|
Crown and Government Liability - Charter. Canada (Attorney General) v. Power [legislative immunity]
In Canada (Attorney General) v. Power (SCC, 2024) the Supreme Court of Canada re-considers and confirms [the previous case was Mackin (SCC, 2002)] whether and how the Crown can be liable for Charter damages for passing unconstitutional legislation:[4] .... The state is not entitled to an absolute immunity from liability for damages when it enacts unconstitutional legislation that infringes Charter rights. Rather, as this Court held in Mackin v. New Brunswick (Minister of Finance), 2002 SCC 13, [2002] 1 S.C.R. 405, the state enjoys a limited immunity in the exercise of its law-making power. Accordingly, damages may be awarded under s. 24(1) for the enactment of legislation that breaches a Charter right. However, the defence of immunity will be available to the state unless it is established that the law was clearly unconstitutional, or that its enactment was in bad faith or an abuse of power. This is a high threshold. But it is not insurmountable.
[5] An absolute immunity fails to properly reconcile the constitutional principles that protect legislative autonomy, such as parliamentary sovereignty and parliamentary privilege, and the principles that require the government be held accountable for infringing Charter rights, such as constitutionality and the rule of law. Each of these principles constitutes an essential part of our constitutional law and they must all be respected to achieve an appropriate separation of powers. By shielding the government from liability in even the most egregious circumstances, absolute immunity would subvert the principles that demand government accountability. The necessary reconciliation of these principles demands that we affirm the limited immunity threshold recognized in Mackin.
....
[17] This appeal raises a single issue: can damages ever be an appropriate and just remedy under s. 24(1) of the Charter for the enactment of legislation later declared unconstitutional?
....
B. This Court Has Recognized a Limited Immunity for the Enactment of Unconstitutional Legislation
[58] We now turn to how this Court’s jurisprudence has addressed these constitutional principles in the context of state immunity for Charter damages. Mr. Power argues that Mackin directly dealt with the question in this appeal. He submits that Mackin rejected an absolute immunity and that we are bound to do the same here. Canada, however, argues that Mackin is not authoritative on this question and, in any case, subsequent cases have displaced the limited immunity established in Mackin.
(1) Mackin Set a High Threshold for Damages
[59] We agree with Mr. Power and the courts below that in Mackin, this Court considered precisely the same issue: the availability of Charter damages for the enactment of Charter-infringing unconstitutional legislation.
[60] The applicants in Mackin were two provincial court judges. They challenged a provincial statute that eliminated the system of supernumerary judges in favour of a panel of retired judges paid on a per diem basis. They claimed that the statute violated the right to judicial independence enshrined in s. 11(d) of the Charter and sought s. 24(1) damages and s. 52(1) declaratory relief.
[61] The Court agreed with the judges that the legislation was unconstitutional and declared it to be of no force or effect. Turning to the claim for damages, Gonthier J. explained that while legislative bodies enjoy immunity from damages for the “mere enactment or application of a law that is subsequently declared to be unconstitutional” (para. 78 (emphasis added)), such immunity will give way to liability when the law was “clearly wrong, in bad faith or an abuse of power” (para. 79 (emphasis added)). He concluded that damages were not justified in that case because there was no evidence to suggest that this threshold was satisfied (para. 82).
[62] In our view, the following three points can be distilled from Mackin.
[63] First, Mackin concerned only the enactment of legislation. The state’s actions under the law was not at issue: it was the legislation itself that abolished the supernumerary judicial status, thereby depriving the judges of their status and income — and thus, their independence. Accordingly, we do not accept Canada’s argument that Mackin never contemplated a claim for damages against the state for the enactment of unconstitutional legislation. The Mackin principle of limited immunity was plainly set out in the context of the “enactment” of law, and as applying to “legislative bodies” (para. 78).
[64] Second, the Court did not deny the judges’ claim because the state enjoyed an absolute immunity for the enactment of legislation, but because the threshold was not satisfied. Gonthier J. found that there was no evidence to suggest the government of New Brunswick acted in bad faith or abused its power. He explained that the government could not have known of later developments in the law concerning judicial independence at the time the statute was enacted. In reaching this conclusion, Gonthier J. looked to the state’s motives, knowledge and conduct during the legislative process. He found the state had a “perfectly legitimate purpose” when it passed the invalid law: “efficiency, flexibility and cost savings” (para. 70). He also found that there was no evidence the state acted with “wilful blindness with respect to its constitutional obligations” (para. 82), nor that the state enacted the law with “knowledge” of its unconstitutionality, nor for an “ulterior motiv[e]” (para. 83).
