In his resignation statement, David Cameron took it to be the case that triggering the UK’s formal withdrawal from the EU under Article 50 of the Treaty on European Union is a matter for the Prime Minister (or, more specifically, the next Prime Minister). Article 50, about which I have written in another post, provides that once a Member State has decided to withdraw from the EU, it must communicate its intention to leave to the European Council, thereby triggering a two-year negotiation period. A great deal has been written recently about when — and even whether — the UK should trigger Article 50, but the concern of this post is who gets to trigger it.

What is abundantly clear is that the Article 50 starting gun can be fired only by a decision taken by the UK, in accordance with its national constitutional arrangements, to withdraw from the EU. What is also abundantly clear is that the “decision” made by the people of the UK in the referendum is not a “decision” for the purpose of Article 50. The European Parliament, in a resolution adopted on 28 June 2016, therefore fell into error when it said that if — as it thought he should — the Prime Minister were to notify this week’s European Council of the “outcome” of the referendum, that “notification” would “launch the withdrawal procedure”. By eliding the “outcome” of the referendum and an Article 50 “decision” to withdraw, the European Parliament implied that the Prime Minister would initiate formal withdrawal merely by communicating the referendum result to the Council. That is wrong.

Putting to one side the European Parliament’s incorrect take on this, three possibilities remain in respect of who gets to make the Article 50 decision, thereby triggering the withdrawal process:

(1) The Prime Minister, by exercising prerogative power

(2) Parliament, by enacting primary legislation

(3) The Government, by making an Order in Council under powers conferred by the European Communities Act 1972 (“ECA”).

This post argues that the best view is the first one. The reasons for preferring (1) over (2) are explored below. Let me begin, however, with (3).

Is section 2(2) of the ECA an exclusive statutory basis for triggering Article 50?

In an intriguing post published on the UK Constitutional Law Blog, Adam Tucker argues that those who think the key question is whether option (1) or (2) is correct are overlooking a crucial detail in the European Communities Act 1972 — specifically, section 2(2). The relevant part of it provides that:

Subject to Schedule 2 to this Act, at any time after its passing Her Majesty may by Order in Council, and any designated Minister or department may by order, rules, regulations or scheme, make provision …

(a) for the purpose of implementing any EU obligation of the United Kingdom, or enabling any such obligation to be implemented, or of enabling any rights enjoyed or to be enjoyed by the United Kingdom under or by virtue of the Treaties to be exercised …

The burden of Tucker’s argument is that Article 50 vests in the UK a right to leave the EU that it previously lacked; that that right is a relevant right for the purpose of section 2(2)(a); and — crucially — that section 2(2)(a) forms the exclusive mechanism whereby the UK can take advantage of that right. (The exclusivity argument is founded upon an application of the De Keyser principle, which is examined further below when considering the relative merits of approaches (1) and (2).) It follows, it is suggested, that the only way for Article 50 to be triggered is by making an Order in Council under section 2(2) — which would take the form of a statutory instrument that would be subject to parliamentary override.

There are two objections to this. The first is that it is far from clear that the right which the UK would be exercising when triggering Article 50 is a right that it enjoys “under or by virtue of the Treaties”. It is at least arguable — but, I accept, not clear — that Article 50 merely constitutes a mechanism or process for the exercise of a right that the UK already had, and continued to have independently of Article 50, in general international law. If this is so, then Tucker’s argument fails because triggering Article 50 would not involve the exercise of a right arising “under or by virtue of” the EU Treaties.

There is, however, a second, and more fundamental, reason to doubt the argument that section 2(2) forms an exclusive — or, indeed, any — statutory basis for triggering Article 50. It seems to me to be quite clear that the purpose of section 2(2) is to enable the UK to make secondary legislation when the making of such legislation is necessary in order to enable rights under the Treaties to be exercised or in order to implement EU obligations or otherwise give effect to the EU Treaties. Indeed, section 2 is headed “General Implementation of Treaties”. There is, however, no need to resort to section 2(2) when domestic law already permits rights under the Treaties to be exercised — just as there is no need to resort to section 2(2) in order to implement Treaty obligations that are already taken care of by existing UK law. In such circumstances, the implementation powers conferred by section 2(2) are redundant.

It follows that section 2(2) does not bite in relation to Article 50. If there had been no prerogative power to trigger Article 50, section 2(2) could have been used to create a domestic power to do so. But because there is already a domestic power to trigger Article 50, section 2(2) is beside the point. In other words, there is no need to make UK secondary legislation enabling the UK to exercise whatever right Article 50 confers, because HM Government already has a legal basis for exercising that right — viz the royal prerogative. Section 2(2) is permissive (“may”) in relation to the making of secondary legislation in respect of both rights and obligations because it will sometimes simply not be necessary to use section 2(2) thanks to pre-existing position in UK law. Section 2(2) is therefore not an exclusive statutory mechanism for the exercise of the Article 50 power; it is a permissive statutory power that could be used to create a statutory basis for the exercise of the Article 50 power. But there is no obligation to use the section 2(2) power, and no need to do so, because a prerogative mechanism for the exercise of the Article 50 power already exists.

Does the ECA displace the relevant prerogative power?

If, as I think it must, approach (3) is to be put to one side, then the choice is between (1) (prerogative) and (2) (primary legislation). A number of commentators have recently argued in favour of (2), contending, in effect, that the ECA 1972, along with other relevant EU-related domestic legislation, displaces the prerogative power that the Prime Minister would otherwise have been able to use to trigger Article 50. For instance, Scott Styles writes:

In Britain it is Parliament which is sovereign, not the Prime Minister or even the whole Cabinet. The UK entered the EU by means of the European Communities Act 1972. The repeal of an Act of Parliament may only be done by a subsequent Act of Parliament to that effect. As the effect of an Article 50 notification is to trigger a two year timeline at the end of which that even UK would automatically cease to be an EU Member State that would be to nullify the effect of the 1972 Act as a matter of EU law. But as a matter of British law a statute may only be repealed by another statute. It therefore follows as a matter of British law that to have sufficient authority a Prime Minister would need that authority of an Act of Parliament to that effect giving him the authority to make an Article 50 notification and prospectively repealing the 1972 Act with effect from two years of making the notification.

This analysis is, to my mind, problematic, not least because it oversimplifies the way in which international (and EU) law interacts with domestic law. The UK became bound in international law by EU law as a result of an exercise by HM Government of its treaty-making prerogative. It was then necessary for pertinent EU law measures to be incorporated by Act of Parliament into domestic law. That was done via the ECA 1972. That Act, however, simply assumes binding EU obligations: it does not make them permanent or transfer the executive’s function in contracting those obligations to Parliament. Just as it is inaccurate, or incomplete, to say that the UK joined the EU by means of enacting the ECA 1972, that Act’s repeal is not a necessary component of Brexit, if Brexit is understood to mean the extrication of the UK from its EU Treaty obligations. The argument that by triggering Article 50 a Prime Minister would be taking upon herself or himself a power to “prospectively repeal[]” the ECA 1972 — and thereby asserting authority to do something that she or he lacked — is therefore incorrect.

However, a more subtle version of this argument is advanced by Nick Barber, Tom Hickman and Jeff King in a fascinating post. The core of their argument is that any prerogative power that would otherwise be available to the Government in order to make an Article 50 withdrawal decision is displaced by legislation enacted by Parliament, including the European Communities Act 1972 — which, among other things, enables directly effective EU law to have legal effect in the UK without the need for further domestic legislation. Barber et al base this contention on their understanding of how legislation and prerogative power interact. The general principle, as they correctly say, is that “statute beats prerogative” — a proposition that follows from the fact that Parliament is sovereign. However, statute can only “beat” prerogative if there is a relevant conflict between the two. In order to explore whether there is such a conflict, it is helpful to examine three key cases.

Case of Proclamations

In the first, the Case of Proclamations, Sir Edward Coke famously said that

the King by his proclamation … cannot change any part of the common law, or statute law, or the customs of the realm.

Giving notice under Article 50 does none of these things. Clearly, it does not change the common law. But equally, it does not change statute law, not least because — as noted above — withdrawal from the EU pursuant to the Article 50 process does not require or purport to effect any alteration to the ECA 1972. Certainly, an Article 50 process may, some way down the line, alter the effect of the ECA 1972 — by causing it to bite upon no or, in the event of a relevant deal with the EU, fewer EU rights and obligations — but using the prerogative to trigger Article 50 implies no assertion that prerogative power can be used to “change statute law”. It might, of course, be argued that this prohibition upon using the prerogative to “change statute law” should be read more broadly — a point that is considered below in relation to the third of our three cases. For the time being, it is sufficient to note that the use of prerogative power so as to adjust or extinguish Treaty obligations to which domestic legislation refers is some distance from the sort of mischief that Sir Edward Coke had in mind in the Case of Proclamations.

