at the administrative actor rather than the legislation. Unlike the court below, however, Kasirer J did not seek to overtly constrain the litigation options of prospective challengers. Yet the consequence of the Supreme Court’s approach may ultimately be to constrain prospective challengers.

Third, the main reason that the scheme survived constitutional challenge was the presence of legislative provisions preventing or curing breaches of constitutional rights. This really is the crux of the Supreme Court’s decision and also the primary barrier between prospective challengers and a successful constitutional challenge. The Federal Court of Appeal had characterized such provisions as “safety valves” and concluded that the scheme could only be unconstitutional if the safety valves demonstrably misfired. Kasirer J was less categorical — and did not conclude that the challenge was fundamentally flawed because of a failure to focus on the functionality of safety valves — but nonetheless arrived in more or less the same place: where preventative or curative provisions are in place, this will generally support a finding that a system is constitutional; challengers will have to provide evidence that the safety valves are not functioning as intended.

Kasirer J described preventative and curative provisions as follows:

Curative measures are … remedial: they repair a breach that would be caused by a general rule by providing a targeted exemption after the fact … These measures often work together with preventative measures to limit the scope of a provision of general application. Preventative measures narrow a general rule by precluding its application in anticipation of a breach, often through legislative exceptions. These categories are not watertight compartments, nor are they exhaustive. Moreover, if a class of individuals habitually receive individualized exemptions after the fact, a legislature could enact a class‑wide exception that applies in advance (at para. 68).

These characteristics of preventative and curative provisions are legally relevant:

At the engagement stage, preventative provisions can tailor a provision of general application so carefully that it never threatens s. 7interests. For instance, s. 159.6 of the IRPR prevents the threat to life that might emerge from returning individuals subject to the death penalty. In so doing, preventative provisions like s. 159.6 rule out certain s. 7 engagements. By contrast, curative provisions will rarely, if ever, preclude the engagement of s. 7. PHS provides direct support for this proposition, as this Court held that the general prohibition on possession engaged s. 7despite the availability of safety valves. Curative provisions create exceptional departures from a general rule; they are typically available only after a determination that the general rule applies. The possibility of obtaining an exemption is therefore a path through which the risks the general rule poses to life, liberty or security of the person can sometimes be avoided. In such cases, the threat to the s. 7 interests persists, but it does not always materialize (at para. 71).

They can, ultimately, make a scheme lawful even if there are instances of individualized harm: “If the legislature has crafted a scheme that cures potential breaches by providing exemptions that can target the specific deprivations, this can render the legislative scheme Charter‑compliant” (at para. 76). Here, the presence of preventative and curative provisions was fatal to the challengers’ case. While Kasirer J accepted that in many respects the challengers had identified interferences with the security of the person, he held that these interferences did not violate s. 7 of the Charter. A s. 7 violation can only be made out where a person is deprived of security of the person in a way that is not in accordance with the principles of fundamental justice: a challenger must demonstrate, therefore, that the deprivation was arbitrary, overbroad or grossly disproportionate (that is, not in accordance with the principles of fundamental justice). The test is “whether the law’s purpose, taken at face value, is connected to its effects and whether the negative effect is grossly disproportionate to the law’s purpose” (Canada (Attorney General) v. Bedford, 2013 SCC 72, [2013] 3 SCR 1101, at para. 125).

At this stage, the preventative and curative provisions played heavily in the analysis. Canada retained discretion under the STCA to deal with individual refugee claims (and could exercise this discretion on a case-by-case basis under s. 42 of the IRPA) and the IRPA contained various other mechanisms (such as humanitarian and compassionate exemptions and temporary resident permits) that would prevent harm materializing:

When the IRPA’s safety valves are activated, claimants can be exempted from return. If they are not returned to the United States, they do not face any risk of refoulement from the United States. The safety valves can therefore intervene to cure what might otherwise be unconstitutional effects … This is not a case where the scheme’s safety valves are so disconnected from the deprivation that they are irrelevant for determining whether the challenged provision violates s. 7 of the Charter. Properly interpreted, the curative mechanisms in the broader statutory scheme are sufficient to ensure that individuals are not subjected to real and not speculative risks of refoulement, if such risks do exist … The related provisions in the IRPA and IRPR provide various avenues for refugee claimants to be exempted from return on a temporary or permanent basis. For instance, an officer who is persuaded during an initial interview that a real and not speculative risk of refoulement exists could refer the case to Immigration, Refugees and Citizenship Canada for further consideration. That department could then determine whether to recommend that the Minister grant a humanitarian and compassionate exemption under s. 25.1 of the IRPA, including waiving the s. 101(1)(e) ineligibility. Alternatively, the Minister might decide that grounds for granting a temporary resident permit exist, thus providing time for determining whether permanent resident status should be granted on humanitarian and compassionate grounds. These avenues are, of course, not exhaustive. But they suggest that the statutory scheme can limit the risk of harm that the general rule might otherwise occasion (at para. 155. See also at para. 143).

