Gavin Phillipson and Alison L. Young: Wightman: What Would Be the UK’s Constitutional Requirements to Revoke Article 50?

Today the Court of Justice of the European Union delivered its judgment in Wightman. This followed the opinion of Advocate General Campos Sánchez-Bordona, concluding that the UK may unilaterally revoke its notification of its intention to leave the EU. In a similar manner to the AG, the CJEU placed conditions on this unilateral revocation. A formal process would be needed to notify the European Council of the UK’s intention to revoke article 50. Such notice of revocation would have to be unequivocal and unconditional (para 74), and, importantly, ‘in accordance with the constitutional requirements of the Member State’, in this case, the UK, and following a ‘democratic process’ (para 66). It would also have to take place before the end of the Article 50 negotiation period, or any agreed extension, and before a Withdrawal Agreement between the exiting state and the EU had been ‘concluded’ – i.e. entered into force (para 73). In addition, the AG’s opinion was that any revocation would have to be in ‘good faith’ and in line with the requirement of ‘sincere cooperation’ between the Member State and the EU and. Further, although not required, it would be reasonable for the Member State to provide its reasons for revoking the Article 50 notification.

This raises an important question for UK constitutional lawyers:  what would be the constitutional requirements for the UK to lawfully notify the Council of its intention to revoke its Article 50 notification in order to remain in the EU? In particular, is AG Campos Sánchez-Bordona correct in his suggestion that ‘if the national constitutional requirements include, for example, prior parliamentary authorisation for the notification of the intention to withdraw from the European Union (as is the case in the United Kingdom, according to the Miller judgment), it is logical, in my view, that the revocation of that notification also requires parliamentary approval’ (para 145)?

This post will argue that primary legislation would be legally and constitutionally required in order to empower the Prime Minister to revoke Article 50. It will contend that the clear legal arguments in favour of this conclusion are reinforced by concerns of both political reality and arguments of constitutional legitimacy.

The need for legislation in UK law

Is the Advocate-General right to suggest that Miller means that legislation would be needed to empower the Prime Minister to revoke Article 50, in a similar manner to the legislation found to be required to empower the Prime Minister to trigger Article 50? Miller concluded that legislation was needed due to limits on the scope of prerogative powers and their lawful use. It found that triggering Article 50 was beyond the scope of the broad prerogative power to enter into and withdraw from Treaties as it would lead to a modification of domestic law, remove rights, and frustrate the purpose of the European Communities Act 1972 (‘ECA’) and the European Parliamentary Elections Act 2002.

Not all of the above limitations on the prerogative are applicable here. Unlike use of the prerogative to leave the EU, its use to revoke the UK’s Article 50 notification would not remove rights or alter domestic law. The UK remains a member of the EU throughout the negotiation period, and will lose this status only when it leaves the EU. As such, revoking Article 50 would merely maintain the status quo in terms of domestic law and the protection of rights. The rights, powers, liabilities, obligations, restrictions, remedies and procedures UK citizens currently enjoy through their membership of the EU would simply continue as Article 50 was revoked.

The principle that we do argue is applicable and engaged is that the prerogative may not be exercised in a way that frustrates the intention of Parliament as expressed in statute.  Critics of Miller might argue that the majority judgment did not rest on this principle but on the more controversial contentions that notifying Article 50 would result in the removal of a source of law in the UK (EU law) and hence precipitate a major constitutional change. However, those arguments may be regarded as flowing from the majority’s recognition of the unique constitutional importance of the ECA, which lead them to emphasise that rendering it a dead letter via the prerogative would actually remove a (de facto) source of law from the UK Constitution and thereby make a major change to the constitution.  However, we contend that even if one disagrees with those arguments as applied to the ECA, that does not affect the central argument we make in this blog.