[65] Third, the threshold for liability established in Mackin was expressly designed to reconcile competing constitutional principles. Gonthier J. explained that the “limited immunity given to government is specifically a means of creating a balance between the protection of constitutional rights and the need for effective government” (para. 79). To award damages merely because legislation was unconstitutional would fail to strike the right balance because it would not give effect to the need for effective government. Yet, at the same time, he recognized that “the government and its representatives are required to exercise their powers in good faith and to respect the ‘established and indisputable’ laws that define the constitutional rights of individuals” (ibid.).
[66] For these reasons, Gonthier J. held that the “clearly wrong, in bad faith or an abuse of power” threshold provides the appropriate qualification on the state’s liability for the enactment of an unconstitutional law.
(2) The Mackin Threshold Has Not Been Overruled by Subsequent Cases
[67] Canada further argues that if Mackin established a limited immunity, this Court has overruled that immunity in subsequent cases.
[68] We do not agree. Rather, we agree with Mr. Power and the courts below that the post-Mackin jurisprudence does not depart from the limited immunity threshold.
[69] The first category of cases said to have overruled Mackin are those Canada says demonstrate that the limited immunity threshold in Mackin has been restricted in application to situations of executive action, such that it no longer applies in the realm of legislative power. We reject this argument.
[70] True, some of this Court’s cases have referred to the threshold applying to executive conduct under the law. For example, in Ward, McLachlin C.J. referred to Mackin as applying to “state conduct under the law” (para. 39). And in Henry v. British Columbia (Attorney General), 2015 SCC 24, [2015] 2 S.C.R. 214, Moldaver J. referred to it as applying to “state action taken pursuant to a law” (para. 42). But those cases involved such executive conduct. There is no inconsistency here. As explained above, Gonthier J. said in Mackin that the threshold applied to the “enactment or application” of legislation later declared unconstitutional (para. 78 (emphasis added)). Indeed, the Mackin threshold was described in general terms in Ward as recognizing “that the state must be afforded some immunity from liability in damages resulting from the conduct of certain functions that only the state can perform. Legislative and policy-making functions are one such area of state activity” (para. 40 (emphasis added)).
[71] Similarly, in Conseil scolaire francophone de la Colombie-Britannique v. British Columbia, 2020 SCC 13, [2020] 1 S.C.R. 678, this Court considered a different type of situation again: whether the Mackin threshold applied to government decisions made under government policies. Canada points out that at one point the Mackin threshold is described as applying to “acts carried out pursuant to a law that is subsequently declared to be invalid” (para. 168). But, at another point, the threshold is described as applying to “those who make laws” or “the legislature” (ibid.). Both are consistent with Mackin, and the Court was unanimous on this point. Although dissenting in the result, Brown and Rowe JJ. noted that “a key holding” in Mackin was its broad formulation of the immunity threshold and that it “plainly encompasses acts of both the legislatures and other public officials” (paras. 286-87 (emphasis deleted)).
[72] Canada also argues that the principles set down in Mikisew overrule Mackin. We do not agree. Canada is right that Mikisew says that courts should not meddle with the law-making process, including in the enactment of legislation. However, Mikisew is readily distinguishable. This appeal concerns remedies for an invalid law that breaches Charter rights. Mikisew concerned whether to impose a procedural step in the form of a duty to consult under s. 35 of the Constitution Act, 1982 within the legislative process (paras. 31 and 52). And, while the majority in Mikisew accepted that it is “rarely appropriate for courts to scrutinize the law-making process” (para. 2), and that the judiciary “should forebear from intervening” in this process (para. 32), it also noted that after-the-fact review remains available in some circumstances (para. 52).
