|
Class Proceeding (Ont) - Certification - 'Common Claims and Defences' [CPA 5(1)(c)] (2). Richard v. Canada (Attorney General)
In Richard v. Canada (Attorney General) (Ont CA, 2025) the Ontario Court of Appeal dismissed a Crown appeal, here brought against a motion judge's certification of a class proceeding where "the Canada Border Services Agency (“CBSA”) placed some immigration detainees ... in provincial prisons instead of Immigration Holding Centers (“IHCs”)".
The court considers the 'common issues' element of class proceedings certification, here in an appeal context:Common Issues
[48] Canada argues that the PCIs [SS: 'proposed common issues'] cannot be resolved without extensive individual findings. In its factum, it submits that “inevitably, should this action be allowed to proceed as a class action, it would break down into individual proceedings, or at a minimum, into groups based on the claimants’ province or territory of residence.”
[49] When determining whether issues are sufficiently common to warrant certification, the “some basis in fact” test applies. As this court noted in Palmer v. Teva Canada Ltd., 2024 ONCA 220, 495 D.L.R. (4th) 151, “[w]hile the ‘some basis in fact’ test is a low evidentiary standard, and a court should not resolve conflicting facts and evidence, the court retains a gatekeeping function and certification will be denied if there is an insufficient evidentiary basis for the facts to establish the existence of common issues”: para. 104
[50] More recently in Stolove, this court provided the following instructions, at paras. 27-28:[A] class proceeding can only be certified if it raises common issues. For an issue to be a common issue, it must be a substantial ingredient of each class member’s claim such that its resolution will avoid duplication of fact-finding or legal analysis, thereby facilitating judicial economy and access to justice: Hollick v. Toronto (City), 2001 SCC 68, [2001] 3 S.C.R. 158, at para. 18. The answer to a question raised by a common issue must be capable of extrapolation, in the same manner, to each member of the class.
[T]he motion judge acknowledged that an issue can be a common issue even if it makes up a very limited aspect of the liability question and even though many individual issues remain to be decided after its resolution: Western Canadian Shopping Centres Inc. v. Dutton, 2001 SCC 46, [2001] 2 S.C.R. 354, at para. 39. Even a significant level of individuality does not preclude a finding of commonality: Hodge v. Neinstein, 2017 ONCA 494, 136 O.R. (3d) 81, at para. 114. Moreover, a common issue need not dispose of the litigation. All that is required is that it be an issue of fact or law common to all claims such that its resolution will advance the litigation. [Emphasis added.] [51] As discussed, the motion judge’s positive finding on common issues is a question of mixed fact and law, and therefore reviewable on a standard of palpable and overriding error: see Stolove, at paras. 45-46. After reviewing the motion judge’s findings on these issues, immediately below, I find no palpable and overriding errors.
[52] Pursuant to s. 5(1)(c) of the CPA, the motion judge accepted that there was “some basis in fact” for the PCIs. The PCIs certified by the motion judge in his order dated August 28, 2024, are reproduced in an Appendix to these reasons.
[53] Canada did not object to PCIs 8-10 which dealt with damages generally. The motion judge rejected Canada’s objection to the remaining PCIs on liability and aggregate damages. He held that a finding by the trial court that immigration detention in provincial prisons is contrary to the Charter and not saved by s. 1 (PCIs 1 to 3), or that Canada owes a duty of care to the class and breached it (PCIs 4 to 7), would avoid duplication of fact-finding and the legal analysis of issues that are a substantial ingredient of each class member’s claim: Pro-Sys Consultants Ltd. v. Microsoft Corporation, 2013 SCC 57, [2013] 3 S.C.R. 477, at para. 108; referring to Western Canadian Shopping Centres Inc. v. Dutton, 2001 SCC 46, [2001] 2 S.C.R. 534, at paras. 39-40. Likewise, the motion judge was satisfied that the common issues court could determine aggregate damages (PCI 11) by finding a “base level” of damages without evidence from individual class members or by awarding aggregate damages on the principles of deterrence and vindication: see Good v. Toronto (Police Services Board), 2016 ONCA 250, 130 O.R. (3d) 24, leave to appeal refused, [2016] S.C.C.A. No. 255, at paras. 72-75, and Brazeau, at paras. 102-103.
