UK Constitutional Law Association

affiliated to the International Association of Constitutional Law

Cormac Mac Amhlaigh: Can Brexit Be Stopped under EU Law?

Ominous clouds are gathering and the terrain underfoot increasingly resembles a quagmire on the Brexiteers’ ‘sunlit uplands’. The latest reminders that the reality will be significantly different from their utopia of a prosperous global Buccaneering Britain has come in the form of a trade dispute between the U.S. and a Canadian aircraft manufacturer which could have a devastating impact on the Northern-Irish economy where the manufacturer has a significant base; and the threat from a gang of countries that they will not accept a proposed agreement (one of the few agreements for now) between the EU and UK as to the divvying up of agricultural import quotas after Brexit.  Perhaps most galling on this front is the fact that the gang involves those with whom it was hoped trade deals would be swiftly struck; including  the U.S. and New Zealand.

It is therefore unsurprising that the chatter about revoking the Art. 50 notification to withdraw from the EU – itself waxing and waning since the referendum vote – has become louder in recent days; spurred on by a freedom of information request seeking the government’s legal advice on the question.  The debate on whether notification could be revoked had reached something of a standstill in the aftermath of the  Miller litigation before the UK Supreme Court, where an opportunity for clarity on the question was lost as both parties to the litigation proceeded on the basis that such notification was not revocable, and attempts to get the CJEU to rule on the issue in separate litigation was thwarted by the Irish Courts.  Thus, the issue is still unresolved.  In this post I want to make the case that an Art. 50 notification is not revocable.  The argument, in summary, is this:  the text of Art. 50 itself is inconclusive and there are compelling consequential reasons for finding that such notification is not revocable in that allowing the UK to revoke its notification could create serious moral hazard risks which could undermine the rule of law in the EU, the authority of the EU Courts and make the Union ungovernable in the future.

Arguments based on Art. 50 itself

Influential voices have weighed in on the question, arguing that an Art. 50 notification is indeed revocable, including former director-general of the Council of the EU’s legal service, Jean Claude Piris as well as the House of Lords EU Committee.  The argument for its revocability centre on two main pillars:  arguments based on international law and the text of Art. 50 itself.

Arguments based on international law

This argument runs as follows:  as the Treaties are agreements between sovereign states, the sovereignty of each EU member state includes the right to notify to withdraw but also the right to revoke such notification at any stage prior to actual withdrawal.  As such, the right to revoke is essentially read into the sovereignty of EU Member States.

The sovereignty of the EU’s member states – and the agency, rights and duties that accompany it – are, of course, recognised and upheld by International law.  However the problem here is that international law is not directly relevant to the question.  It is uncontroversial that international law recognises the general right to accede to and withdraw from international treaties as a function of the sovereignty of states, however international law also recognises that where states have agreed to regulate matters in a particular way, particularly in the terms of an international treaty, the treaty provisions will be applied over general international law.  As such, Art. 50 is a lex specialis, which would be enforced in any dispute about the logistics of withdrawal over the provisions of general international law and in particular the general rights and duties of sovereign states.  Thus, the general point about international law is subject to the Art. 50 procedure which simply begs the question:  does Art. 50 allow for a revocation of notification of withdrawal?

Another tack taken along these lines is recourse to Art. 68 of the Vienna Convention of the law of treaties.  This provision clearly and unambiguously states that ‘A notification [for withdrawal] may be revoked at any time before it takes effect’.   However, it is not clear that this is completely relevant to Art. 50 (for one thing, it stipulates a slightly different procedure which includes the possibility of other states objecting to the withdrawal), but even if it was, the European Union is not bound by it.  Not all EU member states are signatories to the convention (Romania, for example, is not a signatory) but more significantly, the EU itself is not a signatory.  As such, it does not bind EU law (I’m assuming, I think correctly, that this issue does not form part of customary international law as per, for example, the CJEU’s Firma Brita decision) and can therefore have no direct bearing on the revocation question under Art. 50.  Therefore, the legal question invariably turns to the interpretation of Art. 50 itself.

None of this is to discount the possibility that the CJEU may draw on international law in answering this question, however it would treat the issue as entirely a matter of EU law where Art. 50, and particularly the protection of the EU’s constitutional architecture, would take precedence over international law on the matter.

The text of Art. 50

Two issues in the text of art. 50 have been prominent in the revocation debate:  the legal relevance of ‘intention’ and the relationship between ‘intention’ in Art. 50 (2) and ‘withdrawal’ from the EU in Art. 50(5).