[73] A Charter damages analysis under s. 24(1) is conceptually distinct from the recognition of a duty to consult. As the motion judge and the Court of Appeal in this case explained, there is an important difference between, on the one hand, courts requiring the legislature to implement a substantive step within the legislative process, such as the pre-enactment consultations contemplated in Mikisew, and on the other, courts enforcing the Charter by requiring the state to pay damages for a Charter violation, even when that violation results from an unconstitutional law (motion judge’s reasons, at paras. 54-55; C.A. reasons, at para. 23). Compelling the government to consult as part of the legislative process would be a clear interference with Parliament’s exclusive control over its own procedures. By contrast, post-enactment damages do not “unduly interfer[e]” with Parliament, including its control over its own procedures (Mikisew, at para. 35, citing Criminal Lawyers’ Association, at para. 29). Such damages do not compel the legislature to regulate its own internal affairs in a certain way. Mikisew does not determine the outcome of this appeal.
[74] Finally, Canada raises Ernst v. Alberta Energy Regulator, 2017 SCC 1, [2017] 1 S.C.R. 3, as an example of the approach the Court should take in this case. In Ernst, this Court held that an administrative board, in exercising its adjudicative function, is immune from liability for damages. Ernst is of limited assistance to this appeal. Although Ernst engaged some of the same constitutional principles at work in this case, it did so in a different context. Ernst was concerned with protecting a different state function and implicated a different balance of underlying constitutional principles. An assessment into immunity must focus on the branches of government implicated by the claim (K. Cooper-Stephenson, Charter Damages Claims (1990), at p. 316). It is not surprising that a different form of state action raised different concerns about constitutional design and institutional relationships. Nothing in Ernst suggests a retreat from Mackin.
[75] Accordingly, we cannot accept Canada’s submission that the Mackin threshold has been overruled.
C. Mackin Should Not Be Overruled
[76] Canada’s final argument is that if this Court has not overruled Mackin already, it should do so now. Canada submits that the constitutional principles underpinning legislative autonomy and good governance require absolute immunity for the enactment of legislation subsequently declared unconstitutional. Mr. Power responds that an absolute immunity would be inconsistent with the other fundamental constitutional principles engaged by the state’s law-making function.
[77] We agree with Mr. Power. Canada has not provided a compelling reason to overrule a precedent of this Court. Accordingly, we would not overturn Mackin. Limited immunity respects the constitutional principles underpinning both legislative autonomy and accountability. As this Court has said before, effective government and respect for constitutional rights are both “important pillars of our democracy” (Ernst, at para. 25). In order to fulfill its institutional function, the legislative branch requires an independent space for elected representatives to carry out their parliamentary duties, to freely debate and decide what laws should govern, and to exercise the unfettered ability to hold the executive branch of the state to account. But absolute immunity would subvert the principles that command government compliance with the Charter and the courts’ role in enforcing its fundamental guarantees.
....
[93] Fourth, an absolute immunity does not accommodate the principles recognized in this Court’s jurisprudence on constitutional remedies. It leaves little room for the principles that underpin legislative accountability — including the broad and purposive approach to rights and remedial provisions in the Charter, as well as constitutionalism and the rule of law. All these principles militate against absolute immunity. We agree in this respect with Mr. Power, who submits that “[a]n absolute immunity is . . . incompatible with the remedial discretion of the courts — ‘a fundamental feature of the Charter’ — and with the idea that ‘flexibility is necessary to arrive at appropriate remedies involving legislation’” (R.F., at para. 81, citing G, at paras. 101 and 146).
[94] As discussed above, the Charter effected a “revolutionary transformation of the Canadian polity” under which courts were “mandated to bring the entire legal system into conformity with a complex new structure of rights-protection” (L. E. Weinrib, “Canada’s Charter of Rights: Paradigm Lost?” (2002), 6 Rev. Const. Stud. 119, at p. 120). Even before the Charter, the court’s role in holding the legislature accountable was recognized as part of the fabric of Canada’s constitutional order. As Dickson J. (as he then was) explained in Amax Potash Ltd. v. Government of Saskatchewan, 1976 CanLII 15 (SCC), [1977] 2 S.C.R. 576, at p. 590:A state, it is said, is sovereign and it is not for the Courts to pass upon the policy or wisdom of legislative will. As a broad statement of principle that is undoubtedly correct, but the general principle must yield to the requisites of the constitution in a federal state. By it the bounds of sovereignty are defined and supremacy circumscribed. The Courts will not question the wisdom of enactments which, by the terms of the Canadian Constitution are within the competence of the Legislatures, but it is the high duty of this Court to insure that the Legislatures do not transgress the limits of their constitutional mandate and engage in the illegal exercise of power. [Emphasis added.]