That is because Miller did not invent the principle that the prerogative may not be exercised in a way that frustrates the intention of Parliament. Rather the majority judgment explicitly endorsed it as a pre-existing principle, flowing from the prior, unchallenged authorities of Fire Brigades Union (‘FBU’) and Laker Airways, which they cited with approval (para 51). Lord Browne Wilkinson summarised the basic principle in FBU:

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…

The majority in Miller affirmed this basic principle. ‘Ministers cannot frustrate the purpose of a statute or a statutory provision, for example by emptying it of content or preventing its effectual operation’ (para 51).  This principle follows logically from a basic aspect of parliamentary sovereignty – that Acts of Parliament rank higher than other sources of law, including the prerogative. Moreover, Article 1 of the Bill of Rights 1689 makes clear that the prerogative may not be used to ‘dispense with or suspend the execution of the laws’.  The dissenting judgments in Miller did not call the frustration principle into question; they merely considered that it did not apply, given their construction of the nature and purpose of the ECA. Thus our argument does not need to rely on any controversial aspect of the majority judgment in Miller and indeed would be the same even if Miller had never been decided. It simply applies the well-established frustration principle, which derives from FBU and Laker, and was expressly affirmed in Miller. 

What are the statutes whose purposes may be frustrated by revocation of Article 50? We suggest that there are two – the European Union (Notification of Withdrawal) Act 2017 and the European Union (Withdrawal) Act 2018 – and we take them in turn.

Frustration and the European Union (Notification of Withdrawal) Act 2017

Of the two Acts, it is less clear that the European Union (Notification of Withdrawal) Act would be frustrated. If we merely look at the title of the legislation, then its purpose may appear to be frustrated, given that it refers to the UK’s notification of its intention to withdraw. Moreover, section 1 states, ‘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’. Whilst the purpose of this legislation was to empower the Prime Minister to notify the EU of the UK’s intention to leave in order to implement the wishes of the majority who voted in the 2015 referendum, the statute itself only grants a power. It does not require the Prime Minister to notify the European Union. The words ‘intention to withdraw’ also play an important role. These words mirror the requirements of Article 50(2) TEU. As the Advocate-General recognised, intentions may change. The decision of Parliament to give the Prime Minister power to notify the EU of the UK’s then intention to leave is not thwarted if the UK’s intention changes and the PM subsequently also notifies that changed intent.

Nor would the 2017 Act be rendered devoid of purpose by revocation. The legislation was used to trigger Article 50, negotiations have taken place, and a Withdrawal Agreement and Future Framework drawn up and agreed. Indeed, if the statute’s purpose had not been fulfilled then there would no need for the Prime Minister to revoke article 50; the UK would not need to revoke a notification of intention to withdraw if that notification had never been given to the EU under the power bestowed by the Act. So, whilst an argument may be made that this legislation would be frustrated, we do not consider it a particularly strong one. However, since we will argue that primary legislation is in any event needed due to the 2018 Withdrawal Act, our conclusion on this Act is not decisive for our overall argument.

Frustration and the European Union (Withdrawal) Act 2018

In relation to the European Union (Withdrawal) Act 2018 (EUWA), a much stronger argument can be made that using the prerogative to revoke Article 50 would frustrate its purposes. First, the whole purpose of the Act is to facilitate the UK’s actual exit (not its mere initial intention to exit) from the European Union. In our view, this would be frustrated were the UK to revoke Article 50 using the treaty prerogative. It could be argued that the 2018 Act merely provides evidence of an intention to leave the European Union, but that it is not a direct expression of that intention. But this is unpersuasive: the Act provides for the implementation of this intention, aiming to achieve certainty and stability by enabling existing EU law to continue to be applied in the UK as EU-derived law, unless and until this is amended by delegated or primary legislation. It also provides a means of remedying deficiencies in retained EU law which arise due to leaving the EU, or making provision to implement the Withdrawal Agreement prior to exit day.

However, an argument about the EUWA could be made that was similar to one advanced in Miller in relation to the European Communities Act 1972 (ECA). That statute was said by the Government side merely to provide a means of facilitating EU membership should the Government decide to join the then EEC, by allowing, inter alia, such EU rights and obligations as existed from ‘time to time’ to take effect in UK law. As such, it was argued that the ECA’s purpose would not be frustrated were the category of those rights and obligations to become an empty one (as would be the case when the EU Treaties ceased to apply to the UK).