Various claims have been made about the legal significance of ‘intention’ to the effect that it has no legal status or that the law recognises that intention can be overturned as it is not a binding commitment.  However, the role and relevance of intention in the law is contingent;  in some contexts intention is very relevant, even central (such as in the crime of murder) in others completely irrelevant (such as strict liability for harm caused).  ‘The law’, therefore, does not have any general position on the role and relevance of intention; it will depend on the ‘the law’ in question  and here we are concerned with EU law.  I know of no provision of EU law, nor do proponents offer any specific authority in support, on the status of an ‘intention’ under EU law.

However, even were intention to have an autonomous meaning under EU law,  there are good reasons to believe that the legally relevant issue under Art. 50 is not ‘intention’ but ‘notification’.  Only a formal and clear act of notification can provide the requisite certainty and clarity which the law is called upon to provide in the withdrawal procedure.  It is notification which starts the clock ticking on the two year time period under Art. 50(3) TEU, not the ‘intention’.

Relatedly, the arguments based on the difference between Art. 50(2) and 50(5) do not change this position.  The idea here is that given that 50(2) only speaks of ‘intention’ to withdraw, and the fact that 50(5) on the procedures for readmission only relate to withdrawal,  that Art. 50 envisages that there is a difference between an intention to withdraw and actual withdrawal such that the former is revocable but that latter is not.  Put another way, if notification and withdrawal were synonymous under Art. 50, then the procedure for reapplying under art. 50(5) would explicitly refer also to notification.  However the problem here is that there is no definition of ‘withdrawal’ under Art. 50(5).  The legal effect of leaving the EU is that EU law ceases to apply (subject to whatever agreement is put in place between the parties in its lieu).  This is covered by art. 50(3) which clearly states that this state of affairs (the closest Art. 50 comes to a definition of withdrawal), occurs two years from notification (unless otherwise agreed by the parties).  There is nothing here to suggest that Art. 50(3) envisages that this notification could be contingent on some other intervening act such as a revocation of such notification.

Therefore, it is submitted, that the evidence on the text of art. 50 TEU as to whether notification can be revoked is at best inconclusive, at worst, points in favour of a no-revocation position.

Consequential reasons against revoking notification

However this is not the only reason why notification cannot be revoked.  Arguably a more significant reason against permitting revocation is the moral hazard risk involved.  Moral hazard is a familiar term in economic theory which entails the idea that an actor can make a risky or potentially costly decision knowing that it is unlikely to bear the cost of that decision based on past behaviour.  It was prominent in debates about Greece’s bail-outs after the Euro-crisis (‘if we cancel Greek debt now, then future Greek governments will continue to borrow and spend recklessly, safe in the knowledge that they will not be left shouldering the costs of their spending as we’ve bailed them out before’).  Let’s imagine, then, that things come to a head: the UK is so unprepared, banks leave London en masse and Japanese car manufacturers flee such that the UK government decides the only way to prevent economic catastrophe is to notify the European Council of its intention to withdraw its notification of its intention to withdraw from the EU.  EU Member states deliberate on whether to accept it and the Court of Justice of the EU (CJEU) has to decide.  It may suggest that this is a political matter and, if everyone is in agreement, then such revocation would be valid under EU law. There would probably be a lot of peeved EU Member States; annoyed that the UK is playing games with its own economy, its international standing and the EU project.  They might be tempted to not accept the revocation and force the UK out.  But their better judgement will prevail, they will look at the figures, they’ll look at the economic damage Brexit will cause to the EU’s economy (and particularly their national economies), they’ll think about the potential loss of citizenship rights of 65 million EU citizens and think better of it.  Revocation of notification will be accepted and everything will return as it was to 22 June 2016.