De Keyser

The second case is De Keyser, in which the Appellate Committee of the House of Lords had to determine whether the Government could use a broad prerogative power to requisition property, rather than using a statutory power which allowed the same but imported an obligation to pay compensation. The Appellate Committee held that only the statutory power could be used. Explaining this conclusion, Lord Dunedin said:

Inasmuch as the Crown is a party to every Act of Parliament it is logical enough to consider that when the Act deals with something which before the Act could be effected by the prerogative, and specially empowers the Crown to do the same thing, but subject to conditions, the Crown assents to that, and by that Act, to the prerogative being curtailed.

The principle established by De Keyser is clear. If legislation grants the Government a power to do something which the Government is also empowered to do under the prerogative, then the statutory power displaces the prerogative power if the statutory power is subject to conditions to which the prerogative power is not. That ensures that the Government is unable to circumvent legislation designed to safeguard the position of individuals by ignoring such legislation and instead relying upon broader prerogative powers.

But this principle reveals no relevant tension between prerogative and legislation in the Article 50 context. The ECA 1972 is not concerned with the extent of the Government’s powers to conduct foreign policy, either generally or specifically in relation to the extrication of the UK from its obligations under the EU Treaties (whether by means of Brexit or something falling short of that). And while it is true that the European Union Act 2011 regulates the exercise of the prerogative treaty-making power in certain EU-related respects, none of those respects is relevant to the present discussion. In the Article 50 context, there simply is not the form of specific overlap between statute and prerogative that is evident in either the De Keyser or Fire Brigades Union cases.

Fire Brigades Union

In the Fire Brigades Union case, the question was whether the prerogative could be used to establish a criminal injuries compensation scheme, given that such a scheme — a more expensive one — was already on the statute book albeit not yet in force. The Appellate Committee of the House of Lords concluded that the prerogative could not be used in such a way. As Barber et al point out, in this case Lord Browne-Wilkinson said that

it would be most surprising if, at the present day, prerogative powers could be validly exercised by the executive so as to frustrate the will of Parliament as expressed in a statute and, to an extent, to pre-empt the decision of Parliament whether or not to continue with the statutory scheme.

Barber et al go on to acknowledge that the scope of the principle for which the Fire Brigades Union case stands is contestable. A narrow reading of the case, they say, “would limit its application to situations where the statute proscribes in detail how Government must act, but where the Government circumvents that guidance by recourse to the prerogative”. Such a reading of the case plainly does not establish that the prerogative cannot be used to trigger Article 50.

It is then argued that on a broader reading the principle is that “it is not open to Government to turn a statute into what is in substance a dead letter by exercise of the prerogative powers; and that it is not open to the Government to act in a way which cuts across the object and purpose of an existing statute”. And it is contended by Barber et al that, on this reading of the Fire Brigades Union case, legislation should be taken to have precluded the use of prerogative power to trigger Article 50, such that an Act of Parliament would be necessary to start the Article 50 process.

This forms the strongest strand of the argument put forward by Barber et al, and I do not reject out of hand the possibility that a court, presented with the crucible of legal, constitutional democratic arguments that would be in play were this matter ever to be tested, might take such an expansive view of legislation’s capacity to delimit the prerogative as to make the argument tenable. Nor, however, do I reject the possibility that a court would do the opposite bearing in mind the macro-political nature of the issues at stake.

In any event, there are two aspects to the argument advanced by Barber et al that lead me to the conclusion that the better view is that the prerogative remains. First, it is not the case that triggering Article 50 amounts to the Government’s turning the ECA 1972 into a dead letter, since the outcome of any Article 50 process cannot be known. Such a process might result in an agreement that the UK should remain a member of the EU on altered terms, such that the ECA 1972 would continue to bite upon a substantial set of EU-related matters, or that the UK should become a member of the European Economic Area, in which case a substantial corpus of EU law, upon which an amended ECA 1972 might continue to bite, would remain pertinent to the UK. Equally, an Article 50 process would ultimately amount to nothing if it were to be aborted. (It is plain that it can be aborted by agreement, and arguable that the UK could abort it unilaterally.)

