In the decision it released yesterday in Mikisew Cree First Nation v Canada, the Supreme Court of Canada dealt with the question of whether an Aboriginal people’s right to be consulted applies to the « legislative process ». Such a right to consultation has been judicially added to « historical » treaty no. 8, to which the Mikisew Cree First Nation is a party, in a previous case, where the Court took its inspiration from its jurisprudence pertaining to a free standing right to be consulted prior to the taking of a measure that might impact a right that is still « just » claimed (as opposed to a requirement for actual infringement of an established aboriginal or treaty right). Yesterday’s judgement thus concerns both types of the right to be consulted that the Court has read into the protection of aboriginal and treaty rights of s. 35 of the Constitution Act 1982. By a 7-2 majority, the Court’s answer to the question of whether this right applies to the legislative process is « no. » The case’s unanimous ratio decidendi, however, is that the Federal Court did not have the jurisdiction to consider the application for judicial review. Still, what I want to talk about here is a short obiter dictum that is found in three of the four sets of reasons, subscribed to by eight of the nine judges.
At para. 51 of her reasons, Karakatsanis J. provides the following justification for the broader dictum that is her negative answer to the « duty to consult » (it’d make more sense to talk of a right) question by reassuring Aboriginal Peoples regarding the other remedies or solutions they may benefit from: « […] this conclusion does not affect the enforceability of treaty provisions, implemented through legislation, that explicitly require pre-legislative consultation (see e.g. Nisga’a Final Agreement, Chapter 11, paras. 30-31; Nisga’a Final Agreement Act, S.B.C. 1999, c. 2; Nisga’a Final Agreement Act, S.C. 2000, c. 7). Manner and form requirements (i.e. procedural restraints on enactments) imposed by legislation are binding (Hogg, at s. 12.3(b); see also R. v. Mercure,  1 S.C.R. 234). » Writing also for Martin J., Abella J. writes similarly, at para. 96, that « [f]ailure to comply with a manner and form requirement will result in the legislation being invalid, as there is “no doubt as to the binding character of the rules in the Constitution” (Hogg, at pp. 12-11, 12-18 and 12-19) ». Again, Rowe J., writing also for Moldaver and Côté JJ., at para. 167, contends that « [i]f Parliament or a provincial legislature wishes to bind itself to a manner and form requirement incorporating the duty to consult Indigenous peoples before the passing of legislation, it is free to do so (Hogg, at p. 12-12) ».
They thus all assert that legislative provisions enacted, by the same legislature, contrary to other (preceding) ones that required consultation, would be invalid by virtue, not of s. 35 of the Constitution Act 1982, but of self-imposed, through formally ordinary, legislation, of manner and form requirements. I respectfully submit this is wrong, for reasons that follow, excerpted from a chapter manuscript by Patrick Baud and me.
As a result of the Statute of Westminster 1931 and the Canada Act 1982, the Colonial Laws Validity Act 1865 no longer applies to Canada. Unlike in Australia, however, there was no effort to reproduce the legal basis for “manner and form” requirements as part of Canada’s supreme law. Indeed, before its repeal in 1982, subsection 92(1) of the Constitution Act 1867 conferred jurisdiction to provincial legislatures respecting “[t]he Amendment from Time to Time, notwithstanding anything in this Act, of the Constitution of the Province, except as regards the Office of Lieutenant Governor.” Likewise, section 45 of the CA 1982 now provides that, “[s]ubject to section 41, the legislature of each province may exclusively make laws amending the constitution of the province.” There is no mention in either provision that a particular manner or form of legislation need be followed to make such changes nor, in contrast to Australia, is there a mechanism for the recognition of such “manner and form” requirements even if a provincial legislature purported to establish them.
In the 1991 Canada Assistance Plan reference, the Supreme Court, obiter dictum, said (at 563) “[i]t is clear that parliamentary sovereignty prevents a legislative body from binding itself as to the substance of its future legislation.” Sopinka J, writing for a unanimous Court, also suggested that a requirement that a person or group outside the legislature consent to the enactment of legislation was a requirement as to substance, rather than “manner and form” (at 563–564).
