A Referendum with no Legal Effect?
By Ira Ryk-Lakhman, on 13 July 2016
The people have spoken; the UK has decided to leave the EU. Or has it? As lawyers pore over the legal consequences of the result there have been a large variety of views expressed on the constitutional steps necessary for the UK to leave the EU. But on one question there appears to be unanimity: the referendum result had no legal effect.
Some consider that the UK can only decide to leave through an Act of Parliament (Barber, Hickman and King; Ewan Smith; Lord Pannick QC). A second view is that withdrawal is a matter for the Royal prerogative (Mark Elliott; Kenneth Armstrong). While yet a third contends that the decision must be taken pursuant to the statutory powers in s.2(2) European Communities Act 1972 (Adam Tucker). However each of these views – mostly without further elaboration – accepts the notion that, legally speaking, the British people have really not decided anything at all.
Mark Elliott has gone so far as to say that it is ‘abundantly clear’ that the referendum is not a decision to leave the EU. Adam Tucker opines that it is ‘plainly wrong’ to count the referendum as a decision. This post seeks to challenge this absolutist view. As a matter of principle and authority, there are good reasons for saying that the UK has already decided to leave the EU in a manner which satisfies Article 50(1) of the Lisbon Treaty: (1) the intention of Parliament was to give the final decision to the electorate, (2) the notion that, in the absence of an express term as to its legal effect, a referendum is obviously ‘advisory’ is contradicted by highly persuasive authority in the Supreme Court case of Moohan v Lord Advocate. However, this conclusion does not entail that the UK’s decision to leave is irreversible; as will be argued below, the UK can yet change its mind.
The procedural requirements for withdrawing from the EU are determined by Article 50 of the Treaty of Lisbon. For present purposes the relevant part of the provision reads as follows:
- Any Member State may decide to withdraw from the Union in accordance with its own constitutional requirements.
- A Member State which decides to withdraw shall notify the European Council of its intention. […]
The procedure for withdrawal adopts a two-step process for getting the ball rolling: Article 50(1) decision and subsequent Article 50(2) notification. According to the ordinary meaning of Article 50, the constitutional requirement only applies in relation to the decision, not to notification. Barber et al. elide the two provisions under the heading of an Article 50 ‘declaration’ and appear to question the constitutionality of notification. While there may also be constitutional issues in relation to notification, for the strict purposes of Article 50 these are irrelevant. Thus, a constitutionally valid decision followed by an unconstitutional (perhaps ultra vires) notification could still satisfy Article 50. This is the case under the rules of international law which holds the State responsible for acts performed by its agents under apparent authority even if they are unlawful under domestic law (see Articles 3, 7, Articles on Responsibility of State for Internationally Wrongful Acts; Articles 7, 46, 47, Vienna Convention on the Law of Treaties; and, inter alia, dicta in Treatment of Polish Nationals case, pp.24-5 and ELSI case, para.73; further see Ewan Smith).
The pressing question is, therefore, whether the referendum was a decision to withdraw ‘in accordance with [our] constitutional requirements’.
Common-sense dictates that the referendum, in response to a specific question on whether the UK should leave/remain in the EU, constitutes a decision to withdraw from the EU. Richard Ekins makes the point (outside the context of Article 50): ‘Parliament made clear that the decision about whether to leave the EU was to be settled by referendum’.
‘[t]his is so because the referendum – legally speaking – was purely advisory. The legislation that allowed the referendum to take place did not invest the outcome of the referendum with any sort of legal effect. The UK Government is therefore not legally obliged by the referendum to trigger the Article 50 process, either at any particular point in time or at all.’
In support of his argument, Elliott might have cited the difference between the EU Referendum Act 2015, which says nothing of its legal effect, and the AV Referendum Act 2011, which had certain self-executing effects and engendered a legal obligation on the executive to perform specific acts. On Elliott’s view, a referendum is only binding in the latter case.
Except that this argument is flawed. As a matter of domestic law, there is no need for a referendum to have self-executing effects for it to be a decision to withdraw. Could the executive now decide to remain? Such a decision would seem to render the 2015 Act nugatory. If Parliament delegated the decision to the electorate, this would suggest that it has removed the power to decide from the executive (cf. Laker Airways Ltd. v Department for Trade  1 QB 643).
