Philosopher James Tully once argued that the essential objective of the Supreme Court in the Reference case on the secession of Quebec (1998) was to show that the 1982 constitutional order was not a straitjacket. And yet, in 1990, we were forced to acknowledge the immense difficulty of amending the Canadian constitution. The five minimum conditions demanded by Robert Bourassa in the 1987 Meech Lake Accord were refused, since they had to be ratified by all the provincial legislatures before they could be included in the Canadian constitution. Two provinces, Manitoba and Newfoundland, failed to approve this constitutional change within the three-year period prescribed by the amending formula found in Section V of the Constitution Order 1982.

Those who argue that Canada’s constitutional order of 1982 is a straitjacket are relying on the amending formula.

Admittedly, Section V stipulates that the Canadian constitution can be amended in various ways: a province can unilaterally amend the constitution in the section that exclusively concerns it (§45). The constitution can sometimes be amended bilaterally by agreement between the federal government and the province (§43). It can also sometimes be amended according to the « 7/50 » formula, requiring the support of seven provinces representing at least 50% of the Canadian population (§38 and §42). Finally, in certain cases, it may require the unanimous support of all provincial legislatures (§41). If unanimity of the provinces was required to include the five Meech conditions in the constitution, it’s easy to imagine that Quebec’s secession, requiring a substantial amendment to the Canadian constitution, would also easily require such unanimity.

Yet, in an article published in Le Devoir on August 19, 2023 (« Le renvoi sur la sécession du Québec 25 ans plus tard »), André Binette and André Joli-cœur write that « the Supreme Court refused to apply to secession the amending formula that had been imposed on Quebec by the odious Constitution of 1982. »

This statement is astonishing. The authors clearly imply that the Court rejected the idea that the amending formula would apply to a process such as Quebec’s secession. The Supreme Court would then have removed the straitjacket? When we read section 105 of the Reference, however, we hear a completely different story:

 » §105 It will be noted that Question 1 does not ask how secession  could be achieved in a constitutional manner, but addresses one form of secession only, namely unilateral secession .  Although the applicability of various procedures to achieve lawful secession was raised in argument, each option would require us to assume the existence of facts that at this stage are unknown.  In accordance with the usual rule of prudence in constitutional cases, we refrain from pronouncing on the applicability of any particular constitutional procedure to effect secession unless and until sufficiently clear facts exist to squarely raise an issue for judicial determination. . »

So the judges don’t refuse anything, they abstain! Section 105 is not a refusal to use the 1982 amending formula. It expresses a refusal to rule on the how and on the procedure, the question not being for the moment justiciable. For the time being, the Court leaves this question open. In 1998, only three years after the 1995 referendum, the Court had to avoid another constitutional coup. It could not impose an obligation to comply with a constitutional amendment procedure that had been imposed on Quebec. Prudence was the order of the day. It could be otherwise in another context.

The Court had to rule only on unilateral secession, which for the Court meant secession without negotiation, and therefore without submitting to the Constitution and its underlying principles. It replied that, under these principles, a unilateral secession would be illegal. To achieve secession in accordance with the underlying principles, sovereignty would require not only a referendum with a clear positive result on a clear question (under the principles of democracy and federalism), but also a negotiated amendment to the Canadian constitution (under the rule of law and constitutional order). In other words, the Canadian constitutional order would have to be amended to take into account the new situation resulting from Quebec sovereignty, in order to ensure a smooth transition. Such an amendment cannot be imposed. It must be negotiated. This is why unilateral secession (in the sense of secession without negotiation) would not be consistent with Canadian domestic law.

The Court simply stated that, to be constitutional, secession would require a constitutional amendment approved by two legitimate majorities, the Quebec majority and the Canadian majority:

« §152   The negotiation process would require the reconciliation of various rights and obligations by negotiation between two legitimate majorities, namely, the majority of the population of Quebec, and that of Canada as a whole. »

The Court also clarified that a victorious referendum on secession does not oblige the majority to accept secession and negotiate nothing more than the terms of secession:

« §90   The conduct of the parties in such negotiations would be governed by the same constitutional principles which give rise to the duty to negotiate: federalism, democracy, constitutionalism and the rule of law, and the protection of minorities.  Those principles lead us to reject two absolutist propositions.  One of those propositions is that there would be a legal obligation on the other provinces and federal government to accede to the secession of a province, subject only to negotiation of the logistical details of secession.  This proposition is attributed either to the supposed implications of the democratic principle of the Constitution, or to the international law principle of self-determination of peoples. »

Of course, it would be equally unacceptable to claim that a clear majority in favor of secession would not entail any obligation to negotiate on the part of the Canadian majority (§92). But the latter would perfectly respect the obligation to negotiate if it made a proposal for reform of the Canadian constitutional order that did not entail secession.

Let’s imagine, then, that the Canadian majority were to do just that. It would respond to Quebec’s attempt to secede by proposing to negotiate the reform of Canadian federalism. Suppose they propose a specific reform inspired by some of Quebec’s traditional demands. The agreement could also include the following idea. Should it not be possible to agree on the content of the proposed reform, the Canadian majority could accept secession on condition that it be carried out, for example, in accordance with the application of the 7/50 formula set out in Section V.

This would in no way contradict the Reference. In this precise context, the Canadian majority would have shown a real willingness to negotiate. The Court would be faced with a justiciable situation. It could be called upon to rule on the constitutionality of this proposal. It could consider the proposal made by the Canadian majority to be reasonable. The framework of the amending formula could then be invoked favorably in its judgment. In such a context, Binette and Joli-Coeur might try to show that the Court had « refused » to use the amending formula in 1998, but the Court would have them understand that it had only refused to rule, as the question was not yet justiciable at the time.

Rather than paint a picture that may turn out not to be true, it’s better to look reality in the face, if we want to avoid a future that brings us disappointment.