But the Court in the Canada Assistance Plan Reference did not entirely rule out “manner and form” requirements. As Sopinka J explains (at 563), “[i]t may be that where a statute is of a constitutional nature and governs legislation generally, rather than dealing with a specific statute, it can impose requirements as to manner and form.” Such a statute would also need to demonstrate parliament or the legislature’s “intent[ion]…to bind itself or restrict the legislative powers of those of its members who are also members of the executive” (at 562).
Sopinka J cited R v Drybones and R v Mercure as examples of effective “manner and form” requirements. In R v Drybones, the Supreme Court concluded that a provision of the Indian Act was inoperative under the Canadian Bill of Rights. But to say that a statute is inoperative is not the same thing as saying that it is invalid. The Canadian Bill of Rights, like other quasi-constitutional legislation, takes primacy over other legislation such that where another statute is inconsistent with a right set out in the Bill of Rights, the right takes precedence and the statute does not operate. Compliance with the Canadian Bill of Rights is not, however, a “manner and form” requirement that goes to the validity of a statute (see Maxime St-Hilaire).
R v Mercure may appear to pose more difficulty for the position we take here. Section 110 of the North-West Territories Act required that the territory’s legislature enact its ordinances in English and French. From the territorial legislature’s perspective, the North-West Territories Act was “supreme law” because Parliament had plenary power over the territory (Constitution Act, 1871, s 4). As a result, if the territorial legislature failed to comply with section 110, its ordinance would be beyond its powers and consequently, invalid.
When the federal parliament –doing so under the authority of the Constitution Act, 1871, s 2, which provides what s 6 of that Act makes clear is a one-time authority to establish a new province and provide for its constitution– created Saskatchewan from part of the North-West Territories by enacting the Saskatchewan Act, it continued all laws existing in Saskatchewan, including section 110 of the North-West Territories Act. And expressly (s 14) continued “all provisions of the law with regard to the constitution of the Legislative Assembly of the North-West Territories”, a category which, according to La Forest J, includes section 110 (at 262, 314–315, Estey J, dissenting, disagrees). Unlike the territorial legislature which preceded it, the Saskatchewan legislature could repeal any laws that fell within its jurisdiction even if, as was the case with section 110, they had originally been enacted by the federal parliament (at 271-272). La Forest J, writing for a majority of the Supreme Court, took the view that the legislature could repeal section 110, but that it would have to comply with section 110’s “manner and form” requirement to do so (at 277, 280–281).
The reason, however, that section 110 was an effective “manner and form” requirement in the North-West Territories, namely the hierarchical relationship between the federal parliament and the territorial legislature, disappeared when the federal parliament created Saskatchewan. Once Saskatchewan came into being, its legislature was in no way subordinate to parliament.
The only limits on the Saskatchewan legislature’s authority were those that flowed from “supreme law”, including the division of powers between the federal parliament and the legislature, and the provisions of the Saskatchewan Act that would require a constitutional amendment to alter or repeal, such as the guarantee of denominational education rights. (Today, this list would also, of course, include the guarantee of rights in the Charter and the recognition and affirmation of aboriginal and treaty rights in section 35 of CA 1982.) Although section 110 became part of Saskatchewan law in 1905, it took on a less significant weight than it had in the North-West Territories because of the differences between the Saskatchewan and territorial legislatures.
The Supreme Court reaffirmed the impossibility of a legislator’s binding itself as to substance in the 2011 Securities Act reference, but still has not, at least with any degree of clarity, decided the point as to “manner and form”. Again, in the absence of a legal basis to that effect in Canadian law, our view is that “manner and form” requirements cannot be understood as conditions precedent to a statute’s validity, that is, as prerequisites to the existence of a legal norm.
