Skip to content.

Can the Scots secede? Lessons learned from Canada

On November 23, 2022, the United Kingdom Supreme Court ruled that Scotland’s Parliament may not hold a second independence referendum at this time. The last independence referendum was held in 2014, with the legal approval of the UK. Scotland narrowly voted to remain within the UK (55%).


Scotland’s constitutional relationship to the United Kingdom is governed by the Scotland Act 1998, c. 46 (UK). This UK Act establishes the Parliament for Scotland as a devolved legislature within the British legal system. The Act lays out the legislative competencies of the Scottish Parliament, reserving certain matters to the UK, including constitutional matters and the integrity of the Union itself.

During the first referendum, adjudication of whether Scotland needed UK approval to legislate an “advisory” (i.e. nonbinding) independence referendum was avoided, because the UK government ultimately issued an Order in Council temporarily transferring putatively reserved legislative powers to Scotland to carry out the referendum. However, this issue came to a head when the UK refused to cooperate with a second referendum. Then-Prime Minister Boris Johnson stressed the need for finality in a letter to First Minister Nicola Sturgeon:

‘You and your predecessor made a personal promise that the 2014 Independence Referendum was a "once in a generation" vote. The people of Scotland voted decisively on that promise to keep our United Kingdom together, a result which both the Scottish and UK Governments committed to respect in the Edinburgh Agreement. The UK Government will continue to uphold the democratic decision of the Scottish people and the promise that you made to them.’

The Reference: Reference by the Lord Advocate of devolution issues under paragraph 34 of Schedule 6 to the Scotland Act 1998, [2022] UKSC 31

The Lord Advocate (the senior law office of the Scottish Government) put the following reference question to the UK Supreme Court: Does the Scottish Parliament have power to legislate for the holding of a referendum on Scottish independence?

The UK Supreme Court approached the issue by addressing three questions:

  1. Does the Court have jurisdiction to consider the reference?
  2. If so, should the Court accept the reference?
  3. If so, then does the Scottish Parliament have power to legislate for the holding of a referendum on Scottish independence?

The UK’s law office in Scotland, the Advocate General, submitted the answer to all three questions was “no”.

On questions 1 and 2, Court held at paras 13-54 that it could and should hear the reference. Canadians may make fruitful comparison to similar jurisdictional and prematurity arguments raised by the amicus curiae in the Reference re Secession of Quebec, [1998] 2 SCR 217.

Question 3: Scottish Parliament cannot pass the referendum legislation

The critical question was whether the proposed referendum Bill would “relate to the Union of the Kingdoms of Scotland and England or the Parliament of the United Kingdom,” as such matters are reserved to the UK.

Section 29(3) of the Act provides that whether a Scottish legislative provision relates to the Union or the Parliament of the UK “is to be determined … by reference to the purpose of the provision, having regard (among other things) to its effect in all the circumstances.” The Court observed that “relates to” indicates more than a loose or consequential connection. However, a legislative provision need not have a “legal or direct effect” on a reserved matter to meet this threshold. 

The Court noted that it has repeatedly held that the purpose of the reservation of Union and United Kingdom Parliament matters is that “matters in which the United Kingdom as a whole has an interest should continue to be the responsibility of the United Kingdom Parliament at Westminster.” Such matters are “all fundamental elements of the constitution of the UK, and of Scotland’s place within it”.’ Measures which question the integrity of the United Kingdom, even if non-binding and “advisory”, are captured by this conception.

The Court rejected that ascertaining the purpose of the Bill is the same exercise as “purposive interpretation” of legislation. When assessing legislative competence under s. 29(3) the Court “is not attempting to construe the legislation in question”. On the contrary, the clearest indication of a Bill’s purpose may come not from its text, but from “a report that gave rise to the legislation, or in the report of an Assembly committee; or its purpose may be clear from its context.”

Similarly, the “effects” of a Bill for the purposes of a s. 29(3) reservation analysis are not confined to the direct effects prescribed by the legislation. Section 29(3) expressly contemplates the effects of a provision “in all the circumstances”, not just its legal effect. A provision does not have to modify the law applicable to a reserved matter in order to relate to that reserved matter.

Bearing this framework in mind, the Court easily concluded that the purpose of the proposed independence referendum Bill related to the reserved subject matters of the Union and the United Kingdom Parliament. The effects of the Bill would reach far beyond its non-binding legal effect. A lawfully held referendum, even if non-binding, is more than mere public consultation or public opinion polling. It purports to democratically reflect the views of the electorate on a public policy issue at a point in time. Statutory authority and public resources would be needed to define and carry out the referendum procedures. The referendum’s results would receive legitimacy from the official and formal character of the legal process, “even if the United Kingdom Government had not given any political commitment to act upon it.”

Indeed, the Court had previously held of the 2014 referendum, “the referendum is a very important political decision for both Scotland and the rest of the United Kingdom”. Of the Brexit referendum, the Court had written that while it did not change the law so as to allow withdrawal from the European Union, “that in no way means that it is devoid of effect”. Rather, it was “of great political significance.” Indeed, the Lord Advocate herself conceded the momentous political importance of the instant referendum in her submissions. “A clear outcome … either to remain within the United Kingdom or to pursue secession would strengthen or weaken the democratic legitimacy of the Union, depending on which view prevailed, and support or undermine the democratic credentials of the independence movement.”

The Bill’s relation to Union matters was even clearer than its relation to UK Parliament matters, as a referendum on independence is a referendum on ending the Union.

The Court addressed submissions, from the Scottish National Party intervener, that the right to self-determination is a fundamental and inalienable right in international law and that there is a strong presumption in favour of the interpretation of domestic legislation in a manner which is compatible with international law. The intervener submitted that these principles supported allowing Scotland to hold the non-binding independence referendum without UK permission.

The Court acknowledged the importance of these principles, and the well-established presumption that domestic legislation should be interpreted to comply with the UK’s international commitments. However, this presumption only applies to resolve ambiguity. The presumption favoring interpretations compatible with international law obligations only applies where domestic legislation is not clear on its face.

Relying on the Canadian Secession Reference, the Court held the principle of self-determination was inapplicable. The UK, like Canada, is a sovereign and independent state conducting itself in compliance with the principle of equal rights and self-determination of peoples and thus possessed of a government representing the whole people belonging to the territory without distinction. An international law right to external self-determination is only available in situations like former colonies, oppressed peoples (e.g. foreign military occupation), or a definable group denied meaningful access to government to pursue their political, economic, social and cultural development. The Court rejected that Scotland qualifies as any of the above. International law favors the territorial integrity of states. Barring scenarios like the abovementioned, there is no “right to secede”. 

Nor did these principles, even if used as an interpretive tool for the Act, support an interpretation that does not reserve the proposed referendum to the UK. The Act is clear, and it is to be interpreted the same way as any other Act. Nothing in the Act’s reservation of the proposed referendum to the UK was inconsistent with the principle of self-determination, for the reasons described above. The devolution of powers in the Act was carefully calibrated, and it “would be inappropriate to apply any interpretative presumption with the purpose of achieving a greater or lesser devolution of powers.”


Canadian readers can immediately take note of several points:

  • The presumption favouring territorial integrity of states is reinforced.
  • The Canadian Secession Reference continues to be internationally respected and influential.
  • The conditions of the Scottish people in 2022 are an additional data point, along with those of the Quebecois people in 1998, deemed not so intolerable as to legally justify unilateral independence maneuvers—even if they are non-binding.

While the Court did not weigh in on the wisdom of Scotland remaining in or leaving the UK, its conclusion that the Scottish people are not sufficiently oppressed to legally justify a unilateral referendum necessarily involves an implicit assessment of the plight (or lack thereof) of the Scottish people.  

Generally, it is doubtful whether the highest court of a country in fact denying a constituent people the right to self-determination would acknowledge such denial. Such acknowledgement would display an incredible degree of judicial independence in a host state engaged in repression of a constituent people. This may reflect a structural problem with referring secession matters to the highest court of a host state, and underscores the importance of maintaining faith in the independence of the judiciary in any state containing secessionist regions. Notably, Ms. Sturgeon announced the Scottish government’s intention to respect the legality of the Court’s ruling, despite commenting, “A law that doesn’t allow Scotland to choose our own future without Westminster consent exposes as myth any notion of the UK as a voluntary partnership”.

The reference decision raises interesting questions around repeat referendums. The Court readily acknowledged the democratic legitimacy of non-binding referendums; indeed, its legal analysis actually relied on their popular legitimacy as a reason why the proposed referendum would have sufficient political impact to qualify it as within reserved subject matters. A superficially counterintuitive result of this reasoning is that a non-binding referendum can be sufficiently democratic as to justify not putting it to the people.

However, as expressed by the former Prime Minister, the decision to remain in the UK was also an expression of Scottish democratic will less than ten years ago. Fundamental organizational changes, be they changing articles of incorporation or amending a national constitution, often have supermajority requirements or other such mechanisms for dampening the ability of shifting narrow majorities to frequently change the fundamental terms of association. In a similar vein, the UK government clearly seeks finality when it says the previous referendum has concluded the matter of Scottish independence for a generation. The Court, in using the language of a constitutional “settlement”, is consistent with a desire for predictability and stability. If bare majorities suffice to make legal changes of seismic constitutional importance, then referendums will need to be infrequent if the constitution is to be “settled” in any sense.  

Going forward, other governments contemplating temporary devolved “permission” for a major referendum may wish to promote finality by including in the terms of such permission a condition that the referendum result will be final and no further permission will be requested for a stipulated time period. Such language in the 2013 Order in Council and corresponding Edinburgh Agreement may have been more useful to the UK than the “personal promise” referred to by Mr. Boris Johnson in his letter refusing the UK’s cooperation. Conversely, constituent peoples yearning for independence may try to bargain for exceptions to such finality clauses for material changes in circumstances. For example, the 2014 referendum occurred two years before Brexit, in which Scotland overwhelmingly voted to remain in the EU. However, the Court already held in R (Miller) v Secretary of State for Exiting the European Union, [2017] UKSC 5, that exiting the European Union did not require the consent of devolved legislatures, even though the devolution Acts were passed on the assumption that the UK would be a member of the EU.

Interestingly, when the Advocate General tried to invoke the explanatory notes to the Act to argue the Court lacked jurisdiction to hear the reference, the Court commented that such explanatory notes post-date the Act and are therefore irrelevant to its statutory interpretation. This bears some contrast to the “living tree” approach described in the Secession Reference. While the British constitution is famously unentrenched in the sense that constitutional legislation is no harder to override than ordinary legislation, notably the Court has seemingly taken a special interpretive approach to the Scotland Act in the past because of its fundamental constitutional significance: e.g. BH & Anor v The Lord Advocate & Anor (Scotland), [2012] UKSC 24, at para 30.




Stay Connected

Get the latest posts from this blog

Please enter a valid email address