However, then think about the long term implications of this acceptance of revocation.  Even if it is generally accepted that Brexit will leave the UK worse off than the EU, the EU will still suffer.  It has undergone a crisis of confidence after Brexit along with all the other fires it has recently fought including Greece, the Eurozone and its twin rule of law and refugee crises with a new separatist crisis looming in Spain.  Even if the domino effect of other Member States following the UK out the exit door did not transpire, the threatened withdrawal of any state does raise profound existential questions for the bloc.  Allowing the revocation of an Art. 50 notification would create significant moral hazard risks by putting a potentially powerful weapon in the hands of individual member states.  This could be used to gain leverage in bargaining over unpopular laws or policies or even to overturn unpopular CJEU decisions.  We don’t need to look very far to see how this could play out.  A future Greece wishing to force euro debt cancellation or mutualisation could invoke Art. 50 to push others around to their position; a disgruntled Poland or Hungary, smarting at rule of law procedures, refugee quotas, or both, could communicate its intent to withdraw under Art. 50; even Europhile Ireland, uncomfortable with an adverse CJEU judgment on state aid involving Apple tax breaks or how policy-making on corporate tax harmonisation is proceeding undermining its efforts to keep corporate taxes low to attract inward investment, could trigger Art. 50.  This would hinder the development of EU policies on a whole host of issues and threaten the authority of EU law and the EU Courts.  Of course their bluff could be called, other Member States could ignore their pleas for policy-change or overturning a CJEU ruling.  However if the gambit didn’t pay off, the threatening state could simply withdraw the notification of intent before they formally left following the Brexit precedent.  Leaving the door open to such strategic manipulation of EU decision-making processes would undermine the rule of law in the Union, the authority of its courts, and the integrity of EU governance.

Moreover, these moral hazard risks apply regardless of whether the revocation of notification is unilateral or not.  As such, it matters little whether the revocation is done with the agreement of all involved (which the Commission seems to suggest would be sufficient for a valid revocation of withdrawal) or whether the revocation is attempted in the absence of the agreement of the remaining Member States.  Either way the risks to the integrity of the EU’s constitutional architecture and its decision-making processes and general governance remain.  It is pretty clear that the CJEU could not open the door to such risks, potentially undermining its authority and more importantly, the integrity of the EU’s constitutional architecture, under its watch.   In this light, it is almost certain to find that notification cannot be withdrawn for these significant policy reasons.

Of course, even if these reasons apply also to cases of general agreement among the member states that a notification could be revoked, it doesn’t mean that politics, and the political will of member states have no role to play here.   Where the political will was present to prevent Brexit, some sort of solution could be found under EU law.  A stage-managed ‘Potemkin Brexit’ could be arranged whereby the UK formally leaves, but keeps the rights and duties of membership under a status-quo-preserving transnational agreement, and re-joins under Art. 50(5) – if all went very well for the UK – on the terms it currently enjoys.  Such a move is not without political risks – not least for the UK – and the requisite political will is currently in short supply on both sides.  And it would merely vindicate the position that an Art. 50 notification cannot be withdrawn under EU law.

My thanks to Alison Young for helpful comments on a previous draft.

Cormac Mac Amhlaigh is Senior Lecturer at the University of Edinburgh.

(Suggested citation: C. Mac Amhlaigh, ‘Can Brexit Be Stopped under EU Law?’, U.K. Const. L. Blog (10th Oct. 2017) (available at

17 comments on “Cormac Mac Amhlaigh: Can Brexit Be Stopped under EU Law?

  1. Roger
    October 10, 2017

    It doesn’t matter if it’s revocable or not. The vote is done, it is not the disaster you try to make out and we will be leaving; the UK’s Article 50 notice will not be revoked.

    It would be the end for European countries, the EU bureaucrats would ride roughshod over its peoples. Get used to it and stop undermining your own country’s position.

  2. mark hardy
    October 10, 2017

    The question remains, under EU law, what happens if it is judicially determined that there needed to be a decision by Parliament to leave the EU before the PM gave notice of intention for that notice to be of domestic or international effect. Perhaps the word “may” rather than “must” in the EU(Withdrawal) Act will be seen to be of far more relevance than “intention” in 50(2)

    I suspect the ECJ will merely rule that the 50(2) notification was of no effect under EU law given the failure to meet the domestic constitutional requirements and revocation would not need to be considered for a judicial opinion.

    A most interesting preliminary point is does the Divisional Court have jurisidiction to even hear such an application in the light of the SC judgment?

    It is worth reading Dominic Chambers QC original skeleton argument and the transcripts in the divisional court where HMG conceded the decision matter and that is why the judgment was only in relation to the Royal Prerogative point (para 4) and therefore not even on appeal to the SC.

    You may find it interesting to consider this in the light of HMG’s recent verbatim response (see below) to pre-action protocol letters in anticipation of a JR application on that matter.