Hence Kasirer J’s observation (perhaps a strong hint) that individualized remedies are appropriate to catch instances of the preventative and curative provisions misfiring: “It may be that administrative actors, such as CBSA officers, acted unreasonably or unconstitutionally in their treatment of some returnees or in their interpretation of the legislative scheme, including its safety valves … If administrative malfeasance results in returning individuals to circumstances that would shock the conscience of Canadians, such as returning individuals to face a real and not speculative risk of refoulement, constitutional and administrative remedies remain available” (at para. 164). Without evidence that the provisions are not performing as intended, a systemic challenge will be difficult, if not impossible, to maintain:

It may not always be obvious whether the source of an alleged breach is the legislation or the administrative conduct implementing it … As is true at every stage of proving a s. 7 violation, challengers bear the evidentiary burden to establish that the legislation causes difficulties for individuals seeking access to curative mechanisms … They must therefore show that the legislation causes the exemption to be illusory in their individual circumstances. Challengers need not show that the legislation causes exemptions to be illusory in general, as the s. 7 analysis is not focused on “the percentage of the population that is negatively impacted” (Bedford, para. 123). When courts conclude that the legislation causes the alleged difficulties, they should explain how this burden is met for each of the relevant curative mechanisms (at para. 159).

On the one hand, courts should “rarely preclude consideration” of s. 7 breaches but on the other, “challengers are better advised to produce evidence of misfiring safety valves: “the success or failure of a Charter claim may turn on arguments or evidence related to preventative or curative provisions” (at para. 63). Harm to one person is not enough, on its own, to justify a declaration of unconstitutionality of a complex statutory and regulatory scheme: it must be harm that was neither prevented nor cured by the safety valves in the scheme.

Judicial review, incidentally, was neither a preventive nor curative provision on this definition (at para. 77) as the Federal Court of Appeal had suggested. Nor were the federal cabinet reviews under s. 102(3), as the reviews “do not make after-the-fact relief available on an individual basis” (at para. 81). This is an overly narrow view (I think) of the potential relevance of curative provisions, as the removal of the designation would invariably cure all potential harm under the scheme but clearly and in contrast to the court below, the Supreme Court saw the possibility of individualized relief as inherent to the concept of a curative provision.

In the end, however, the Supreme Court ended up in more or less the same place as the court below. The Federal Court of Appeal inferred from the absence of evidence about the misfiring of safety valves that the challenge was not properly constituted (and though that there was a wider range of safety valves). But although the Supreme Court disagreed on this particular point, the challengers’ failure to focus attention on the preventive and curative provisions was held to be fatal. Knowing that a challenge was properly constituted but fatally flawed rather than improperly constituted tout court will come as small satisfaction. And I suspect prospective challengers in the future will think long and hard about bringing systemic challenges. In this, the merit of the Federal Court of Appeal’s approach is that it faced up to overtly constraining the scope of such challenges (and provided a principled explanation for such constraints).

Lastly, Kasirer J also addressed the issue of causation. To some extent, the second and third aspects discussed above also relate to causation, in the sense that they require a challenger to demonstrate a connection between a statutory and regulatory scheme and harm suffered. Indeed, it is clear that this connection must have an evidential basis. Kasirer J did not discuss these aspects under the rubric of causation, which he reserved for the issue of determining whether Canadian state action was responsible for the harm identified by the challengers. Specifically, Canadian state action must be a necessary precondition to the harm materializing, which must also be a foreseeable consequence of the operation of the scheme (at paras. 112-114). Here, again, the challengers faced hurdles and only some of the identified harms were causally related to Canadian state action, in the sense that Canada knew or ought to have known they could occur (at paras. 116-117).

But given the lack of evidence on key points, there was no breach of s. 7 and no need to consider whether the breach could be justified under s. 1. Kasirer J observed, in this regard, that justification is possible in principle: “the government may argue — as Canada did here — that a scheme’s safety valves, if they proved to be imperfect under s. 7, nonetheless render the legislation minimally impairing. Further, the government might seek to rely on safety valves in the final balancing stage of Oakes” (at para. 171).

Yet another hurdle for prospective systemic challengers to surmount! Again, evidence is key. And again, the Federal Court of Appeal and Supreme Court ended up in the same place. Whereas the Federal Court of Appeal frontloaded the analysis and assessed causation as a logical prerequisite to considering whether s. 7 had been breached, the Supreme Court was content to first consider the evidence but in both instances the absence of an evidentially based link between the conduct and harm meant that the systemic challenge could not succeed. 

Prospective challengers to complex statutory and regulatory schemes will in future have to accurately identify preventative and curative provisions (on which, it is worth noting, experienced judges on the Federal Court of Appeal and Supreme Court diverged); lead evidence on the operation of these preventative and curative provisions (which might not be easily obtained); and link the evidence of harm and misfiring safety valves to Canadian state action. These are considerable hurdles, of which prospective challengers will no doubt take careful note.

This content has been updated on September 4, 2023 at 10:29.