Editor's Note: The series this article is part of was originally submitted as one academic paper. Due to its length we divided it up into several parts which will be posted in installments. Additionally we have not included the original footnotes . We will include the whole original piece with footnotes as an attached document to the last installment of the series for readers who wish to read it in its original form. Read Part 1 here, Part 2 here, and Part 3 here.

IV. The First Attempt at Canadian Constitutionalism (The B.N.A. Act)

Canada, broadly speaking, was founded by ultramontanist French Catholics and English-speaking heirs to high-church Anglican Loyalists who had made treaties with the First Nations people. Even though not all the Fathers of Confederation shared the faintly theocratic views of their Loyalist predecessors, who looked disdainfully on the perceived destructive individualism and insubordination of the American Revolutionaries, a “Tory touch” has been recognized in the Canadian founding. There was generally a desire for a more communitarian society where the Church had a more prominent social role, as had existed in Upper and Lower Canada (witness the clergy reserves). Thus, the British North America Act 1867 (henceforth the B.N.A. Act) protected group rights and religious rights, such as the rights of publicly-funded religious (“separate”) schools in section 93.

Sadly, despite its attempted traditionalism, we must admit that the constitution of the B.N.A. Act, far from restoring checks and balances, instead contained the absolutizing dangers of both the British and the American constitutions.

On the one hand, its adaptation of the Westminster parliamentary system brought with it the tendency towards executive sovereignty, and, in the Canadian context, the sovereignty of the federal executive specifically. The complaints about how tyrannically Trudeau oversaw the genesis of the Charter inadvertently reveal this: With his majority government, Canada’s Parliamentary system permitted him and his cabinet to behave autocratically without any checks to slow them down.

The lieutenant governor of a province, representing the reigning monarch, has, according to section 90 of the B.N.A. Act, reserve powers to either refuse royal assent or to reserve it. The right of outright refusal seems like a restoration of the medieval royal veto, and was used 38 times between 1870 and 1945. However, this was normally done on the advice of the provincial cabinet (executive sovereignty again). Its final use was in 1945, when Lt. Gov. Bradford LePage of Prince Edward Island refused assent to a modification to the province’s prohibition legislation based on his moral convictions about alcohol. His term having expired anyhow, he was promptly replaced. However, the Supreme Court of P.E.I. did subsequently find that the amended statute was not binding law since it never did receive royal assent. This was the last time royal assent was outright refused.

The lieutenant governor also has the prerogative to reserve legislation and refer it up to the governor general-in-council, who can then disallow that act under instructions from the federal cabinet. Much like the provincial veto, this was not uncommon (federal disallowance occurred 112 times between 1867 and 1943) until a lieutenant-governor acted on their own initiative. In 1961, Saskatchewan Lt. Gov. Frank Bastedo deemed the CCF’s Act to Provide for the Alteration of Certain Mineral Contracts not to be in the public interest and reserved assent from it without the advice of the federal government (which promptly issued an order-in-council granting royal assent). This was so controversial that there was talk of abolishing the office of lieutenant governor after this. As with refusal, reservation stopped being used as soon as a lieutenant governor exercised their own discretion in using it rather than taking orders from the cabinet, although Alberta NDP leader Rachel Notley did make an unsuccessful appeal to Lt. Gov. Lois Mitchell to refuse assent to Bill 22 in 2019.

Nor has the viceroy been an effective curb on corrupt or ineffective governments. Although in 1878 Quebec’s Lt. Gov. Luc Letellier de St-Just dismissed the government simply because he deemed them incompetent, despite them holding a majority in the legislative assembly, since the King-Byng affair of 1926, the monarch’s representative has been subordinate to the executive in matters relating to the status of the legislature.

Further, as in the U.K., Canada’s model of “Parliamentary sovereignty” was equally vulnerable to the risk of judicial supremacy. Decades before the U.K.’s Supreme Court suggested it had the authority to recognize “constitutional statutes”, Canada’s Supreme Court acknowledged a class of legislation as being “quasi-constitutional”, which were not subject to the doctrine of implied repeal and could only be repealed by explicit language setting out so to do. An example of this was the Canadian Bill of Rights 1960. But there was an even bigger opening for judicial supremacy to flow in through.

In his previously mentioned work on the British constitution, A.V. Dicey remarked that the preamble to the B.N.A. Act was inaccurate when it said it was establishing “a Constitution similar in Principle to that of the United Kingdom”. Dicey wanly suggested that the Act should have said “States” instead of “Kingdom”. Although the Fathers of Confederation may have expected that granting the federal Parliament so much power would avoid the necessity of going to the courts to interpret the Constitution, the “futility” of this “hope...is proved by the existence of two thick volumes of reports filled with cases on the constitutionality of legislative enactments...In Canada, as in the United States, the Courts inevitably become the interpreters of the Constitution.” The phenomenon of Supreme Court justices (who, like their American counterparts, cannot be removed so long as they observe “good behaviour”) is not a phenomenon that began with the Charter.