Second, and more fundamentally, it is far from clear that invoking Article 50 would — to use Lord Browne-Wilkinson’s words in the Fire Brigades Union case — “frustrate the will of Parliament” vis-à-vis the ECA 1972. Whether invoking Article 50 would have the effect of frustrating the will of Parliament depends on what it is that we deem Parliament to have been trying to do when it enacted the ECA 1972. The Act is centrally concerned to ensure that “such rights, powers, liabilities, obligations” and so on as are “from time to time provided for by or under the Treaties” have effect in UK law. The Act does not, however, confer any particular rights upon anyone. Instead, the purpose of the Act — or of the will of Parliament in enacting it — can be understood to be enabling the UK to discharge such obligations as it has from time to time under the EU Treaties, by enabling EU law to have such effect in the UK as those Treaties require. The ECA 1972 leaves entirely open the possibility that those Treaties might cease to so require. If the EU Treaties were no longer to require EU law to have effect in the UK because the UK was no longer a party to them, that would therefore not amount to a frustration of the purpose of the ECA 1972. Rather, the Act would continue to do the job for which it was enacted, namely giving such effect to EU law in the UK as the Treaties might require at any given point in time. A similar analysis can be applied to the European Parliamentary Elections Act 2002, to which Barber et al also refer, the purpose of which can best be understood as the fulfilment of such Treaty obligations in respect of the European Parliament as the UK had when the legislation was enacted.

Another way of looking at this is to acknowledge that the Government and Parliament played different, and complementary, roles in securing EU membership, and that they will (or may) play different, and complementary, roles in terminating such membership. Just as it was the UK Government, exercising prerogative power, that caused the UK to be bound by EU Treaty obligations, so it is for the Government, using prerogative power, to extricate the UK from those obligations — including by triggering the Article 50 extrication process itself. Meanwhile, just as it was for Parliament to enact such domestic legislation as EU membership required (such as the ECA 1972), it is equally for Parliament to enact any domestic legislation that Brexit may in due course require. On this analysis, no tension between the ECA 1972 and the prerogative arises because they concerned with distinct spheres of activity, the one operating on the plane of diplomacy and international law, and the other operating on the plane of domestic law.

This is not to deny that domestic law could not have the effect of curtailing the prerogative, but as legislation that simply facilitates discharge of Treaty obligations entered into under the prerogative, the ECA 1972 does not amount to a statute that cuts across the prerogative. Indeed, if it were the case that legislation giving effect to Treaty obligations were to extinguish any prerogative power to renegotiate or extricate the UK from such obligations, it would be necessary to enact legislation every time the Government wished to secure such renegotiation or extrication — and that simply does not happen. The fundamental point, then, is that legislation — like the ECA 1972 — facilitating the discharge of treaty obligations does not occupy the same legal space as, and therefore does not conflict with, the Government’s prerogative power to contract, renegotiate or extricate the UK from treaty obligations.

Closing remarks

Just because, on my analysis, the Prime Minister can trigger Article 50 without reference to Parliament, it does not follow that that would be a wise or sensible thing to do. Triggering Article 50 would be a highly significant step, given that it would open up the possibility of — even though, as discussed above, it certainly would not render inevitable — the wholesale departure of the UK from the EU. In such circumstances, the case for parliamentary involvement is strong. Indeed, in other contexts — most notably the use of the prerogative to deploy the armed forces abroad — there is an increasing expectation, and arguably a constitutional convention, concerning parliamentary involvement. In the Article 50 context, there is no equivalent established convention that requires parliamentary involvement, but there is certainly a normative argument in favour of such involvement that could in due course form the basis of a convention.

Even if one dismisses the possibility of “ignoring” the result of the referendum, much remains to be decided — including about the UK’s interests would best be served by triggering Article 50 or seeking to proceed in some other manner — and there are excellent democratic reasons for arguing that Parliament should play a full part in those deliberations. As we are rapidly discovering, the volume and complexity of the issues left unresolved by the binary view expressed by the electorate is immense, and Parliament has a crucial role to play in shaping the way forward. For all that the UK has experimented with direct democracy through the holding of a referendum on EU membership and on other constitutional matters, the UK remains, fundamentally, a parliamentary democracy, and it cannot plausibly be argued that the referendum substitutes for proper parliamentary involvement.

But such normative arguments are a distinct issue from the question whether Parliament, as a matter of law, must be involved at the outset, by way of enacting primary legislation firing the Article 50 starting gun. For the reasons given in this post, the better view is that Article 50 can be invoked by the Prime Minister using prerogative power, without the involvement of Parliament.

I am grateful to a number of academic colleagues, including Catherine Barnard, Lorand Bartels, David Feldman, Hayley Hooper, Aileen McHarg, David Mead, Philip Murray, Gavin Phillipson and Alison Young, and to Jack Williams, a pupil barrister at Monckton Chambers, for their invaluable comments on and discussion of matters related to this post. The usual disclaimer applies.