“Saved by the Valve:” Refugees and Concerns with the Safe Third Country Agreement

In Canadian Council for Refugees v Canada (Citizenship and Immigration), 2023 SCC 17 [CCR], the Supreme Court of Canada (the “Court” or “SCC”) unanimously confirmed the designation of the United States (“U.S.”) as a “safe third country” and appropriate partner with which Canada can share responsibility for considering refugee claims. The Court also took time to make several clarifications and comments surrounding Canadian Charter of Rights and Freedoms (“Charter“) claims, and should serve as a touchstone for s. 7 challenges against interrelated legislative regimes for years to come.


Statutory and Regulatory Context

The statutory and regulatory context is crucial to fully understand what is going on here.

Canada and the U.S. are parties to a bilateral treaty, the Safe Third Country Agreement (the “treaty”), which aims to enhance their sharing of responsibilities in considering refugee claims (CCR, para 1). Article 4(1) sets out the treaty’s animating principle: “the country of last presence shall examine, in accordance with its refugee status determination system, the refugee status claim of any person who arrives at a land border port of entry” (CCR, para 37). Therefore, refugee claimants generally must seek protection in whichever of Canada or the U.S. that they first enter, after leaving their country of origin (CCR, para 1). 

In Canada, the treaty is given domestic effect through the Immigration and Refugee Protection Act (“IRPA”) and the Immigration and Refugee Protection Regulations (“IRPR”). Section 101(1)(e) of the IRPA is implements the treaty’s core principle by providing that : 

101 (1) A claim is ineligible to be referred to the Refugee Protection Division if

(e) the claimant came directly or indirectly to Canada from a country designated by the regulations, other than a country of their nationality or their former habitual residence…

Several provisions in the IRPA and IRPR affect the application of this general ineligibility rule, including (1) designation and review mechanisms; (2) limits to the scope of the general rule; and (3) exemptions (CCR, para 37). The latter two essentially function as legislative safety valves: provisions that are tailored to survive constitutional scrutiny arising from concerns over the ineligibility and removal of refugee claimants from Canada. In the context of a s. 7 Charter claim, the second provision does this by preventing a constitutional breach before it happens when certain conditions are met, while the third provision achieves this by creating discretionary exemptions that ensure compliance with the principles of fundamental justice (“PFJs”) (CCR, para 10).

With respect to the first provision affecting the general ineligibility rule (designation and review mechanisms), there are two key mechanisms of note. First, s. 101(1) (a) of the IRPA provides that only countries that are compliant with the non-refoulement obligations under Article 33 of the Refugee Convention and Article 3 of the Convention Against Torture are eligible for “designation” under the IRPR (CCR, para 38). Under international law, the non-refoulement obligations prohibit States from directly or indirectly returning a person to a place where they would face certain kinds of irreparable harm, including threats to their life or freedom, torture and cruel or degrading treatment (CCR, para 2). Second, ss. 101(2) and 101(3) of the IRPA ensures for the continued review of a designated country’s compliance with its non-refoulement obligations (CCR, paras 39-40). Importantly, the U.S. is the only designated country under s. 159.3 of the IRPR (CCR, para 40). 

The IRPR also narrows the application of the general ineligibility rule by ensuring that not all refugee claimants arriving from the U.S. are returned there. For example, s. 101(1)(e) only applies to claimants arriving at land ports of entry (IRPR, s. 159.4). There are also many exceptions under ss 159.5 and 159.6 of the IRPR that apply to certain circumstances or persons (like unaccompanied minors), or to achieve certain outcomes (like family reunification) (CCR, para 41). What’s more, the IRPR preserves Canada’s discretion under Article 6 of the treaty to consider “any refugee status claims” when it is in the public interest (CCR, para 42; also see paras 36, 153 and 162).