[54] Canada argued that PCIs 1 to 7 and 11 would break down into individual issues because immigration detainees were placed in provincial prisons for different reasons based on individualized assessments. It referred to the NRAD form used by CBSA officers to determine if an immigration detainee should be held in an IHC or provincial prison. The NRAD form considers the risk that a detainee will fail to appear for a future immigration-related process, risks they may pose to themselves or others, and other factors including pregnancy, restricted mobility, and suspected or known mental illness.
[55] The motion judge rejected this argument. He held that the reasons for placement in a provincial prison are irrelevant to the PCIs given the respondents’ pleading that Canada cannot ever lawfully impose punitive detention on an immigration detainee, regardless of the circumstances. Even if the reasons for detention were relevant, the motion judge held that there was some basis in fact for systemic commonality in the CBSA’s decision-making process as the NRAD form itself “provides a common and categorical set of reasons informing the selection of an IHC versus a provincial prison.” Further, there was some basis in fact that the CBSA’s placement decisions were guided by policies and practices that had nothing to do with individualized assessments of detainees, including overcrowding and IHC capacity.
[56] The motion judge also dismissed Canada’s challenge to the PCIs on the basis that class members were placed in different provincial prisons and therefore experienced different conditions. Pointing to the expert evidence of Canada’s own witnesses, and reports of the Red Cross (the organization in charge of monitoring CBSA’s compliance with the national mandate on the treatment of immigration detainees), the motion judge took note of the “core” conditions faced by immigration detainees in provincial prisons. In his view, these conditions, and their effects on class members, could be compared to a similarly consistent or “core” set of conditions and effects in IHCs. The motion judge noted further that immigration detainees in IHCs were afforded greater privacy, autonomy, and access to information regarding their legal status, while those in provincial prisons experienced strip searches, co-mingling, and restrictions on family visits. In this context, the motion judge held that the trial court would be well equipped to explore the PCIs without requiring evidence from individual class members.
[57] I conclude, first, that there is no basis to interfere with the motion judge’s finding that “[a] difference in the reasons for detention does not lead to a breakdown of PCIs 1 to 7. To the contrary, the [respondents] submit that any detention of an immigration detainee in a provincial prison is unlawful, regardless of the reason” (emphasis in original). Canada’s argument that liability for both the alleged Charter breaches and negligence depend on the individual assessment of CBSA officers seeks to circumvent what has been presented as the core issue: that the detention of any immigration detainee in prison is unlawful. In staking out its position Canada appears to assume that because detention under the IRPA is supposed to be administrative, that the detention in question is in fact administrative and therefore lawful. That could very well be the finding at the end of the day, but if there is some basis in fact to suggest that immigration detention in provincial prisons is unlawful across the proposed class, this issue is for the common issues judge to decide. Canada has not identified a palpable and overriding error in the motion judge’s reasons on this point.
[58] Second, the motion judge recognized that there may be differences in the conditions from one province or institution to another, but in the end correctly concluded that there is no requirement that conditions be identical to find commonality. There is no basis to disturb this conclusion.
[59] There is considerable jurisprudential guidance on the certification of common issues in class actions alleging systemic negligence. For example, in both Rumley v. British Columbia, 2001 SCC 69, [2001] 3 S.C.R. 184 and Cloud v. Canada (Attorney General), 2004 CanLII 45444 (ON CA), 73 O.R. (3d) 401 (C.A.), leave to appeal refused, [2005] S.C.C.A. No. 50, the courts held that questions about duty of care, standard of care, and breach, could meaningfully advance class members’ claims even if individualized trials on causation and damages would still be necessary: Rumley at paras. 27-30; Cloud at paras. 60, 63-64. Similar guidance exists for common issues arising out of Charter claims related to “a single course of conduct” (like the conditions experienced by inmates in a jail based on general operational methods): Cirillo v. Ontario, 2021 ONCA 353, 486 C.R.R. (2d) 25, leave to appeal refused, [2021] S.C.C.A. No. 296, at paras. 62, 65, referring to Johnson v. Ontario, 2016 ONSC 5314, 364 C.R.R. (2d) 17.
[60] There is no basis to disturb the motion judge’s finding of commonality. . Carcillo v. Ontario Major Junior Hockey League
In Carcillo v. Ontario Major Junior Hockey League (Ont CA, 2025) the Ontario Court of Appeal dismissed a plaintiff appeal where class proceeding certification had been denied.
Here, while finding that the representative plaintiff-appellants had made out the s.5(1)(c) 'common issues' requirement, the court summarizes applicable 'common issues' law - particularly as it relates to cases of systemic negligence:(1) Governing Principles
[39] Section 5(1)(b) [SS: should be "(c)"] of the Class Proceedings Act requires that certification promote efficiency by ensuring that “allowing the suit to proceed as a [class proceeding] will avoid duplication of fact-finding or legal analysis”: Pioneer Corp, at para. 104. An issue is considered common where (1) it does not inevitably break down into individual issues, and (2) it represents a substantial and necessary component of each class member’s claim, such that resolving it would meaningfully advance those claims: Fulawka v. Bank of Nova Scotia, 2012 ONCA 443, 112 O.R. (3d) 346, at para. 81, leave to appeal refused, [2012] S.C.C.A. No. 326; Canada v. Greenwood, 2021 FCA 186, [2021] 4 F.C.R. 635, at para. 180, leave to appeal refused, [2021] S.C.C.A. No. 377.
[40] The common issues threshold is intentionally low. As this court explained in Cloud v. Canada (Attorney General), proposed issues may qualify even if they form only “a very limited aspect of the liability question,” and even where many individual issues will remain: at paras. 52-53. They need not predominate over individual questions, nor must they completely determine each class member’s claim: Western Canadian Shopping Centres Inc. v. Dutton, 2001 SCC 46, [2001] 2 S.C.R. 534, at para. 39.
[41] The evidentiary standard — “some basis in fact” — reflects this low bar. This standard requires only minimal evidence showing that the proposed common issues exist and could be addressed class-wide. The appellants need not demonstrate a prima facie case or establish the likelihood of trial success, and the court cannot weigh the evidence or evaluate the merits at certification: Lilleyman v. Bumble Bee Foods LLC, 2024 ONCA 606, 173 O.R. (3d) 682, at paras. 67–77, leave to appeal refused, [2024] S.C.C.A. No. 406.
[42] Specific guidance exists for institutional abuse class actions alleging systemic negligence. In both Rumley v. British Columbia and Cloud, the courts held that questions about duty of care, standard of care, and breach could meaningfully advance class members’ claims even though individualized trials on causation and damages would still be necessary. They further held that whether institutions should have adopted abuse-prevention policies could be determined class-wide without devolving into individualized inquiries: Rumley, at para. 30; Cloud, at paras. 58, 66, 69. Courts must apply these precedents consistently and treat like cases alike: Black v. Owen, 2017 ONCA 397, 173 O.R. (3d) 334, at paras. 41–42, 46; Sriskandarajah v. United States of America, 2012 SCC 70, [2012] 3 S.C.R. 609, at para. 18.
[43] On appeal, extricable questions of law are reviewed for correctness. In the absence of such errors, a motion judge’s decision is reviewed for palpable and overriding error: Pioneer Corp., at para. 94; Pearson v. Inco Ltd. (2006), 2006 CanLII 913 (ON CA), 78 O.R. (3d) 641 (C.A.), at para. 43, leave to appeal refused, [2006] S.C.C.A. No. 1; Fulawka, at paras. 76–77. References to “errors in principle” in some cases are simply a subset of legal error: Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 235, at para. 36.
(2) Application: The Motion Judge Applied an Overly Stringent Test
[44] Although the motion judge accurately summarized the governing principles, he applied them incorrectly in three key ways: (1) he imposed an improperly high bar, (2) he exceeded the limits of the “some basis in fact” standard, and (3) he disregarded binding guidance from Rumley and Cloud. These legal errors justify appellate correction: Housen, at para. 27; Black, at paras. 36, 41. I would not, however, determine whether the appellants meet the correct commonality test because it would not affect the result.
[45] First, the motion judge erred in applying a high threshold. He reasoned that because liability could not be determined without individual trials, the proposed common issues (1) would not significantly advance the litigation, and (2) would inevitably fracture into individual issues. Both conclusions are legally incorrect. As Cloud makes clear, common issues need not establish liability or eliminate individual trials; they need only advance claims in some meaningful way. Moreover, issues are only disqualified if they are inherently dependent on individual circumstances, not simply because individualized inquiries may follow them: Fulawka, at para. 81; Dennis v. Ontario Lottery and Gaming Corporation, 2013 ONCA 501, 116 O.R. (3d) 321, at para. 68, leave to appeal refused, [2013] S.C.C.A. No. 323. It is unnecessary to decide whether the proposed common issues meet these correct thresholds for the reasons I have explained.
[46] Second, the motion judge overstepped the evidentiary limits of certification. In finding that “[n]o one system exists here,” he effectively decided a merits question. That was not his role. The correct inquiry was whether the appellants had offered “some minimal evidence” to support their allegations of systemic negligence: Lilleyman, at para. 74. He did not address why the evidence adduced by the appellants – that each regional league has its own safety policies, the CHL has conducted national reviews, and those reviews repeatedly urged adoption of comprehensive abuse-prevention polices – failed to meet this low threshold. I need not resolve this question either because it would not affect the result.
[47] Third, the motion judge failed to engage with the precedents in Rumley and Cloud. He neither explained why the systemic issues in this case differed from those in which courts had previously certified duty, standard, and breach as common issues, nor reconciled his conclusion with the binding authority that institutional abuse-prevention policy questions can be resolved class-wide because they “affect all class members irrespective of their personal circumstances”: Cavanaugh, at para. 22. As well, he did not address why the solutions employed in Rumley and Cloud – creating subclasses or limiting liability to certain periods – would be inadequate to address his concerns about changes in the standard of care over time: Rumley, at paras. 31-33; Cloud, at para. 59.
[48] The motion judge also did not address Rumley and Cloud’s guidance concerning the availability and quantum of punitive damages. Both cases held that those were proper common issues: Rumley, at para. 34; Cloud, at paras. 70, 72. I note that, as Rumley and Cloud demonstrate, the question of whether punitive damages are available for systemic misconduct can sometimes be resolved without reference to the extent of harm suffered by each individual: Waldman v. Thomson Reuters Corporation, 2012 ONSC 1138, 22 C.P.C. (7th) 33, leave to appeal refused, 2012 ONSC 3436 (Div. Ct.). Any concern about quantifying punitive damages before compensatory damages can usually be addressed by sequencing trials so that punitive damages are assessed after compensatory damages have been determined: Robinson v. Rochester Financial Ltd., 2010 ONSC 463, 89 C.P.C. (6th) 91, at paras. 57–61, leave to appeal refused, 2010 ONSC 1899, 262 O.A.C. 148 (Div. Ct.).
[49] To be clear, I would not decide whether the proposed class action is sufficiently similar to Rumley and Cloud so as to establish commonality. That is an arguable question in light of the differences between those cases and this proposed class action which I address later in this judgment. My point is merely that the motion judge should have meaningfully engaged with those precedents and explained which distinguishing features justified a different result in this case.
|
