|
Appeals - SOR - Errors of Mixed Fact and Law. Fair Voting BC v. Canada (Attorney General)
In Fair Voting BC v. Canada (Attorney General) (Ont CA, 2025) the Ontario Court of Appeal dismisses (IMHO unnecessarily harshly) a Charter s.3 ['democratic rights'] and s.15 ['discrimination'] challenge to the federal electoral system.
Here the court notes the obvious point that in an appeal, deference still applies wrt non-legal fact issues (even constitutional/charter ones) - though points out (interestingly) recent SCC doctrine that varies from that wrt mixed fact and law, where a 'correctness' standard has been applied [see para 82-83] (albeit in a JR constitutional context):The limits of deference
[74] The appellants’ s. 15(1) argument can be dismissed on the basis that the application judge’s decision is entitled to deference. His conclusion that causation was not established can be characterized as a finding of fact based on his review of the expert social science evidence. The appellants have not established that the application judge made any palpable and overriding errors that would allow this court to intervene.
[75] The decision to uphold the application judge’s decision on this basis follows the Supreme Court’s instruction in Bedford as to how social science findings are to be treated on appeal. But we should be clear about what this means: that the constitutionality of the federal electoral system rests on a requirement to defer to the findings of a single judge concerning highly contestable social science evidence.
[76] Consider a counterfactual. Suppose that, instead of finding that causation was not established, the application judge made the findings that the appellants submit he should have: that PR systems moderate the effect of sexist attitudes and make it easier for women to get elected; that SMP disadvantages geographically dispersed minority and women voters and candidates; and that New Zealand’s experience with PR is a useful model for Canada – all findings that, according to the appellants, were supported by the expert social science evidence.
[77] These too would be findings to which this court would be expected to defer, but they would support the opposite result: a finding that the SMP electoral system causes a disproportionate impact on women and racial minorities. Given this finding, it would not be difficult to establish that this disproportionate impact has the effect of reinforcing, perpetuating, or exacerbating the disadvantage of women and racial minorities in politics. The result would be a conclusion that the electoral system infringes s. 15(1) of the Charter.
[78] It is difficult to accept that a conclusion that the federal electoral system is discriminatory should depend on the social science findings of a single judge. And yet, that is the result of the Supreme Court’s instruction in Bedford, at paras. 48-56, that appellate courts are to defer to social science findings made at trial.
[79] The problem is not simply that findings of fact based on highly contestable social science evidence are entitled to deference; it is that those findings may essentially determine the alleged Charter infringement. And even to the extent that findings can be characterized as a mixed question of fact and law, Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 235 instructs that the standard of review remains palpable and overriding error – although as I discuss below, there is now some uncertainty in this regard.
[80] The nature of the problem is put in sharp relief by the application judge when he says: “[S]ince the evidence does not establish that implementing PR in Canadian elections would do any better than SMP, the section 15 claim is not made out.” I do not see why the constitutionality of the federal electoral system should depend on this sort of judgment, still less why this sort of judgment should be entitled to deference in this court. How could evidence establish that a foreign electoral system would deliver “better” results than the SMP electoral system? At the end of the day it is academic conjecture about how a different electoral system would operate in Canada. It is usefully discussed in academic and policy development settings, but it is hardly the stuff of constitutional law.
[81] The application judge seems to recognize as much, stating that there was nothing in the record to suggest, let alone prove, that Canada would follow New Zealand’s lead and so achieve similar outcomes if PR were adopted. And yet he strained to reject the relevance of New Zealand’s experience with PR, pointing to relatively minor differences between Canada and New Zealand, including New Zealand’s distinctive approach to tort reform, its smaller population and land mass, and its climate and geography. He could just as easily have noted more salient similarities between the two countries, including the fact that both are Westminster parliamentary democracies with common law legal orders complete with a bill of rights. And had he chosen to do so, he could just as easily have concluded that the representation of women and racial minorities would improve if PR were adopted in Canada, and that the appellants had established discrimination under s. 15(1).
Uncertainty in deference doctrine
[82] This appeal was argued on the basis that the application judge’s legal conclusions on ss. 3 and 15(1) were reviewable for correctness, but that his underlying factual findings were entitled to deference. The parties made no submissions concerning the relevance, if any, of the Supreme Court’s decision in Société des casinos du Québec inc. v. Association des cadres de la Société des casinos du Québec, 2024 SCC 13, 491 D.L.R. (4th) 385. In that case, the majority of the court endorsed Côté J.’s concurring reasons, which held that findings of mixed fact and law made in connection with a constitutional question are subject to review for correctness. Only findings of “pure” fact that can be isolated from the constitutional analysis are entitled to deference: see paras. 45, 92-97.