(See also Manitoba Language Rights, at p. 745.) [95] The Charter demands that legislative power be constrained by constitutional rights. Courts are constitutionally obliged to hold the government accountable when it breaches such rights, including by providing meaningful remedies in the face of their violation. An absolute immunity would undermine the purpose and text of s. 24(1), which asks courts to look at the specific context of a given violation to determine whether a remedy is appropriate and just. The Charter requires courts to enforce constitutional rights. Enforcement means ensuring that remedies are commensurate with the extent of the violation (Corbiere, at para. 46). In this way, the separation of powers also protects the judiciary’s independence to carry out its constitutional duties: “Nothing less [is] required to maintain the normative ordering of the Canadian legal system” (Doucet-Boudreau, at para. 109).
[96] An absolute immunity would protect the government from any claim for damages for any unconstitutional legislation, no matter how egregious. We accept Mr. Power’s assertion that an absolute immunity allows a narrow set of constitutional interests to dominate the analysis (R.F., at para. 81).
[97] In setting a high bar for immunity, Mackin has stood the test of time. It has been good law for over two decades. It has resulted neither in chilling good governance, nor in a floodgate of claims against the state for damages. The state will continue to benefit from immunity unless the high threshold is satisfied. This exacting threshold functions to limit the scope of causes of action for damages. And, as always, the state can apply for a s. 24(1) claim to be dismissed summarily if the claimant fails to plead circumstances which could, if accepted, satisfy the Mackin threshold for liability (Henry (2015), at para. 43).
....
(2) Clarifying the Mackin Threshold
[99] The question remains how to best articulate the qualified immunity threshold. The parties and interveners point to a number of different descriptions within Mackin and ask us to clarify the threshold.
[100] In Mackin, Gonthier J. used several formulations to describe the circumstances in which damages may be an appropriate and just remedy for legislation later declared unconstitutional. Gonthier J. initially sets out the threshold as capturing state conduct that was “clearly wrong, in bad faith or an abuse of power” (paras. 78‑79). Later, in applying the threshold, he looked to whether the state enacted the unconstitutional law “negligently”, or with “wilful blindness with respect to its constitutional obligations at that time” (para. 82). He concluded there was no evidence “the legislation was enacted wrongly, for ulterior motives or with knowledge of its unconstitutionality” (para. 83).
[101] These different formulations reflect different lines of inquiry aimed at assessing the context of the enactment of the Charter-infringing law in that case. Gonthier J. focused on the state of the law at the time of enactment, and on any evidence of the state’s motives, conduct, and knowledge in relation to its unconstitutionality.
[102] Gonthier J.’s various formulations, with the exception of negligence, describe instances in which the enactment would have been clearly wrong, in bad faith or an abuse of power. In our view, however, negligence does not connote the gravity of misconduct this threshold was designed to capture. While Gonthier J.’s statements indicate that there was no evidence that would meet even this lower standard in that case, in our view, the use of the term negligence was not intended to lower the general threshold. It does not form part of the threshold. The concept of “negligence” is unhelpful as it does not reflect the high standard demanded by the constitutional principles underlying the analysis.
[103] Nonetheless, an objective assessment of the unconstitutionality of the legislation can assist in identifying whether the threshold is met, provided the standard remains high. Indeed, like negligence, the French translation in Mackin of “clearly wrong” — “clairement fautif” — signals such an objective standard. A number of interveners favoured a threshold that focussed on an objective review of the legislation under the “clearly wrong” component of the threshold. Other interveners point out that the concept of “wrongfulness” has led to some confusion. We would clarify that this inquiry is better understood as a focus on whether the legislation is “clearly unconstitutional”, which directs a judge to look objectively at the legislation itself, particularly the nature and extent of its constitutional invalidity. Underlying this objective assessment is a presumption of the legislature’s knowledge of and respect for basic Charter rights.
[104] However, we would reject any of the formulations suggested by the interveners that would set the threshold so high that it immunizes the government from liability for unprecedented but egregious constitutional breaches. Thus, the threshold will be met where the legislation was “clearly unconstitutional” in the sense that, at the time of its enactment, it would clearly violate Charter rights (M. L. Pilkington, “Monetary Redress for Charter Infringement”, in R. J. Sharpe, ed., Charter Litigation (1987), 307, at pp. 319-20, cited with approval in Guimond v. Quebec (Attorney General), 1996 CanLII 175 (SCC), [1996] 3 S.C.R. 347, at para. 15; R. J. Sharpe and K. Roach, The Charter of Rights and Freedoms (7th ed. 2021), at p. 511). Such egregious or obvious violations of Charter rights are clearly wrong. We would not set the test as high as Jamal J.’s proposed articulation of the clearly unconstitutional threshold.
[105] A finding of clear unconstitutionality will usually imply that the state either knew that the law was clearly unconstitutional, or was reckless or wilfully blind as to its unconstitutionality. As Sharpe and Juriansz JJ.A. helpfully explained in Brazeau, where the law is clearly unconstitutional, the state may have shown a “‘clear disregard’ for Charter rights” by “proceeding with a course of action in the face of a known risk that the Charter will be violated or by deliberately failing to inquire about the likelihood of a Charter breach when the state knows that there is a good reason to inquire” (para. 87, citing Ward, at para. 43).
[106] While the clearly unconstitutional standard will likely resolve most issues of whether the limited immunity applies, other rare situations may require judges to ask whether there is evidence that the state acted in bad faith or abused its power in enacting the invalid law. Principles of constitutionalism and legality require that the threshold leaves room for meaningful recourse for breach of constitutional rights involving such intentional state misconduct.
[107] We would not attempt to define bad faith or abuse of power in the law‑making process with exactitude without the benefit of a full record and submissions. This standard may, for example, be met in cases where the state acted for an improper purpose, or was dishonest. We would not, however, limit the concepts of bad faith and abuse of power to an examination of the legislation’s purpose.
[108] When it comes time to assess an allegation of bad faith or abuse of power on the basis of specific facts, other contexts of bad faith and abuse of power may provide guidance (see, e.g., Roncarelli, at p. 141; Finney v. Barreau du Québec, 2004 SCC 36, [2004] 2 S.C.R. 17, at para. 39; Hinse v. Canada (Attorney General), 2015 SCC 35, [2015] 2 S.C.R. 621, at paras. 48-53; see also Special Lectures of the Law Society of Upper Canada 1979 — The Abuse of Power and the Role of an Independent Judicial System in Its Regulation and Control (1979)). However, we note that these concepts are “flexible” and their “content will vary from one area of law to another” (Entreprises Sibeca Inc. v. Frelighsburg (Municipality), 2004 SCC 61, [2004] 3 S.C.R. 304, at para. 25). In the context of the state’s law-making function, in which Parliament and legislative bodies are entitled to enact any law within their constitutional confines, bad faith and abuse of power may require a higher degree of misconduct than in other contexts. These are legal standards applied by courts, they are not means of evaluating the wisdom or policy of the enactment process or the enacted law.
....
[110] We appreciate that discerning institutional motivation or the knowledge of legislative bodies when enacting legislation is a difficult task. We further appreciate that although Charter damages lie against the state and not individual lawmakers or government officials, the state acts through the vehicle of individuals. As with other contexts of institutional state conduct, whether it is possible to attribute the bad faith or abuse of power of an individual or group to the institution itself will depend on the facts of a given case. It bears repeating here that the basis for liability under s. 24(1) is the state’s breach of a Charter right. In cases like Mr. Power’s claim, it is the invalid law that breached his right. Any inquiry into state misconduct in enacting the invalid law is to assess whether damages are just and appropriate for that breach, not to create an independent basis for liability.
[111] Where the claimant puts forward a particularized allegation that the Mackin threshold has been met, the claim must be assessed on the basis of evidence obtained in a way that does not violate parliamentary privilege, such as statements made outside of the parliamentary process.
[112] Thus, we would clarify that “clearly wrong” reflects an objective assessment into whether the legislation was clearly unconstitutional at the time it was enacted, and that bad faith and abuse of power remain part of the threshold. To that extent we would restate the Mackin threshold relating to the enactment of legislation later found to be unconstitutional: the good governance defence will prevail unless the law was clearly unconstitutional, in bad faith or an abuse of power. The exacting nature of the threshold means that an applicant’s failure to provide detailed particulars will be fatal to their claim at the pleadings stage (Henry (2015), at para. 43). Bald or vague assertions will necessarily fall short.
|