The relatively few authorities from the nearly three decades since the Supreme Court delivered its opinion in the Canada Assistance Plan reference are divided on the possibility of effective manner and form requirements. (Canadian Taxpayers Federation v Ontario (Minister of Finance) (2004), 73 OR (3d) 621 (SC); Greater Vancouver Regional Distrct v British Columbia (Attorney General), 2011 BCCA 345; Canada (Attorney General) v Friends of the Canadian Wheat Board, 2012 FCA 183; Oberg v Canada (Attorney General), 2012 MBQB 64; Amalgamated Transit Union Local 1374 v Saskatchewan (Finance), 2017 SKQB 152.) In 2014, however, the Court of Queen’s Bench of Manitoba, despite referring an attempt by the Manitoba legislature to bind itself as to the substance of legislation, refused to give effect to what was actually a manner and form requirement. There was no appeal.
The federal parliament and provincial legislatures can probably provide for the mandatory holding of a referendum by the executive, even on a popular initiative, as long as this referendum is not a formally legislative one, that is, one whose result is the making of legislation without further action (In re Initiative and Referendum Act,  AC 935 (UKPC); The King v Nat Bell Liquors Ltd.,  2 AC 128 (UKPC)). The “ratio decidendi” (it was a reference) of the Judicial Committee of the Privy Council’s opinion in the Initiative and Referendum Act reference was that the providing of a legislative referendum by provincial ordinary legislation amounted to an unconstitutional amendment relating to the “office of Lieutenant Governor” pursuant to then in force subsection 92(1) of the CA 1867 because it would deprive the Lieutenant Governor of the power to assent to the legislation or to reserve it under sections 55 and 90 of that Act. The “office of the Queen…and the lieutenant governor of a province” is a matter now dealt with at paragraph 41(a) of the CA 1982, and still falls outside the competencies of provincial legislatures.
But in the Initiative and Referendum Act,  AC 935 (UKPC), reference, where the Manitoba statute also gave the electorate a form of legislative initiative, the Privy Council suggested (at 945) in obiter dictum that a provincial legislature may not “create and endow with its own capacity a new legislative power […].” This line of reasoning that was later applied (at 73) to the federal parliament by the Supreme Court in the first Senate reference in 1979, which answered questions now dealt with at Part V of the CA 1982 and revisited in the second Senate reference in 2014. Generalizing, this would mean that neither parliament nor a provincial legislature could alter the core features of the legislative process, beginning with the formulation and introduction of a bill, including its adoption by a legislative house, and culminating in royal assent. It might also imply that a “Henry VIII” clause, through which parliament or a legislature empowers the executive to amend primary legislation, is unconstitutional.
In this regard, let us recall that in the Canada Assistance Plan reference, the Supreme Court may have recognized a constitutional protection for the government’s legislative initiative by saying (at 560) that “[a] restraint on the Executive in the introduction of legislation is a fetter on the sovereignty of Parliament itself.” In its opinion on the proposed pan-Canadian securities regulation scheme, the Quebec Court of Appeal gave effect to this view, concluding that an agreement not to make amendments to securities legislation unless approved by a Council of Ministers was an impermissible restriction on the executive’s legislative initiative and consequently, on the sovereignty of provincial legislatures (Renvoi relatif à la réglementation pancanadienne des valeurs mobilières, 2017 QCCA 756, para 69–70, appeal as of right to SCC).
Should this principle hold be true, we cannot see how ordinary legislation could prevent a minister from ignoring a requirement that the government hold a purportedly “mandatory” referendum and instead introducing incompatible legislation. This principle would mean that neither parliament nor a provincial legislature could make the result of a referendum is binding on the executive (See e.g. Referendum Act, SNB 2011, c 23, ss 12-13; The Referendum and Plebiscite Act, SS 1990-91, c 8.01, ss 3-4). Moreover, there is a decision of the Manitoba Court of Queen’s Bench that suggest that ordinary legislation might not be able to provide for a mandatory referendum, even a consultative one, as part of a legislative process (Progressive Conservative Party of Manitoba v Government of Manitoba, 2014 MBQB 155, para 36).