    HMG’s current position as expressed by the Treasury Solicitor:
    “Article 50 sets out the procedure by which a Member State which has decided to withdraw from the EU may achieve that result. That decision having been taken, the nest stage in the process is for the state to notify the European Council of its intention to withdraw. Parliamentary approval for the Prime Minister to notify the UK’s intention to withdraw from the EU was given under the European Union (Notification of Withdrawal) Act 2017. Parliament passed this Act in full knowledge of the Government’s intention to notify under Article 50.”
    “……the Supreme Court in [Miller] did not hold that the effect of Article 50 was that “Parliament” had to make a “decision” to leave the EU before notification could be given. Rather, the Court held that the UK’s constitution meant that authorisation of the notice of the UK’s intention to withdraw must be given in primary legislation passed in both Houses of Parliament and assented to by Her Majesty the Queen. The European Union (Withdrawal) Act 2017 satisfied that requirement. The Prime Minister then acted within the power given to her in the 2017 ACt in giving the notification. No other constitutional or legislative steps were required”

  3. Dr Mike Tremblay
    October 11, 2017

    Moral hazard is a type of “gaming”, in this case of treaties. It is commonplace and practised by governments all the time, including within the EU. The UK games the treaties on all sorts of things so why stop with Art 50? Indeed, it appears to be a common practice in the UK civil service to view treaties as contingent subject to whatever advantage can be gained. As such, therefore, the goal is bargaining, not moral hazard per se.

    We, whether states or individuals, are always looking for advantage in negotiations, and sometimes invoke conditions that create the impression of moral hazard, but simply represent a presentation of a position designed to create an unpleasant alternative to the other side (read Getting to Yes for instance). I suspect that the current game of chicken the UK is playing will end badly for the UK; studies of French bargaining behaviour (given Barnier’s background) suggest they are unlikely to play this game, but rather precondition for a given outcome. The UK hasn’t understood this for perhaps 400 years.

    So what if countries game Art 50? Even despite the best efforts of the EU, nation states exist in a state of anarchy with respect to each other (they can put national interests first for instance). They game just about everything possible in the EU itself, to gain advantage, protect advantage, to avoid loss, etc. Plus ça change, plus c’est la même chose.

    Gaming may be evidence of policy failure and/or weak democratic processes in the EU itself. Nuclear options are a last resort, though, as they have, as noted, serious consequences. But when your back is against the wall, that is usually when you read the handwriting on it. Brexit has forced both the EU and the UK to read that handwriting. This is both therapeutic and risky, depending on what outcome you seek.

    I suspect when it was drafted, leaving the EU was seen as a black swan event and therefore the clause was not subjected to very close analysis. On a simple Wilson matrix, it has unintended and perverse consequences. That suggests to me that it is poorly worded and deserves to be gamed.

  4. Malcolm Ramsay
    October 11, 2017

    There’s another level of moral hazard that the author doesn’t seem to have considered: if courts take too much responsibility for saving people from the consequences of their own failures, they encourage sloppy drafting of contracts and treaties.

    I would expect the CJEU to take into account that current EU treaty provisions have been shown to be inadequate for the complexity of a member state withdrawing, that Britain’s Article 50 notification has been the means of bringing that inadequacy to light and that EU member states are in a position to agree new treaties that properly reflect the difficulties. Any moral hazard is purely short term and could anyway be short-circuited by the EU making clear that any future notifications will not be taken seriously unless the member state concerned explicitly agrees in advance that it is irrevocable. I don’t see any good reason for the CJEU to base a judgement on the moral hazard argument.

    Personally, though, I think anyone desperate to remain would be foolish to count on our current crop of politicians being sensible enough to revoke notification, even if leaving is clearly going to be a disaster. A better option, to my mind, would be to challenge the constitutional validity of the original notification on the grounds that, by failing to establish that there is either a clear public interest in taking such a drastic step or an unequivocal public mandate to do so regardless of any adverse consequences, the government have breached their duty of care to the public. Of course, making that argument would involve opening a constitutional can of worms that many people would rather keep the lid on, so perhaps no-one will be bold enough to try it.

  5. Mr. Wayne Ramwell
    October 11, 2017

    Ultimately, assuming your position on irrevocability is correct, although I am not convinced because nothing has materially changed in terms of membership yet (and would not even during a fixed transitional period) and some who were involved in drafting article 50 intimate that it is indeed revocable, and notwithstanding that the final decision on the interpretation of EU law rests with the CJEU, the threat(s) to the authority and constitutional architecture of the EU order could easily be mitigated against by the UK losing some benefit of current membership or facing some form of consequence devised by the EU so as to discourage invocations of article 50 by other member-states merely to exert political pressure etc. Whilst I accept that some form of punishment or negative consequence to the UK suggests that revocation differs from current membership, it does not necessarily mean revocation becomes impossible based on EU policy reasons. At best, revocation arguably becomes undesirable from an EU perspective, but still nevertheless possible.

  6. Alberta
    October 11, 2017

    Why would you do that? Negotiations are the key to find some solutions that work for both.