[83] The constitutional question in Société des casinos arose on an appeal from a judicial review and was decided in accordance with the framework established in Canada (Minister of Citizenship and Immigration) v. Vavilov, 2019 SCC 65, [2019] 4 S.C.R. 653. Neither Housen nor Bedford was mentioned by the court, nor did the court acknowledge that it was altering the approach established by those cases and others. Nevertheless, two decisions in this court have assumed from Société des casinos that the correctness standard also applies to findings of mixed fact and law arising from civil constitutional appeals: see R. v. Pike, 2024 ONCA 608, 171 O.R. (3d) 241, at para. 31, and Jacob v. Canada (Attorney General), 2024 ONCA 648, 172 O.R. (3d) 721, at para. 53, leave to appeal refused, [2024] S.C.C.A. No. 488.
[84] Two other decisions from this court, however, applied the palpable and overriding error standard to findings of fact relevant to the legal question at issue: Ontario English Catholic Teachers Association v. Ontario (Attorney General), 2024 ONCA 101 at para. 51 and Amalgamated Transit Union, Local 113 v. Ontario, 2024 ONCA 407, 172 O.R. (3d) 571, at paras. 31-33. Dissenting in the Amalgamated Transit Union case, Nordheimer J.A. would have applied the approach in Société des casinos and reviewed the mixed findings below for correctness: at paras. 190-94. There also appears to be disagreement on the Supreme Court as to whether the decision in Société des casinos extends to civil constitutional appeals. In their dissenting opinion, Chief Justice Wagner and Justice Moreau cited Housen in asserting that findings of mixed fact and law are reviewable only for palpable and overriding error: at paras. 108-9.
[85] This lack of clarity in the law has attracted academic commentary. See Anthony Sangiuliano & Mark Friedman, “What is the Standard of Review for (Mixed) Constitutional Questions?” (U.B.C. L. Rev., forthcoming). It is not necessary to weigh in on the debate for purposes of this appeal; neither party on appeal has characterized the application judge’s findings as findings of mixed fact and law. It is enough to note that the law requires clarification, not least because the distinction between findings of pure fact and findings of mixed fact and law – where both are made in connection with a constitutional question – may be subtle at best. That is certainly the case here. . Patel v. Dermaspark Products Inc.
In Patel v. Dermaspark Products Inc. (Fed CA, 2025) the Federal Court of Appeal dismissed an appeal, here where the lower court "decided that the appellants were liable, jointly and severally, in the amount of $45,000, representing statutory damages of $5,000 for copyright infringement, $20,000 for trademark infringement, passing off, depreciation of goodwill and unfair competition, and $20,000 for punitive damages".
Here the court considers (perhaps critically) the appellate 'mixed questions of fact and law' category with it's 'palpable and overriding' standard of review:[6] Appellate courts bandy around the phrase "“questions of mixed fact and law”" but seldom define it, much to the prejudice of young practitioners, most of whom graduated from law schools that did not instruct them on the standard of review. For their benefit, questions of mixed fact and law are those where appellate courts apply the law to the facts of the case. This includes so-called discretionary questions where courts apply legal standards to a set of facts: Hospira Healthcare Corporation v. Kennedy Institute of Rheumatology, 2016 FCA 215, [2017] 1 F.C.R. 331, citing Decor Grates Incorporated v. Imperial Manufacturing Group Inc., 2015 FCA 100, [2016] 1 F.C.R. 246 at paras. 15-29; see also Mahjoub v. Canada (Citizenship and Immigration), 2017 FCA 157, [2018] 2 F.C.R. 344 at paras. 71-72.
[7] In some cases, legal questions predominate or fundamentally taint the question of mixed fact and law. In the parlance of appellate standards of review, this is called "“an extricable question of law”". When there is an extricable question of law, the appellate court can examine that question of law and decide it on a standard of correctness—i.e., without any deference at all to the first-instance court.
[8] But where legal questions are not extricable, i.e., do not predominate or fundamentally taint the question of mixed fact and law—in other words, where the question of mixed fact and law is factually suffused or the facts predominate—the appellate court can interfere only for palpable and overriding error.
[9] Palpable means obvious. And overriding means capable of changing the result of the case. As a practical matter, these two things very seldom happen together. First-instance judges almost never make obvious factual errors that can change the result of the case. Thus, reversal on this ground is rare indeed.