The limits on changes to the legislative process may not be quite as restrictive as these dicta would have us believe. After all, all of the Canadian provincial legislatures that used to include an (appointed) legislative council abolished those councils through ordinary legislation enacted under former subsection 92(1) of the CA 1867. Nowadays at least, nobody seems to think that the abolition of the legislative councils might have been unconstitutional (Montplaisir v Quebec (AG),  RJQ 109 (CS); R v Somers,  MJ No 57 (QB)).
This is an area in need of further scholarship and clearer, more consistent legal doctrine. But, in the meantime, yesterday’s rather lightly-made dicta probably just add to the confusion.
5 réflexions sur « Mikisew Cree First Nation v Canada and « manner and form » requirements: tiny dicta with huge implications »
This result is particularly surprising given that the s. 2d) jurisprudence requires pre-legislative consultations with trade unions in some cases, on pain of the declaration of invalidity of the impugned legislation.
See : Health Services and Support – Facilities Subsector Bargaining Assn. c. Colombie-Britannique,  2 RCS 391 and the more detailed analysis in British Columbia Teachers’ Federation v. British Columbia, 2015 BCCA 18 (appeal to the SCC denied).
Merci de cette excellente observation.
Si la source est la loi suprême, ne peut-on pas justement plus facilement admettre qu’il puisse y avoir invalidation de la loi qui est assujettie? Tu pourras m’objecter qu’ici, dans l’arrêt Mikisew Cree FN, il est question de la mise en oeuvre législative de traités, donc de droits issus de traité et donc de l’article 35 LC 1982. Ce à quoi je te répondrais deux choses. Premièrement, les dicta ici en cause, notamment à la lumière des références indiquées, sont de portée beaucoup plus générale. Secondement, la jurisprudence de la CSC semble « détacher » formellement les traités et droits issus de traités de leur mise en oeuvre législative, par exemple en se refusant à sanctionner des défauts de mise en oeuvre par l’État (voir l’arrêt Beckman).
Selon moi, dans un État constitutionnel, pourvu d’une loi constitutionnelle formelle, supra-législative, il est capital de ne jamais amalgamer loi (formellement) ordinaire et loi (formellement) constitutionnelle, pas davantage que d’amalgamer compétence formellement constitutionnelle (la véritable procédure de modification de la constitution) et compétence législative, y compris celle qui n’est que matériellement « constitutionnelle » (par exemple celle prévue aux articles 44 et 45 LC 1982). J’explique ça dans cette série de billets: http://www.iconnectblog.com/2017/09/150-years-on-what-is-the-constitution-of-canada-part-1-of-3-the-problem-of-identification/
Il faudrait donc, à mon sens, insister sur la différence entre l’inopérabilité d’une loi qui est mise à l’écart par suite de résolution judiciaire d’un conflit entre deux lois de même niveau formel, d’une part, et, de l’autre, l’invalidité d’une loi pour cause de violation de la loi formellement constitutionnelle.
C’est ce qui est actuellement en jeu dans l’amalgame, stratégiquement cultivé, au sujet de la dérogation aux droits fondamentaux, entre les chartes constitutionnelle (canadienne) et quasi constitutionnelle québécoise, ainsi que sur la question de la « compétence constitutionnelle » du Québec, dans l’affaire Henderson et dans certains projets de constitution québécoise.
À un autre niveau, c’est l’État de droit constitutionnel comme bien social qui est ici en jeu, à mon humble avis, et comme Kelsen le pensait, je pense à mon tour que cela exige de maintenir une certaine rigueur conceptuelle afin d’éviter certains dérapages.
But if by « this result » you mean, not the topic of this post, but the court’s answer to the duty to consult question, then I’d just remind you that the court did not deny the fact that lack of consultation may lead to a conclusion of unconstitutionality of a statute. It’s important to distinguish consultation in the meaning of Haïda Nation (which was at stake in this case) and consultation under the Sparrow test, as I say in the first lines of this post. Now, on every judge’s acknowledging of the relevance of consultation when examining whether an aboriginal or treaty right infringement, including by legislative provisions, is justified, see paras. 43 (Karakatsanis), 64-6 (Abella), 145 (Brown), 154 (Rowe).