The IRPA contains several mechanisms for temporary or permanent exemptions from return to the U.S. under the general ineligibility rule. For example, officers may grant an administrative deferral from the enforcement of a removal order (s. 48(2)), or they may grant a temporary resident permit (s. 24) (see CCR, paras 44-45). Additionally, while returnees are not eligible for a pre‑removal risk assessment before their removal from Canada and there is no administrative appeal body set out in the IRPA for them, they may file for judicial review at the Federal Court and judges may grant stays of removal while such proceedings are ongoing (CCR, para 43). Other carve outs exist: see ss. 24(3), 25.1(1), 25.2(1), and 233 of IRPA, and CCR paras 44-48 for additional details.



The appellants are a mix of individual refugee claimants and public interest litigants. Each individual appellant arrived at a land port of entry from the U.S. to claim refugee protection in Canada. Their claims were deemed ineligible for referral to the Refugee Protection Division for review pursuant to the combined effect of s. 101(1) (e) of the IRPA and s. 159.3 of the IRPR (CCR, paras 14-15). The appellants claimed that returning them to the U.S. exposed them to significant harm, including: (1) the risk of detention in harsh conditions in the U.S.; and (2) the risk of refoulement that arose from both detention and American asylum policies (CCR, paras 16-19; also see paras 85-86, 93 and 97). 

The appellants challenged the constitutionality of s. 159.3 of the IRPR on three bases. First, s. 159.3 was ultra vires (‘beyond the powers’) because of post-promulgation constraints on the Governor in Council’s statutory authority to maintain a designation (CCR, para 49). Second, the risks that returnees to the U.S. were exposed to infringed s. 7 of the Charter (CCR, para 85). Third, s. 159.3 adversely affected women fearing gender-based persecution because the American interpretation of “particular social group” (one of the classes protected from refoulement) was overly restrictive with respect to women (CCR, para 173).


Judicial History

Federal Court (2020 FC 770)

At trial, the claimants argued that s. 159.3 of the IRPR was unconstitutional on the basis that the designation of the U.S. as a safe third country was ultra vires the authority granted by the IPRA. They also argued that s. 159.3 of the IRPR and s. 101(1) (e) of the IRPA violated ss. 7 and 15 of the Charter.

McDonald J. rejected the vires argument because an earlier case – Canadian Council for Refugees v Canada, 2008 FCA 229 [FC] – held that whether a regulation is within the authority delegated by a statute is assessed based on facts at the time of promulgation (CCR, paras 20-21). 

However, she found that the claimants’ s. 7 rights to liberty and security of the person (“SOP”) were unjustifiably infringed (CCR, para 20). The right to liberty was infringed because, under Canada’s legislative scheme, claimants who were ineligible under the treaty were “immediately and automatically imprisoned by US authorities” (CCR, para 23). Additionally, the right to SOP was infringed because: (1) there was a real risk of one appellant being refouled from the U.S. due to “the challenges in advancing an asylum clai[m] for those detained”, such as barriers to accessing legal advice; and, (2) because being subjected to some detention conditions flowed directly from Canada’s conduct of returning claimants to the U.S. (CCR, para 23). These breaches violated the PFJs because some available safeguards to protect against overbreadth and gross disproportionality were “illusory” and did not constitute a reasonable limit under s. 1 of the Charter (CCR, paras 24-25). 

McDonald J. declared s. 101(1)(e) of the IRPA and s. 159.3 of the IRPR of no force or effect and suspended the declaration for six months (CCR, paras 20 and 25). She did not rule on the s. 15 claims.

Federal Court of Appeal (2021 FCA 72)

The Federal Court of Appeal (“FCA”), allowed Canada’s appeal, dismissed the claimants cross-appeal on the s. 15 claim and the vires argument, set aside the Federal Court’s judgment, and dismissed the applications for judicial review (CCR, para 26). It also declined to adjudicate the s. 15 claim and remitted this issue to trial (CCR, para 29).

