Democracy Watch v. Canada (Prime Minister), 2023 FCA 41

Date: 20230306

Docket: A-54-22

Citation: 2023 FCA 41

Heard at Ottawa, Ontario, on February 13, 2023.

Judgment delivered at Ottawa, Ontario, on March 6, 2023.

Date: 20230306

Docket: A-54-22

Citation: 2023 FCA 41

REASONS FOR JUDGMENT

LEBLANC J.A.

[1] This is an appeal of an order of the Federal Court (per Zinn J.) dated February 22, 2022. This order struck the appellants’ judicial review application (the Application) challenging the decision of the Prime Minister of Canada made on August 15, 2021 by way of Order in Council 2021-0892 issued by the Committee of the Privy Council, advising the Governor General to dissolve Parliament and call a general election.

[2] The appellants claim that the Prime Minister’s advice to the Governor General violates section 56.1 of the Canada Elections Act, S.C. 2000, c. 9 (the Act).

[3] Section 56.1 of the Act reads as follows:

[4] According to the appellants, that provision prohibits the calling of an election before the fixed election date set out in subsection 56.1(2) of the Act unless, consistent with an emerging unwritten constitutional convention stemming from the 2011, 2015 and 2019 general elections, a vote of non-confidence occurs in Parliament before that fixed date. They contend that this new constitutional convention prohibits election advice driven for purely partisan electoral advantage and that section 56.1 of the Act must now be read accordingly.

[5] The election called by the Governor General on the Prime Minister’s advice was held on September 20, 2021, whereas the fixed election date under subsection 56.1(2) of the Act was October 16, 2023.

[6] The respondents moved for an order striking the Application on the basis that it was moot, bereft of any chance of success, and an abuse of the Court’s process. The respondents also argued that the appellants lacked the requisite standing.

[7] After setting out the test applicable on motions to strike, the Federal Court determined that the Application lacked legal merit and was doomed to fail because this Court had decided the same issues raised therein in Conacher v. Canada (Prime Minister), 2010 FCA 131, [2011] 4 F.C.R. 22, leave to appeal to the Supreme Court of Canada dismissed (2011 CanLII 2101 (SCC)) (Conacher). Conacher concerned the advice given by the Prime Minister to the Governor General, on September 7, 2008, to dissolve Parliament and set a polling date for October 14, 2008. This was a year or so after the Act had been amended to include section 56.1.

[8] The Federal Court considered, but rejected, the appellants’ contention that the factual and legal matrix underlying the present matter is different from the one in Conacher. Aside from the fact that the new constitutional convention asserted by the appellants was “far from established”, it determined that Conacher clearly established that section 56.1 of the Act, “as drafted, [did] not affect the Prime Minister’s ability to give advice to the Governor General” (Conacher at para. 7). It further held that, to the extent that constitutional conventions were relevant in deciding whether to call an election, they were only relevant insofar as “in the Governor General’s opinion, [they] may bear upon or determine the matter” (Reasons for Order at paras. 17-18 [my emphasis]).

[9] The Federal Court also considered, but rejected, the appellants’ reliance on the UK Supreme Court decision in R (Miller) v. The Prime Minister, [2019] UKSC 41 (Miller) as forming part of the new legal landscape underlying the present matter. In addition to noting that the relevance of that judgment to the facts at issue in the present matter was “far from clear”, the Federal Court held that however persuasive said judgement might be, it was bound by Conacher (Reasons for Order at para. 21).

[10] Having determined that the issues raised in the Application were fully resolved by Conacher, the Federal Court declined to rule on the question of mootness.

[11] Decisions made on motions to strike are discretionary in nature (Lafrenière v. Canada (Attorney General), 2020 FCA 110 at para. 2; Feeney v. Canada, 2022 FCA 190 at para. 4). They are subject, on appeal, to the standards of review set out in Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 23 (Housen); (Imperial Manufacturing Group Inc. v. Decor Grates Incorporated, 2015 FCA 100, [2016] 1 F.C.R. 246 at para. 29; Hospira Healthcare Corporation v. Kennedy Institute of Rheumatology, 2016 FCA 215, [2017] 1 F.C.R. 331 at para. 28; Bewsher v. Canada, 2020 FCA 216 at para. 7). This means that the Court will only intervene in such matters if it is satisfied that the Federal Court erred on a question of law or committed a palpable and overriding error on a question of fact or of mixed fact and law (Housen, at paras. 8, 10 and 36-37).