[10] On occasion, some have said that a lower standard, such as "“unsafe verdict”", might have been better in furthering accountability and high-quality decision-making. In some countries, that is the standard. But that is not our standard. The Supreme Court decided upon "“palpable and overriding error” "as our standard. It has kept that standard almost for a quarter-century.
[11] Truly, palpable and overriding error is a tough standard:Palpable and overriding error is a highly deferential standard of review: H.L. v. Canada (Attorney General), 2005 SCC 25, [2005] 1 S.C.R. 401; Peart v. Peel Regional Police Services (2006), 2006 CanLII 37566 (ON CA), 217 O.A.C. 269 (C.A.) at paragraphs 158-59; Waxman, [2004 CanLII 39040 (ON CA), 186 O.A.C. 201 at paragraphs 278-84]. “Palpable” means an error that is obvious. “Overriding” means an error that goes to the very core of the outcome of the case. When arguing palpable and overriding error, it is not enough to pull at leaves and branches and leave the tree standing. The entire tree must fall. (Canada v. South Yukon Forest Corporation, 2012 FCA 165, 4 B.L.R. (5th) 31 at para. 46, adopted by the Supreme Court in Benhaim v. St‑Germain, 2016 SCC 48, [2016] 2 S.C.R. 352 at para. 38.)
[12] Later cases have clarified that the palpable and overriding error standard can be met not only by "“one decisive chop”" at the "“tree”" but by "“several telling ones”": Mahjoub at paras. 64-65.
[13] Examples of things that can qualify under this difficult-to-meet standard include a number of different types of errors: "“obvious illogic in the reasons (such as factual findings that cannot sit together), findings made without any admissible evidence or evidence received [not] in accordance with the doctrine of judicial notice [""R. v. Spence, ""2005 SCC 71"", ""[2005] 3 S.C.R. 458""], findings based on improper inferences [""Pfizer Canada Inc. v. Teva Canada Limited, ""2016 FCA 161"", ""400 D.L.R. (4th) 723"" at paras. ""168-170]"" or logical error, and the failure to make findings due to a complete or near-complete disregard of evidence”": Mahjoub at para. 62. But, as said before, only errors on central points that can change the result of the case will qualify.
[14] This discussion is not meant to suggest that the judgment of the Federal Court in this case survives only because the palpable and overriding error standard is hard to meet. But it is to suggest that many of the appellants’ submissions in this Court — vigorously argued and gamely pursued—run straight into this unforgiving and uncompromising standard. As a result, they must fail. . Kantoor v. City of Hamilton
In Kantoor v. City of Hamilton (Ont Divisional Ct, 2025) the Divisional Court allowed an appeal from a Small Claims Court action, and here comments on the appellate SOR for issues of 'mixed fact and law' with a novel perspective ('along a spectrum'):[8] I agree with my colleague, Justice LeMay, that the applicable standard of review is correctness on a question of law, palpable and overriding error on a question of fact, and somewhere along that spectrum on a question of mixed fact and law. Covenoho v. HomeLife Response Realty Inc., 2022 ONSC 5877, at paragraph 26, referring to the decision of the Supreme Court of Canada in Housen v. Nikolaisen, 2022 SCC 33. . Smith v. Gega
In Smith v. Gega (Div Court, 2023) the Divisional Court applied a novel, functional determination of when an issue was one of 'mixed fact and law':[19] This issue is a matter of law. The Landlords argue that this issue is a matter of mixed fact and law, but there are no factual findings that need to be made by me to consider this question. . Abbott v. Canada (Attorney General)
In Abbott v. Canada (Attorney General) (Fed CA, 2021) the Federal Court of Appeal set out the appellate standard of review where the judicial review application generated fresh findings at the first instance:[17] The standard set out in Agraira v. Canada (Public Safety and Emergency Preparedness), 2013 SCC 36, [2013] 2 S.C.R. 559 usually applies to an appeal from the judgment of a first instance court in an application for judicial review. However, this Court has held that "“where the application judge made findings of fact or mixed fact and law based on the consideration of evidence at first instance, rather than on a review of the administrative decision, these findings are reviewable on the Housen standard”" (Apotex Inc. v. Canada (Health), 2018 FCA 147, 157 C.P.R. (4th) 289, para. 57; Sturgeon Lake Cree Nation v. Hamelin, 2018 FCA 131, 424 D.L.R. (4th) 366, paras. 36-37). This is effectively the case here. Therefore, the appellate standard of palpable and overriding error applies to the Judge’s findings of fact, or mixed fact and law, regarding the respondents’ authorization to issue the directive.
|