Stratas J.A., writing for a unanimous FCA, allowed Canada’s appeal on the s. 7 issue because the claim was improperly constituted. In his opinion, the provisions in an interrelated legislative scheme cannot be considered in isolation from the legislative whole, and administrative action or inaction must be challenged when it is the source of a Charter violation, not the legislation (CCR, para 27). Nonetheless, the FCA considered the s. 7 claim and decided that the trial judge erred by findings that returnees to the U.S. are automatically detained and that the scheme’s legislative safety valves were illusory, as well as her analysis of U.S. detention conditions and the PFJs (FCA, paras 138-140, 144, 146, and 161; CCR, para 28). 


Issues On Appeal

  1. Is s. 159.3 of the IRPR ultra vires? 
  2. Does s. 159.3 of the IRPR unjustifiably breach s. 7 of the Charter
  3. Should the s. 15 Charter claim be remitted to trial or decided based on the appeal record?



Writing for a unanimous court, Kasirer J. answered the first and second questions on appeal in the negative and decided to remit the s. 15 claim to the Federal Court for trial. 

Section 159.3 of the IRPR is Intra Vires 

The appellants advanced two arguments on the vires issue. 

First, maintaining the designation of the U.S. was unreasonable because refugees cannot access effective protections there (CCR, para 49). Kasirer J. rejected this because the appellants had misapprehended the contours of s. 102 of the IRPA. Sections 102(1) (a) and 102(2) codified “conditions precedent” for a country to be designated and subsequently promulgated as such under s. 159.3 of the IRPR (CCR, para 52). The relevant date for to ascertain a country’s compliance with those conditions is the date of its promulgation as a designated country (CCR, para 52). Since the appellants focused their argument on post-promulgation developments (events relevant to the maintenance/continuance of the designation), this aspect of the claim failed (CCR, para, 54).

Second, they argued that the Governor in Council failed to carry out their statutory obligations under s. 102(3) of the IRPA to continuously review U.S. compliance with the refugee protection standards listed in s. 102(2) of the IRPA (CCR, para 49). Kasirer J. rejected this too, explaining that while the continuous compliance reviews conducted pursuant to s. 102(3) of the IRPA “may be challenged based on administrative law principles” in a judicial review, such reviews are outside the scope of a vires challenge, because “these reviews are not directed at whether the regulation exceeds the limits imposed by the statute” (CCR, paras 50-53). The appellants did not seek judicial review of particular s. 102(3) compliance reviews, so this aspect of the claim also failed (CCR, para 53).

Section 7 was Violated … but in Accordance with the PFJs 

After laying out the test for a violation of s. 7 of the Charter (CCR, para 56), the Kasirer J. found that both of the courts below erred. 

(1) Legislative Safety Valves & Constituting the Section 7 Claim

Kasirer J. held that the s. 7 claim was properly constituted. The appellants were not required to target legislative safety valves in constituting their arguments, nor were they obliged to target individual administrative conduct. (CCR, para 79). In a Charter claim challenging a law of general application, “[p]rovisions in a complex interrelated scheme should not be viewed in isolation” because “[t]he success or failure of a Charter claim may turn on arguments or evidence related to preventative or curative provisions” (CCR, paras 62-63). Curative and preventative provisions (i.e. legislative safety valves) that form “part of an integrated legislative whole” must be “reviewed” and “accounted for” in determining whether a violation really exists (CCR, paras 63-64). 

He clarified that a properly constituted Charter challenge does not need to target legislative safety valves “when the provision of general application to which they relate is a cause of the alleged mischief” (CCR, para 69). In such cases, it is the reviewing court’s responsibility to review the entire statutory context, including legislative safety valves, irrespective of how claimants choose to structure their s. 7 challenges (CCR, para 69). Indeed, legislative safety valves may be relevant to and can be considered by a reviewing court at various points in the s. 7 analysis (CCR, para 70-71). Here, Kasirer J. found it most appropriate to address them in considering the impugned law’s accordance with the PFJs (CCR, para 73).

Lastly, the Court highlighted that the mere existence of alternative avenues for challenging the constitutionality of the legislative scheme – such as challenging administrative actions, like s. 102(3) reviews – did not insulate s. 159.3 of the IRPR from a Charter challenge, nor did it preclude the appellants from challenging the law itself (CCR, paras 80 and 82).

For all of these reasons, the appellants’ s. 7 challenge of s. 159.3 of the IRPR was properly constituted and the FCA erred in its opposite conclusion. 

(2) The Rights to Liberty and Security of the Person were Infringed

Kasirer J. found that the risks of detention, the harsh conditions of detention upon return to the U.S., the risk of refoulement, and the U.S. approach to gender-based claims engaged the liberty and SOP interests of the appellants (CCR, paras 89-93; para 108). 

To illustrate that a sufficient causal connection existed between Canadian state action and the potential deprivation of the s. 7 interests engaged, the appellants had to show that: (1) Canada’s participation was a necessary precondition of the harmful effects alleged, and (2) that these risks were foreseeable consequences (CCR, para 111). Foreseeability requires actual or constructive knowledge of the risks of potential harms (CCR, para 114). 

The causal connection was clear in this case. The Safe Third Country Agreement, enforced through the IRPA and IRPR, required Canada to send claimants back to the U.S. where they faced harmful effects. Without the treaty and the impugned legislation, “individuals could advance their refugee protection claims in Canada” (CCR, para 112). Thus, Canada’s participation was a necessary precondition to the harmful effects alleged. 

Turning to the foreseeability requirement, Kasirer J. first found that McDonald J. erred by insufficiently explaining her findings that the relevant effects were foreseeable. The record did not support a finding that most of the harms claimed to arise in U.S. detention facilities were foreseeable consequences of Canada’s actions (CCR, paras 115-116). 

However, the record did substantiate the foreseeability of some negative effects. These included the fact that returnees were exposed to the risk of detention, the “widespread practice of medical isolation” in U.S. detention centres, the “one-year bar” (a rule under which asylum claims must be advanced within a year of a claimant’s arrival) and the treatment of gender-based claims upon return to the U.S. (CCR, para 117).

As such, there was a sufficient causal connection between Canada’s legislative regime that implemented the treaty and the deprivations of the appellants’ s. 7 rights to liberty and SOP (CCR, para 117).

(3) The Infringements Accorded with the PFJs / were Saved by Legislative Safety Valves 

Kasirer J. identified the applicable PFJs as overbreadth and gross disproportionality. As described in Canada (Attorney General) v Bedford, 2013 SCC 72 [Bedford], the principle of overbreadth requires asking whether the impugned legislation “is so broad in scope that it includes some conduct that bears no relation to its purpose” (Bedford, para 112 [emphasis in original]; CCR, para 141). Gross disproportionality, meanwhile, requires asking whether the effect(s) of a deprivation of s. 7 is so out of sync with the law’s purpose that it cannot be rationally supported (Bedford, para 120; CCR, para 147).

He recalled that, when assessing whether legislation violates the PFJs, courts consider “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” (CCR, para 127). Courts then assess whether the legislation breaches any applicable PFJs in light of that legislative purpose (CCR, para 127).

After applying the guiding principles for characterizing the legislative purpose in a s. 7 analysis, established in R v Ndhlovu, 2022 SCC 38, he examined intrinsic and extrinsic evidence related to s. 159.3 of the IRPR and found that the legislative purpose was “to share responsibility for fairly considering refugee claims with the United States, in accordance with the principle of non-refoulement” (CCR, paras 128-139). Against this backdrop, he concluded that the s. 7 claims failed because the deprivations of liberty and SOP accorded with the PFJs, and even if they didn’t, any constitutional defects were cured by the legislative safety valves (CCR, paras 147, 152 and 163). 