Similarly it could be argued that the EUWA merely provides the means to implement the withdrawal of the UK from the EU should the UK Government finally decide to do so under the prerogative. However, this narrow approach to statutory purpose was emphatically rejected by the majority in Miller. It would be odd, therefore, if the courts were to prefer this more narrow reading of the EUWA having used a more purposive reading of the ECA in Miller. In a question of equal constitutional importance to that raised in Miller, there are strong constitutional arguments for adopting the same approach.  

This conclusion is reinforced when we examine the following provisions of the EUWA, which are currently in force (much of the Act has not yet been commenced). Sections 8 to 11 of the Act empower Ministers to enact delegated legislation when the Minister considers it appropriate to deal with defects that arise when the UK leaves the EU, or in order to implement the Withdrawal Agreement prior to exit day. These sections would be frustrated by revocation. It would no longer be the case that measures would be required to implement the Withdrawal Agreement, or to remedy defects in delegated or primary legislation arising from the UK’s withdrawal from the EU. Hence these provisions would be rendered devoid of purpose.

Sections 12(2), 12(4) and 12(6) of the Act modify the powers of the devolved legislatures and executives. Prior to the Act coming into force, the devolved legislatures were unable to enact legislation which contravened EU law. Nor could the devolved executives act in a manner which contravened EU law. Section 12 modifies this restriction, replacing it with a restriction not to act contrary to the provisions of section 12 of the EUWA. To revoke Article 50 unilaterally would frustrate the purposes of section 12, the aim of which is to allow powers repatriated from the EU to return to the devolved legislatures, enabling them to modify EU-derived law, unless Ministerial orders of the UK government have temporarily removed this power from the devolved legislatures in order to develop common frameworks.

Section 20 meanwhile defines ‘exit day’ while s 20(4) provides for a power to vary exit day by ministerial Order. This provision too would be rendered devoid of purpose were the UK to revoke; moreover it appears fairly clearly to provide evidence of parliamentary intention to exit, without which the whole notion of ‘exit day’, which runs through the provisions of the EUWA, would be meaningless.

Section 13 provides for the meaningful vote. Whether it would also be frustrated by revocation is perhaps more debatable. It might be argued that the narrow purposes of this section is only to provide for a process of parliamentary approval in relation to the Withdrawal Agreement and the Framework for the Future Relationship should the Government wish to ratify the former. In other words, it could be seen as merely a conditional provision, such that, were the Government to decide instead to revoke the Article 50 notice, the condition required for its application would not have come into play.

The contrary argument is that the aim of this provision is to provide Parliament with the opportunity to debate, and vote on, the Withdrawal Agreement and the Framework for the Future Relationship if they are agreed with the EU, as is the case now. The condition precedent for making section 13 “live” has thus been fulfilled and the Government has laid the relevant documents before the House, thus commencing the application of the process set out in that section. Revoking article 50, using the prerogative, would render both the Withdrawal Agreement and Future Framework otiose and hence frustrate the purpose of section 13, which is to enable Parliament to approve or disapprove those instruments, once agreed. (If the Government’s Withdrawal Agreement were to be approved by the meaningful vote, only for the Government later to revoke Article 50, then section 13 could be said to be frustrated after the fact because the approval of the Withdrawal Agreement that Parliament had bestowed would be rendered retrospectively nugatory).  More broadly, it could be said that the purpose of this provision is to enable Parliament to exert control over the decisive stage of the Brexit process; were the Government to revoke its Article 50 notice using the prerogative, that would seem to run counter to that broader aim.

An implied statutory power to revoke?

The above has assumed that the foreign affairs prerogative would be the domestic law power used to revoke Article 50. It could be argued instead that there is an implied statutory power to revoke, flowing from the 2017 Notification Act. Plainly there is no mention of any such power in the statute. Hence the argument would have to be that a power to notify under Article 50 also included the implied corollary power to withdraw which, the CJEU has now found, inheres in Article 50. Or it could be suggested that the power bestowed by the 2017 Act was to commence and conduct the negotiations with the EU that notifying Article 50 entails, and that this more general power included the power to withdraw the notice. This argument does not seem particularly strong, especially given our conclusion that the basic purpose of the Act was simply to empower the Prime Minister to notify the UK’s then intention to leave. To find an implied positive power to revoke in the face of that basic purpose would seem something a stretch, to say the least.