Consider the Supreme Court of Canada case Reference Re Alberta Statutes [1938]. The Alberta Social Credit government had passed the Accurate News and Information Act, which required newspapers to print government rebuttals to critical articles and to publicly name their sources, moves which were widely seen as restricting the freedom of the press. The B.N.A. Act does not contain any explicit protection of the freedom of expression. However, the Court struck down the legislation. Its reasoning, as expressed by Duff CJ, was that the preamble of the B.N.A. Act envisions a constitution “similar in principles to the United Kingdom.” In the U.K., Parliament “work[s] under the influence of public opinion and public discussion”, which would be impaired by the legislation. Therefore, the statute was unconstitutional.

Some might see this as a rather imaginative construction of the preamble. Nevertheless, this is an example of courts ruling based on the so-called “implied bill of rights” which they had discovered in the Canadian constitution. A standard this fluid is at least as conducive to judicial activism as vague “Charter values” are. Whatever advantages Canadian society and jurisprudence may have had in the past, they were no less susceptible to judicial overreach and even judicial sovereignty.

Synthesis: The Second Attempt at Canadian Constitutionalism (The Charter)

Pierre Trudeau’s original vision of patriation, and the role of the Charter of Rights and Freedoms within it, would have been a full “Americanizing” the country. That this spirit was latent can be seen in how Trudeau had prepared a plan to declare independence from the U.K. if his patriation request was rejected. Unlike the American federation, however, this new constitution was not going to arise from the consent of local states, but would be imposed from above by the federal government.

When Trudeau submitted his patriation request to the U.K. Parliament, he did so without the support of most of the provinces. The concerned British government asked the Select Committee on Foreign Affairs, chaired by Sir Anthony Kershaw, to review the constitutionality of such a unilateral request. Natural law philosopher John Finnis was appointed “special adviser” to the committee.

The Kershaw Reports determined that the U.K. Parliament should only accept a federally spearheaded constitutional change if it “accords with the wishes of the Canadian people as a federally structured community”, that is to say, if it has “at least that degree of Provincial concurrence (expressed by governments, legislatures or referendum majorities) which would be required for a post-patriation amendment affecting the federal structure in a similar way.” Finnis would later note that this was simply a restatement of what Aquinas and Fortescue said about how the “whole kingdom” should be involved in its own governance. This is important, because it lends credence to the idea that we should use a natural law hermeneutic to interpret the constitution which came about from this process.

When this finding was supported by a similar ruling by the Supreme Court of Canada asserting a conventional convention that there should be a substantial amount of provincial consent for a constitutional change, Trudeau was forced by political necessity to negotiate with the provinces in order to get the Charter passed. In November 1981, a conference with the premiers was held in Ottawa. In the course of the discussions, Alberta Premier Peter Lougheed suggested that the Charter include a provision from the Canadian Bill of Rights 1960: a “notwithstanding clause” which would allow the legislature to pass laws “notwithstanding” contrary provisions in the Charter (or, at least, “contrary” as interpreted by the courts). This found its way into a three page proposal submitted by the Newfoundland delegation. In the evening of November 4, Jean Chrétien held the legendary (and perhaps mythologized) “Kitchen Accord” where he agreed to the inclusion of the notwithstanding clause. The provinces agreed to support the Charter, which now contained the fateful section 33.

Opponents of the clause, like the perennially perturbed Andrew Coyne, contemptuously describe this a grubby compromise rather than a high-minded expression of philosophical principles. But Maistre should have taught us that we should not put our trust in aloof political thinkers spinning constitutional systems from their armchairs. In fact, out of this political sausage-making, and seemingly by accident, a fascinating and unique constitutional arrangement has developed. Consider the checks and balances that exist in the Charter’s system.

On the one hand, we have a Westminster-style Parliament; the legislature (though often in practice the executive) initially seems sovereign. On the other hand, the courts have a Bonham-esque ability to strike down legislation they deem immoral by using familiar tricks of judicial creativity in interpreting Charter rights. To that extent, the judiciary seems sovereign, and it is with regard to the interpretation of certain rights (democratic, mobility, language, and education rights, all of which are a bar against the tyranny of the majority). However, in other areas, which relate more obviously to moral judgment (those found in sections 2 and 7-15), the government can in turn veto the judicial veto by invoking section 33. We seem to be back at Parliamentary sovereignty, except that this veto exists on a razor’s edge. Unlike most legislation, which operates until it is repealed by later legislation, the contested statute is only protected for five years, the length of a legislative session (section 4), after which point the clause must be invoked again. If the government has lost the confidence of the legislature (and the electorate) by then, and the new government does not re-apply the clause, the Court’s veto snaps back into operation.

The judiciary has a kind of sovereignty in certain limited areas, and, if Parliamentary sovereignty exists in other realms, it is even more precarious than the king’s sovereignty in the Middle Ages, lasting only as long as it commands the democratic support of its people and not a moment longer. This is perfectly consistent with section 1, which “guarantees the rights and freedoms set out [herein] subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society” (emphasis added).

On paper, the Charter has blundered into a system of checks and balances that is not only more effective than America’s constitution, but is perhaps even more balanced than the medieval constitution. No one has unchallenged sovereignty here; circumstances beyond anyone’s control (but ultimately determined by the electorate) effectively determine who will have final say in a legislative conflict. Seeds from both the British and American constitutions, buried in the soil of Canadian history, here blossom into something new. We have, in both a descriptive and prescriptive constitutional sense, the supremacy of God. Maistre would smile with vindicated satisfaction.


The system we have identified from the text of the Charter is obviously not what we observe in Canadian society today. There is a widespread stigma against invoking the notwithstanding clause, based largely on the misapprehension that section 33 gives the government a pass from having to follow the rest of the constitution.

Peter Hogg’s “dialogue” theory defends this status quo by making the argument that, when legislation is struck down for violating the Charter, legislatures usually go on to pass similar laws that achieve the same effect but without the offending provisions. The courts do not thwart the legislature’s intention, Hogg argues; they just ensure that it is implemented in a constitutional way. But calling this a “dialogue” is an odd use of language. The Vatican can officially censure a theologian; that theologican can then adjust his teaching so that it matches official doctrine, even though it is substantially the same as what he taught before. But it would be odd to call this a “dialogue”. One party is obviously giving the other party binding instructions, not chatting with them. Moreover, in a dialogue, each party is theoretically able to learn something from the other, but, in Hogg’s model, the courts are the Magisterium, not a conversation partner. Yet the legislature can sometimes be a better constitutional interpreter than the courts.

Take the recent ruling by the Queen’s Bench of Saskatchewan that non-Catholic students do not have a right to government funding to attend separate schools. The ruling clearly evinces a belief that the constitutional protection of Catholic schools is outdated, and it therefore concocts a way to kneecap and possibly end Catholic schools as a way of getting around the constitution. By invoking the notwithstanding clause to nullify this ruling, the government of Saskatchewan is being more consistent with the constitution than the court. Here, the notwithstanding clause upholds the spirit of the constitution rather than undermining it.

This essay has argued that statutes and rulings should be judged against the standard of natural law. Natural law, in turn, comes to us interpreted and explained by different religious groups, in various wisdom traditions, and through many philosophical schools. Our need to learn from these communities--whether they call the natural law the Logos, the Dharma, or the Tao--is the best justification for Canada’s policy of multiculturalism. But the expansion of law in the modern state is a perennial threat to their autonomy, as it has been since the medieval Church clashed with absolutist kings. The courts sometimes protect their rights from the government, but the notwithstanding clause allows the government to protect them from the courts.

Obviously, not every use (or threatened use) of the clause has been of this nature. Quebec has used it to suppress the freedom of provincial employees to wear religious symbols. But, appalling as this is, it only directly affects Quebec, and will only last as long as its citizens elect governments who pursue this policy. In contrast, when the Privy Council ruled in Barrett v City of Winnipeg [1891] against the Catholic minority’s right to separate schools in Manitoba, despite the clear constitutional protection of that right, its ruling set a harmful precedent that had a ripple effect in provinces like Ontario and Quebec for decades afterwards, hurting both English Protestant and French Catholic communities and stirring up conflict that could have been avoided.

(Incidentally, it should not be forgotten that the Supreme Court of Canada had already unanimously ruled in favour of the Catholic minority before the Privy Council overturned their decision. Nostalgia for a time when the Council kept the Supreme Court in check should evaporate in the face of cases like Barrett or Cunningham v Homma [1902]. Under the Charter, the legislature can thwart deleterious rulings like this in a way they never could before 1982.)

A legislative misuse of the notwithstanding clause for an unjust law is confined geographically and temporally in a way that an unjust court ruling is not. If Quebec using the clause in an oppressive way gives section 33 an unpleasant association, it is only because other provinces have been unwilling to invoke it under more salutary circumstances and thus develop conventions around the clause’s appropriate use. Saskatchewan has finally got the ball rolling on this. We should keep it rolling until it snowballs.

In closing, then, this paper suggests that legal scholars and legislators adapt a more liberal approach towards invoking section 33, particularly in the protection of religious minorities with practices rooted in unpopular beliefs. This would better instantiate the constitution outlined in the Charter, which not only delineates a balance of powers, but also a system where minorities are protected. A situation where elected representatives and judges vie with each other on how to achieve that is an eminently desirable one.

Full Version with Footnotes