Neither the risk of detention nor medical isolation met the high threshold for overbreadth or gross disproportionality (CCR, paras 140 and 147). Starting with the former, Kasirer J. reasoned that the risk of detention in the U.S., with opportunities for release and review, was related to the legislative objective of sharing responsibility for refugee claims. While the responsibility-sharing purpose exposes refugee claimants to the risk of detention, there is no violation because they are given “… fair consideration in the designated country” (CCR, para 142). Indeed, in the Court’s view, the record showed that returnees’ risk of detention in the U.S. was tempered by “mechanisms that create opportunities for release and provide for review by administrative decision makers and courts” (CCR, para 143). Since nothing in the evidentiary record supported the conclusion that the American asylum system was “fundamentally unfair” and deprived of safeguards, the risk of detention was not overbroad (CCR, paras 142-143).

Nor was the use of medical isolation in the U.S. fundamentally unfair (CCR, para 144). In this case, the claimants that were medically isolated in U.S. detention facilities were awaiting tuberculosis test results. Kasirer J. highlighted Rule 30(d) of the United Nations’ Nelson Mandela Rules, which requires that prisoners with contagious diseases be clinically isolated and provided with adequate treatment (CCR, para 144). Absent any evidence suggesting that U.S. authorities’ utilisation of medical isolation in this case was improper or unnecessary to control public health risks, or that it was for punitive purposes, returnees’ risk of medical isolation in U.S. detention facilities was not an overbroad effect of Canada’s legislative regime (CCR, paras 145-146).

The risks of detention and medical isolation were also not grossly disproportionate, because “these risks are within the mutually held norms accepted by our free and democratic societies” (CCR, para 147).

With respect to the risk of refoulement due to U.S. asylum policies, Kasirer J. resolved that while this risk might otherwise occasion a deprivation of returnees’ s. 7 Charter rights in a manner that is contrary to the PFJs of overbreadth and gross disproportionality, any such effects were guarded against or cured by the legislative safety valves in the IRPA (CCR, paras 148 and 155). Especially relevant here were the safety values in the IPRA providing for administrative deferrals of removal, temporary resident permits, humanitarian and compassionate exemptions, and policy exemptions (CCR, paras 148 and 155).

Further, any ‘real and not speculative risks’ could be reconciled by the power preserved under Article 6 of the treaty and exercising the broad legislative discretion to consider any refugee claim if it is in the public interest (CCR, paras 152-153). He emphasized that the sweeping curative power under Article 6 can be applied to “individuals who would not normally be eligible under an exception but who nevertheless warrant special consideration because of their vulnerability” (CCR, para 156; also see para 162). As such, these safety valves were not “illusory”, and the trial judge erred by not substantively assessing the relevant curative mechanisms that could have limited the breaches of s. 7 rights (CCR, paras 158 and 160).

Considering the foregoing, the SCC unanimously agreed that even if refugee claimants faced a real risk of refoulement from the U.S., Canada’s legislative implementation of the treaty — and in particular s. 159.3 — accorded with the PFJs (CCR, para 163). 

Since the deprivations of liberty and SOP accorded with the PFJs, Kasirer J. declined to engage in or speculate on a hypothetical s. 1 justification analysis (CCR, paras 165-171).

Section 15 Issue

Kasirer J. declined to exercise its fact-finding powers and remitted the s. 15 challenge back to trial (CCR, para 175).

For my analysis and commentary of the SCC’s decision, see the second installment of this post here.

Editor: Farah Abdel Haleem

Daniel Legris

Daniel Legris is a 3L J.D. student at Osgoode Hall Law School. He holds an Honours Bachelor of Arts in Political Science from York University. This year, Daniel will be representing Osgoode as an oralist at the Price Media Law Moot Court Competition, having advanced to the international rounds of the International Criminal Court Moot Court Competition last year. Daniel also serves as Co-Chair for the Osgoode Constitutional Law Society, Vice-President of Internal Affairs for the Osgoode Society for Civil Litigation, and Senior Editor for the Journal of Law and Social Policy. He is also participating in Osgoode’s Disability Law Intensive Clinic. Daniel will be completing his articles at one of Toronto’s leading civil litigation firms. He is interested in all areas of law, cooking, and crying over the Toronto Maple Leafs’ failures.

You may also like...

Join the conversation

Loading Facebook Comments ...