Law, politics and the constitution

The above arguments are reinforced by the following considerations of political and constitutional legitimacy. Given that a majority of those who took part in the referendum voted to leave the EU and Parliament has passed two Acts that allowed for the UK to withdraw from the EU, it would seem inconceivable that the Executive, acting alone through the prerogative, could legitimately make the decision to cancel that process via revocation.

Moreover, giving Parliament a full legislative role in the decision to revoke may be relevant to an EU law requirement. The A-G saw the need for adherence to national constitutional requirements as providing evidence for the EU law requirement that a Member State was not abusing its power to unilaterally revoke Article 50 by using this tactically to negotiate a better deal. To revoke Article 50 without legislation might not satisfy this evidential requirement, especially when coupled with the AG’s suggestion of reciprocity between the criteria for triggering and revoking Article 50. The decision of the CJEU appears to take this requirement further. To allow a Member State to revoke Article 50 without adhering to a democratic process ‘would be inconsistent with the Treaties’ purpose of creating an ever closer union among the peoples of Europe’ (para 67). Moreover,  ‘to force the withdrawal of a Member State which, having notified its intention to withdraw from the European Union in accordance with its constitutional requirements and following a democratic process, decides to revoke the notification of that intention through a democratic process’ would further undermine the EU’s value of respect for democracy (para 67). Use of the royal prerogative alone could surely not satisfy the requirements of ‘a democratic process’.

Our main legal conclusion is clear: should the decision be made to revoke the UK’s Article 50 notice, legislation would be required explicitly empowering the Prime Minister to do so. For the avoidance of any doubt, any such legislation would need to state that the power to revoke applied ‘notwithstanding any provision’ in the 2017 and 2018 Acts ‘or any other enactment’. At the very least, there is a strongly arguable case that legislation would be required: to seek to revoke without it would simply invite immediate litigation.

A further referendum?

Many argue that democratic legitimacy would require not just an Act of Parliament but also a second referendum to be held before a decision to revoke Article 50 could properly be made. Their argument may be reinforced by the decision of the CJEU, with its focus on the need for reciprocity and democratic legitimacy to uphold values of liberty and democracy in the EU – although it is clear that it is for the UK and not for the EU to determine the UK’s constitutional requirements.

Should the Commons refuse to approve the Withdrawal Agreement and Future Framework in the vote on the 11th December then a further motion is required under section 13(4)-(6) of the EUWA, taking note of how the Government proposes to proceed, in light of that rejection. However even were an amendment to be passed to that motion that called for a further referendum, that would not itself have any empowering legal force. Fresh legislation would still be needed to authorise the holding of any such referendum.

Were Parliament to pass such legislation, and were it to provide that remaining in the EU was to be one of the choices on the ballot paper, then that Act could provide the statutory authorisation for revocation that we have argued is constitutionally required. Indeed, to avoid arguments about the referendum’s legal status, it would make sense for the enabling Act to specify that, in the event that a majority voted ‘Remain’, the Prime Minister was required to revoke the Article 50 notice. Thus the referendum legislation itself could provide the necessary legal authority to the Prime Minister. It is hard if not impossible to see that such legislation could be passed and a referendum held before 29 March. Hence such a course would almost certainly require the UK to request an extension under Article 50(3), which would require the unanimous approval of the European Council.

The authors would like to thank Robert Craig for comments on an earlier version.

Gavin Phillipson is Professor of Law at Durham University. He is also currently a Parliamentary Academic Fellow in the Parliament and Constitution Centre in the House of Commons Library. The views above represent his personal view, which should not be taken as representing the views of the Library.

Alison L Young is the Sir David Williams Professor of Public Law, Robinson College, University of Cambridge.

(Suggested citation: G. Phillipson and A.L. Young, ‘Wightman: What Would Be the UK’s Constitutional Requirements to Revoke Article 50?’, U.K. Const. L. Blog (10th Dec. 2018) (available at