[12] Here, I see no such error on the part of the Federal Court, although the respondents urge us to also dismiss the appeal on the basis that the Application is moot.

[13] The appellants reassert before us that the facts and the legal landscape underling the present matter are different from those in Conacher and that, therefore, it was an error on the part of the Federal Court to strike the Application on the ground that it is a relitigation of Conacher. As indicated above, what is different now, according to the appellants, is that there are “new constitutional conventions or other appropriate matters” for the Prime Minister to consider in his advice to the Governor General under section 56.1 of the Act. Amongst those “other appropriate matters” is the UK Supreme Court decision in Miller as well as new evidence of the will of Parliament, which, the appellants claim, must now inform the interpretation and application of section 56.1 of the Act.

[14] These contentions are without merit.

[15] First, I agree with the respondents that the Application simply repackages what was being advanced in Conacher. Despite the appellants’ claim that the scope of their claim is more circumscribed in this case than it was in Conacher, there is nothing therein that fundamentally changes the nature of the debate raised in Conacher.

[16] As is apparent from the judgment of the Federal Court in Conacher v. Canada (Prime Minister), 2009 FC 920, [2010] 3 F.C.R. 411 (Conacher FC), Democracy Watch there was seeking a declaration that the Prime Minister had violated section 56.1 of the Act by advising the Governor General to call an election on a date that was not the one set out in subsection 56.1(2) of the Act (Conacher FC at para. 2). Democracy Watch argued that section 56.1 had crystallized a new convention, which required the Prime Minister to exercise his advisory authority only “in accordance with subsection 56.1(2), or in a situation of a vote of non-confidence” (Conacher FC at para. 13 [my emphasis]).

[17] This is exactly the position put forward by the appellants in the present matter. As described in their notice of application, the appellants claim that the Prime Minister, in advising as he did the Governor General on August 15, 2021, contravened section 56.1 of the Act, “which requires that the next federal election be held on the third Monday in October 2023 unless, under the unwritten constitutional ‘confidence convention’ that underlies section 56.1, a vote of non-confidence occurs in Parliament before that October 2023 fixed election date” (Appeal Book at 22).

[18] As indicated previously, this Court in Conacher held “that section 56.1, as drafted, does not affect the Prime Minister’s ability to give advice to the Governor General” and “leaves the Prime Minister and the Governor General able to act in the way they did” (Conacher at paras. 7-9). This conclusion fully applies to the matter at hand. I would add that since Conacher, there has been no amendment to section 56.1 of the Act. In other words, Parliament’s legislative intent regarding section 56.1 has remained unchanged since adopting the provision in 2007, leaving the legal interpretation of the provision from Conacher unchanged as well.

[19] Second, the changes to the factual and legal landscape alleged to have occurred since Conacher was decided are of no assistance to the appellants. Even assuming that a new constitutional “confidence convention” has emerged since Conacher, which would limit the advisory authority of the Prime Minister in the manner suggested by the appellants, it is trite law that constitutional conventions are not enforceable by courts, although courts may be called upon to recognize their existence and determine whether they have been breached.

[20] As the Supreme Court of Canada stated in Re: Resolution to Amend the Constitution, 1981 CanLII 25 (SCC), [1981] 1 S.C.R. 753 (the Patriation Reference), “[t]he very nature of a convention, as political in inception and as depending on a consistent course of political recognition by those for whose benefit and to whose detriment (if any) the convention developed over a considerable period of time is inconsistent with its legal enforcement.” (Patriation Reference at 774-775; see also The Honourable Malcom Rowe & Nicolas Déplanche, “Canada’s Unwritten Constitutional Order: Conventions and Structural Analysis” (2020) 98:3 Can Bar Rev 430 at 444 (Canada’s Unwritten Constitutional Order)).

[21] However fundamental they may be to the Constitution, constitutional conventions are not part of the law of the Constitution as “[t]hey are not based on judicial precedents but on precedents established by the institutions of the government themselves”, “[n]or are they in the nature of statutory commands which is the function of the courts to obey and enforce.” (Patriation Reference at 880). I pause to recall that constitutional conventions are not to be conflated with the underlying constitutional principles discussed in Reference re Secession of Quebec, 1998 CanLII 793 (SCC), [1998] 2 S.C.R. 217, which can assist in the interpretation of constitutional provisions and may, in certain circumstances, give rise to substantive legal obligations (Reference re Secession of Quebec at para. 52; see also Canada’s Unwritten Constitutional Order at 440).

[22] Here, the appellants, for all intents and purposes, are not merely seeking recognition of the existence of a new constitutional convention in the shape of a “confidence convention”. They seek instead that the convention be legally enforced through a judicial declaration that section 56.1 of the Act must now be interpreted in a manner that curtails the Prime Minister’s advising authority in election matters to the extent provided for by this convention.

[23] In the Patriation Reference at pages 880-881, the Supreme Court opined that perhaps the main reason why constitutional conventions cannot be enforced by courts is that these conventions generally conflict with the legal rules that courts are otherwise bound to enforce, stating that such conflict “results from the fact that legal rules create wide powers, discretions and rights which conventions prescribe should be exercised only in a certain limited manner, if at all.” For example, as a matter of law, the Governor General could refuse to assent to every bill passed by Parliament, but constitutional convention prevents her from doing that. However, if this particular convention were violated, the courts would be bound to enforce the law, not the convention, and would therefore “refuse to recognize the validity of a vetoed bill” (Patriation Reference at 881).

[24] In sum, constitutional conventions cannot be crystallized into laws, “unless it be by statutory adoption” (Patriation Reference at 882).

[25] Therefore, assuming the appellants’ allegation that a “confidence convention” has emerged since Conacher and that it has been breached by the Prime Minister in the manner described in the Application, to be “true” for the purposes of the respondents’ motion to strike and assuming it is successfully established on the merits, that breach, to use the words of Justice Rowe and Me Déplanche, would only “create [sic] a deficit in legitimacy, not legality, which is sanctioned ultimately in the political arena” (Canada’s Unwritten Constitutional Order at 433).

[26] This, in my view, is what ultimately transpired in Conacher, where this Court, consistent with its function as a court, focussed on—and enforced—the law, that is, section 56.1 of the Act, and concluded that that provision “specifically preserved” the Governor General’s powers and discretion to dissolve Parliament and, by extension, the Prime Minister’s advice-giving role. The Court perfectly understood that in order to curtail these powers, Parliament would have had to use “explicit and specific wording”, something it had not done and, I would add, something it has not done since (Conacher at paras. 4-5). Again, as long as they have not been crystallized into law through statutory adoption, constitutional conventions are not legally enforceable and their breaches can raise only legitimacy concerns.

[27] The appellants point to the recent decision of the Court of Appeal of New Brunswick in Democracy Watch v. Premier of New Brunswick, 2022 NBCA 21 (NBCA Judgment), where that Court interpreted provisions of that province’s Legislative Assembly Act, R.S.N.B. 2014, c. 116, as foreclosing dissolution and election advice advanced purely for partisan electoral advantage. They contend that this judgment is part of the new legal landscape that must inform the decision this Court is called upon to make in the present matter.

[28] The problem with this submission is that there, Democracy Watch distinguished Conacher on the basis that, unlike section 56.1 of the Act, “[t]he legislation under consideration […] features wording that purports to explicitly define the dissolution and election advice the Premier is lawfully entitled to provide” (NBCA Judgment at para. 49). The Court of Appeal of New Brunswick agreed with that submission.

[29] Indeed, subsection 3(4) of New Brunswick’s Legislative Assembly Act makes it explicitly incumbent upon the Premier of that province to “provide advice to the Lieutenant-Governor that the Legislative Assembly be dissolved and a provincial election be held” in accordance with the schedule established therein. There is no such language in section 56.1 of the Act, which, as drafted “does not affect the Prime Minister’s ability to give advice to the Governor General” (Conacher at para. 7). In sum, the NBCA Judgment does not assist the appellants.

[30] Third, I see no reason to interfere with the Federal Court’s conclusion that it was bound to follow Conacher and not Miller, however persuasive this UK Supreme Court decision might be, assuming its relevance to the facts at issue in the case at bar. As the respondents point out, there are two problems with relying on that decision.

[31] On the one hand, it is distinguishable on its facts. In Miller, the UK Supreme Court was concerned with the prorogation of Parliament, which it held must be distinguished from dissolution (Miller at para. 4). Miller arose in circumstances related to the UK’s exit from the European Union, “which have never arisen before and are unlikely to arise again” (Miller at para. 1).

[32] The UK Supreme Court found that there were legal limits to the power to prorogue Parliament stemming from the constitutional principles of Parliamentary sovereignty—which, time and again, “courts have protected from threats posed to it by the use of prerogative powers” (Miller at para. 41)—and Parliamentary accountability. It held that the prorogation at issue was unlawful because it “prevented Parliament from carrying out its constitutional role” as a legislature and as the body responsible for the supervision of the executive at a time when a “fundamental change was due to take place in the Constitution of the United Kingdom” (Miller at paras. 50 and 56-57). It found that there was no justification for taking an action “[that] had such an extreme effect upon the fundamentals of [the country’s] democracy” (Miller at para. 58).

[33] Miller was found by the Court of Appeal of Alberta to be of no assistance in determining whether expected election periods set out in statute constrain the executive’s advisory powers, on the basis that the facts in Miller are distinguishable (Engel v. Prentice, 2020 ABCA 462 at para. 25).

[34] I too find Miller to be of no assistance in determining the issues on this appeal, as both the factual and legal underpinnings of that case are quite different from the ones in the present matter.

[35] On the other hand, what was binding on the Federal Court by the virtue of the stare decisis principle was this Court’s decision in Conacher since it is, as we have seen, directly on point (R. v. Comeau, 2018 SCC 15, [2018] 1 S.C.R. 342 at para. 26). Even if Miller were persuasive, it would not be binding on a Canadian court (Quebec (Attorney General) v. 9147-0732 Québec inc., 2020 SCC 32).

[36] Finally, the appellants submit that there is new evidence of the “will of Parliament”, which further supports their claim that the alleged “confidence convention” has been breached by the Prime Minister. This evidence consists of declarations made by political actors or reported in the media in the spring and summer of 2021, which opposed the calling of an election during COVID. The appellants contend that the Prime Minister, in deciding to advise the Governor General to dissolve Parliament and call an election despite maintaining the confidence of the other political parties, acted against the will of Parliament.

[37] I agree with the respondents that this “new evidence” is incapable of changing the outcome of this case, as Conacher definitively concluded that such evidence is irrelevant to the interpretation of section 56.1 of the Act. The “will of Parliament” has been formally expressed through section 56.1 of the Act. This is the “statutory command [sic] which it is the function of the courts to obey and enforce” (Patriation Reference at 880). Section 56.1 was found in Conacher not to affect the Prime Minister’s advice-giving role regarding the dissolution of Parliament and the calling of elections.

[38] It is of course always open to Parliament to amend that “statutory command” but, as mentioned previously, it has not done so. As indicated previously as well, even if a convention pointing in a different direction had emerged since the adoption of section 56.1, this convention would not be legally enforceable and could have no bearing on the interpretation of that provision.

[39] Although as pointed out by the Federal Court, quoting from R. v. Imperial Tobacco Canada Ltd, 2011 SCC 42, [2011] 3 S.C.R. 45 at paragraph 21, motions to strike must be used with care, I am satisfied that the Application has no reasonable prospect of success and must not be allowed to proceed. Therefore, I see no reason to interfere with the Federal Court’s decision.

[40] Having so concluded, I see no need to rule on the mootness question raised by the respondents, which ultimately comes down to determining whether the Federal Court should have declined, at the second step of the mootness analysis (Borowski v. Canada (Attorney General), 1989 CanLII 123 (SCC), [1989] 1 S.C.R. 342 at 353), to entertain the Application because it amounts to an attempt to relitigate Conacher.

[41] For all these reasons, I would dismiss the appeal. The respondents are seeking their costs on appeal. The appellants do not seek their costs, and ask that no costs be awarded against them. However, in the particular circumstances of this case, I would award costs, as was the case in the Federal Court, to the successful party. The respondents proposed an amount of $2,000.00, which the appellants found to be reasonable. I would therefore award costs to the respondents in that amount.

"René LeBlanc"

"I agree.

Donald J. Rennie J.A."

"I agree.

J.B. Laskin J.A."

FEDERAL COURT OF APPEAL

NAMES OF COUNSEL AND SOLICITORS OF RECORD

APPEARANCES:

SOLICITORS OF RECORD: