Power to the people?
What the decision in Miller means for the legal effect of referendums, and what any future enabling legislation should contain in light of the Brexit experience
The Miller ruling
People could be forgiven for thinking that the main question in R (on the application of Miller) v Secretary of State for Exiting the European Union  UKSC 5 was whether the result of the referendum on 23 June 2016 had not concluded the issue that the United Kingdom was to leave the European Union. But the effect of the referendum result was not the main question.
Article 50 of the Treaty on European Union (“TEU”) provides that any member state may decide to withdraw from the Union “in accordance with its own constitutional requirements”. If it so decides, it must notify the European Council of its intention. The main issue in Miller was whether formal notice of withdrawal could lawfully be given by ministers without prior legislation, in exercise of the royal prerogative (judgments, para 2).
In answering that question, the referendum was almost forgotten, as battle raged over two fundamental principles of constitutional law:
- the executive cannot, by exercise of prerogative powers, change the law of the land; and
- the making and unmaking of treaties falls within the prerogative powers of ministers, acting on the international plane.
The UK Government’s position was that legislation was unnecessary, because giving of notice under article 50 fell within the foreign affairs prerogative. That was so irrespective of consequential changes to domestic law. Its argument in support of that position was a clever, technical but ultimately rather narrow one. It found favour with Lord Reed, supported by Lords Carnwath and Hughes, but not with the eight justices in the majority.
The claimants’ argument, which the majority preferred, was a simpler and broader one. In the analogy drawn by Lord Pannick, notice under article 50 was pulling the trigger of a gun. Once pulled, the bullet would inevitably hit the target of statutory rights enjoyed under EU law, changing the law of the land without the consent of Parliament.
Lord Pannick’s argument was, in effect, accepted. EU law is sui generis domestic law. For that reason, the common law rule on the relationship between statute and the prerogative applies to it: “The essential point is that, if, as we consider, what would otherwise be a prerogative act would result in a change in domestic law, the act can only lawfully be carried out with the sanction of primary legislation enacted by the Queen in Parliament” (para 122).
The majority bolster their opinion by reference to broad constitutional principle. Withdrawal would constitute a significant constitutional change. On one view, the reason why Ms Miller’s claim succeeded is captured in a single sentence: “It would be inconsistent with long-standing and fundamental principle for such a far-reaching change to the UK’s constitutional arrangements to be brought about by ministerial decision or ministerial action alone” (para 81).
There are a number of issues worth considering which may reverberate well beyond Miller. These are addressed under four headings framed as questions.
(1) Why was the result of the referendum not decisive?
In his brief introductory remarks, the Attorney General submitted that the traditional view as to the limits of prerogative power should not apply to a ministerial decision authorised by a majority vote in a referendum provided for by Parliament. Towards the end of his submissions, James Eadie QC adopted a suggestion from one of the Justices along the lines of the Attorney’s introduction: even if parliamentary authority would otherwise have been required, the European Union Referendum Act 2015 and the referendum dispensed with that requirement. However, he did not develop the argument.
More spirited submissions attaching importance to the referendum were made by Lawyers for Britain, who argued that the Referendum Act had to be read in historical context. That included the facts that:
- holding referendums on major constitutional issues is a developing practice;
- referendums have been presented to the people as having binding effect;
- the Act being silent as to implementation of the result, it was necessary to have regard to the legislative object, which was to provide for a final and decisive result; and
- the only reading of the Act consistent with the background material, including ministerial statements, was that it confirmed (or, if necessary, re-conferred) authority to give notice under article 50 if a majority voted to leave.
The court’s analysis
Arguments based on the Referendum Act and the vote were rejected. The court’s analysis of the effect of the Referendum Act can be reduced to two sentences in the majority judgment: “The effect of any particular referendum must depend on the terms of the statute which authorises it... Where, as in this case, implementation of a referendum result requires a change in the law of the land, and statute has not provided for that change, the change in the law must be made in the only way in which the UK constitution permits, namely through parliamentary legislation” (paras 118 and 121).
In other words, the default setting for referendums is that they are merely advisory.
(2) Was the result of the referendum legally irrelevant?
The referendum vote was plainly of enormous political significance. However, the unavoidable conclusion which flows from the passage in the majority judgment just quoted is that it was legally irrelevant.
Curiously, that was also the conclusion which seemed to flow from the core argument advanced by the UK Government. Soft pedalling the effect of the referendum vote left that core argument looking unattractive for reasons identified in the majority judgment:
“The improbability of the Secretary of State's case is reinforced by the point that, if, as he contends, prerogative powers could be invoked in relation to the EU Treaties despite the provisions of the 1972 Act, it would have been open to ministers to take such a course on or at any time after 2 January 1973 without authorisation by Parliament. It would also follow that ministers could have taken that course even if there had been no referendum or indeed, at least in theory, even if any referendum had resulted in a vote to remain. Those are implausible propositions” (para 91).
Mr Eadie’s submission that if the Government had invoked article 50 in such circumstances, its decision might have been judicially reviewable, was described by the majority as a “bold suggestion”, but Lord Reed provided some endorsement for it: “it does not necessarily follow from my conclusions that ministers could properly have invoked article 50 whenever they pleased, or, more specifically, in the event of a vote to remain. As Lord Carnwath makes clear at para 266 below, there has been no discussion in this appeal of the question whether there might be any circumstances in which the exercise of the prerogative power in question might be open to review, such as if the referendum held under the 2015 Act had resulted in a vote to remain, and I express no view on that point” (para 239).
This takes us into the rather different territory of the exercise of the prerogative, as distinct from the scope of the prerogative which is what Miller was about. These remarks suggest, however, that this may be an interesting subject for another day.
(3) Need the result of the referendum have been legally irrelevant?
No. The default setting is that referendums are merely advisory, but default settings can be changed. Referendums can be organised in such a way that their outcome is legally, as well as politically, decisive.
An obvious example is the 1979 referendum on devolution. The Scotland Act 1978 provided for devolution. It stipulated that the minister should bring the Act into force if there was a specified majority in the referendum. If there was not, he was required to lay an order repealing the Act. A second example is the alternative vote referendum of 2011.
The Government has introduced the European Union (Notification of Withdrawal) Bill, which contains only one operative clause:
“Power to notify withdrawal from the EU
“1.–(1) The Prime Minister may notify, under Article 50(2) of the Treaty on European Union, the United Kingdom’s intention to withdraw from the EU.
(2) This section has effect despite any provision made by or under the European Communities Act 1972 or any other enactment.”
If Parliament had intended the referendum to be legally decisive, a section making provision along those lines in the event of a vote to leave could have achieved that result. There may be a number of explanations why the 2015 Act did not contain provision explaining what was to happen in the event of a vote to leave.
(4) Are there lessons to be learned for the future?
Leaving aside the result, the referendum has caused concern. The campaign was conducted in a fractious atmosphere. Both sides made claims of doubtful veracity. The result has left the country divided. There was a mini-constitutional crisis about the exit procedure. And the shape of the final Brexit deal remains unclear. What lessons are there to be learned from the referendum?
Be clear about the legal effect
A lesson likely to be learned from the EU referendum is the need to be as clear as possible in the legislation establishing the referendum about what the Government is to do after the vote, including if the result is not the one it is expecting.
Binary and pseudo-binary questions
If Parliament had intended that a leave vote was to empower the Government to give notice under article 50, it could and should have said so. That would have been quite easy, because whether or not to trigger article 50 is a binary question.
The wider question of whether to leave the EU has been described as a “pseudo-binary question”. While it was clear enough what a vote to remain meant, a vote to leave may lead in a number of directions, none of which were on the ballot on 23 June.
While referendums typically pose binary questions, they can offer more than two options. Early proposals for the Scottish independence referendum favoured three. It is difficult to see how this could have worked for the EU referendum. The way that withdrawal from the EU works under article 50 is that the decision to withdraw is taken first and is followed by negotiations on the terms of withdrawal. It would have been well-nigh impossible to have presented a small number of clearly defined options on the ballot.
What of a second referendum?
The British people voted for a departure, but not for a destination. Should they not be given a second vote once the final exit deal has been negotiated? That has some attractions, but a couple of potential drawbacks.
A negotiation between parties who have decided to divorce is not the same as one between parties who are merely thinking about it. The very fact of a second referendum would be bound to impact on the negotiations and likely affect the deal that was reached.
The second drawback is a technical, legal one. Article 50 does not provide for second thoughts. It was common ground in the Miller case that, once notification has been given, the process of withdrawal cannot be stopped. Not everyone agrees that that is a correct interpretation of article 50, but there is certainly no guarantee that withdrawal can be stopped. A second referendum might be too late.
To those who regard the vote to leave as a mistake or worse, the lesson to be learned may be a simple one: don’t hold referendums on important questions.
There may be force in that but, if not by referendums, how are proposals for major constitutional change to be determined? Some suggest that a general election is a better vehicle. There are a number of objections to that suggestion, not least of which is that general elections are never fought on a single issue.
It is said that referendums are alien to the British tradition of representative democracy. When the United Kingdom joined the European Communities in 1972, that was certainly true. But traditions change. The developing pattern in recent decades has been that proposals for important constitutional change have been put to the people in referendums. There was a substantial degree of consensus that a referendum was the right way to determine the question of Scottish independence. If that question is to be put again, there seems little doubt but that it will be by way of another referendum.
“No more referendums” may be too glib a response. But it is certainly worth reflecting upon the experience of the EU referendum and what can be learned from it in light of the Miller case.
Douglas Ross QC, Ampersand stable. A longer version of this article was delivered as a paper at the Ampersand seminar “The Constitution After Brexit and where we go now”, held on 6 February 2017.