The issue is what Parliament intended when enacting the 2015 Act. It seems a common-sense construction that Parliament intended to give the decision over whether we withdraw to the electorate. Moreover, a glance at Hansard will find no reference to ‘advisory’ but reveals repeated statements about giving the people the ‘final say’. The Foreign Secretary, Philip Hammond MP, presented the Bill to Parliament in the following terms: “whether we favour being in or out, we surely should all be able to agree on the simple principle that the decision about our membership should be taken by the British people not by Whitehall bureaucrats, certainly not by Brussels bureaucrats; not even by Government Ministers or parliamentarians in this Chamber. The decision must be for the common sense of the British people.”
That the absence of express terms as to the legal effect of a referendum result can nevertheless produce a legally effective decision is also supported by persuasive Supreme Court authority. In Moohan v Lord Advocate  UKSC 267, the court was faced with a human rights challenge to the franchise for the Scottish Referendum. Like the 2015 Act, the Scottish Referendum Act 2013 contained no express term as to its legal effect. One of the arguments put to the Supreme Court was that the referendum was ‘purely consultative’ as a result. Lord Kerr rejected this, stating at -:
Both the United Kingdom and Scottish governments had agreed that the result of the referendum would be binding […]
Similarly Lord Wilson stated on the same question (at ):
[it] could not be said of the Scottish referendum [that it was ‘purely consultative’].
These statements contradict Elliott’s assertion and fatally undermine the case that it is impossible for the referendum to satisfy Article 50(1) simply because it does not state its legal effect on its face. Instead there is a good case for saying that Parliament intended that the referendum was a decision for the purpose of Article 50. The only power left to the executive, therefore, is when to leave the EU, not whether to do so.
Is it a reversible decision?
If the argument above is correct, this does not entail that the decision as a matter of domestic law is so binding that it is irreversible. It is trite law that Parliament cannot be bound by the decision. Parliamentary sovereignty vests Parliament with the power to overturn any decision (even notwithstanding that it had been taken by a previous Parliament) (see Lord Neuberger at - in Moohan) even if doing so might be ‘far-fetched’ (per Lord Wilson at ; cp. dictum of Lord Sankey LC in British Coal Corporation v King  AC 500, p.520).
The question must instead be over whether the decision is reversible under Article 50. There is an argument to be had, beyond the scope of this post, over whether the process for withdrawal following notification is reversible. The question here is: whether the withdrawal decision is reversible prior to notification?
Article 50(2) says that the state ‘shall notify’ which at first glance might appear to create a binding obligation once the decision is made to issue the notification, suggesting that the decision is irreversible for the purposes of Article 50. However the term is more ambiguous than it first appears (see also Professor Besselink’s post). The two main purposes of Article 50(2) notification are that (1) notification obliges the EU to negotiate a withdrawal agreement with the departing state; (2) it enables the departing state to leave after two years without the agreement of other states (Article 50(3)). As a matter of general international law, a state who decided to withdraw from a treaty would ordinarily be unable to do so except by agreement with the other treaty parties (Articles 54-6 VCLT). Therefore, Articles 50(2)-(3) enable the withdrawing state to force the issue; in this respect, the notification procedure exists for the benefit of the withdrawing state.
A teleological interpretation would suggest that Article 50(2) ought not to be read so as make the decision irreversible. It would defeat the purpose of the sovereign right of unilateral withdrawal if Article 50 were read so as to enable the EU to force a state, which has changed its mind prior to initiating the withdrawal process, to leave the EU.
To say that the referendum result is obviously not a decision to leave the EU for the purposes of Article 50 seems untenable. It is an arguable proposition, both in principle and as a matter of authority that Article 50(1) has been satisfied by the referendum result and the UK could now issue a notification under Article 50(2) that could result in the UK leaving the EU by 2018. But as a matter of both domestic and EU law, until that notification is served, Parliament retains the power to keep us in the EU, no matter how politically ‘far-fetched’.