  7. David Howarth
    October 13, 2017

    I can’t see how the VCLT can be irrelevant to article 50 when the VCLT was expressly referred to the in travaux préparatoires of article 50 itself. See What is irrelevant is the fact that a couple of EU MS have not ratified the VCLT They assented to the Lisbon Treaty and that’s enough.

    Moreover, if states really cared about the moral hazard point there is no way article 68 would have survived into the adopted version of the VCLT in its present form. The moral hazard point would imply the exact opposite of article 68, namely saying that notices of withdrawal should be taken as irrevocable unless the treaty provides otherwise. That article 68 doesn’t say that is highly significant.

  8. AustralianActuary
    October 15, 2017

    Don’t you think that if Brexit could be stopped under EU law that would be the single most potent argument as to why Brexit is necessary?

  9. Pingback: Robert Craig: Why an Act of Parliament Would Be Required to Revoke Notification under Article 50 | UK Constitutional Law Association

  10. Temple10
    October 16, 2017

    To hoist a less than Jolly Roger, pending the inevitable gambit of litigation and regulatory upheaval in perspective in 2019 there seems to have been little thought given anywhere to what Court would regulate the post-Brexit position of the United Kingdom in relation to the EU, given that there will inevitably be both constitutional issues but also legal and regulatory issues to be tried and heard.

    To infer that either the UK system or the CJEU system is presently capable of shouldering the additional workload in this are would be delusional. The effects upon the Dependencies still left within any area in common with the EU post Brexit would be unjudiciable.

    The prospect of the UK leaving Euratom without some form of Court able and competent to regulate the ensuing issues as to transfers of radioactive and strategic atomic materials could be catastrophic.

    Serious thought should be given by the current UK Court system’s guiding minds, and perhaps the Treasury Solicitor, as to the creation of an intermediate but separate CJEU/UK to handle the post Brexit regulatory, compliance and litigation issues. If there is to be an interregnum, there needs to be a court.

    There would need to be either a joint presidency of this institution or a regularly rotating one, one from the EU and one from the UK, two lists of advocate-generals, each list providing an AG for each hearing and submitting before the court to ensure that both sides of any argument are ventilated.

    Either then two official languages, English being by default one, and perhaps French the other. Or, if it is really necessary to do so, any one of the European languages, for a given case but not necessarily all, with dedicated back-up from the current CJEU linguistic department.

    At the end, a place for the Court to sit. In order to show that it is not part of the present EU order, it should not be Luxembourg, but either London or another British city, or why not Paris? There may be good economic reasons for another venue.

    It would however be necessary to ensure a similar level of quasi-diplomatic/international protection and a fiscal and social régime for the judiciary, AGs and persons populating the Courts’ administration.

    By so doing, an indirect jurisprudential influence and counterweight can then be had over the CJEU’s position in Luxembourg without the UK being subservient to it.

    This is but the outline of what could be a tailored and workable investment for the future.

    Just an idea.

    Peter Harris

  11. NarusReport
    December 5, 2017

    According to the latest report capturing the views of 31 legal experts published by the NarusReport, a legal periodical, 74% consider that the UK can unilaterally revoke the Brexit withdrawal notice. The report and underlying raw data are freely available.

  12. Jacq
    January 3, 2018

    Think you are right about lex specialis. After the referendum, a UK Parliament committee heard how the EU used wider international law in its external dealings; obviously internal dealings are defined by the treaties.

    There is no provision for rescinding notification in TEU/TFEU.

    Note also that EU negotiator Michel Barnier was quoted that there was no going back on withdrawal (Evening Standard, 13 Dec 2017) and the Commission refused a Citizens’ Initiative to stop Brexit. So we should accept the reality and rather be focused on making it work for everyone.

  13. Pingback: Time For a Re-Think? – Brexit Time

  14. Pingback: Blocking Brexit: Article 50 in the Scottish Court of Session - Free Movement

  15. Pingback: Robert Brett Taylor and Adelyn L. M. Wilson: Seeking and Implementing a Referral on Revocability of Article 50 Following Wightman | UK Constitutional Law Association

  16. Pingback: Luxembourg court to be asked if the UK can cancel Brexit - Free Movement

  17. Pingback: Can An Article 50 Withdrawal Notice be Revoked? The CJEU is Asked to Decide | Verfassungsblog

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Ordinary membership

UKCLA yearly membership (ordinary)


Student membership

UKCLA yearly membership (student)


Associate membership

UKCLA yearly membership (associate)


%d bloggers like this: