UK Constitutional Law Association

affiliated to the International Association of Constitutional Law

Stijn Smismans: About the Revocability of Withdrawal: Why the EU (Law) Interpretation of Article 50 Matters

stijn-smismansIn the Miller case the High Court of England and Wales did not question the parties’ common understanding that withdrawal under Article 50 cannot be revoked once triggered. However, academic opinion on the issue is divided. Interestingly enough, the academic debate on Article 50 has particularly developed from a UK constitutionalist perspective (understandably as far as the decision to withdraw has to be taken in accordance with the Member State’s own constitutional requirements (Art.50, para.1)), and even from an international law perspective (Aurel Sari).  However, the interpretation of Article 50 is primarily a matter of EU law. This is even more the case if the issue gets to be decided by the CJEU.  While national courts may be more inclined to interpret EU law in the context of national and international law, the CJEU’s frame of reference is above all EU law.  As a referral to the CJEU has become more likely since the Miller case has been passed on to the Supreme Court, the EU (law) interpretation of Article 50 needs to be taken more seriously.

Moreover, even in the absence of a CJEU ruling on Article 50, it may well come to the point that the EU Institutions have to take a stance on accepting or not a revocation of withdrawal.  If the Supreme Court declares Article 50 revocable, the chances increase that the UK might make use of the option to revoke withdrawal, particularly given the current political constellation, which is increasingly set towards a Hard Brexit.  Yet, if the UK does intend to revoke its withdrawal the final word on the acceptability of revocability will lie with the EU Institutions, which may or may not accept the UK’s declared intention.  From this perspective, a referral by the Supreme Court to the CJEU for a preliminary ruling may create legal certainty about what the UK can expect if it does decide to revoke its withdrawal.

In making their decision on whether a request to revoke withdrawal can be accepted, the EU Institutions will be advised by their Legal Services.  If the CJEU has given its interpretation via a preliminary reference they will adhere to that.  In the absence of any such ruling, the Legal Services will still give an interpretation that relies primarily on EU law.  For both the CJEU and the Legal Services, an EU law interpretation also means an ‘EU interest’ interpretation that ensures the integrity of European integration and the balance of interests set out in the Treaties.  It is on this EU (law) perspective I focus my analysis.

I will first analyse the procedural ways through which the ‘EU interpretation’ of Article 50 may come to light.  I will then argue that this EU interpretation of Article 50 will above all rely on a teleological interpretation, referring to the broader objectives and institutional checks and balances set out in the Treaties, rather than on a reference to international or national law.  I will finally argue that the teleological interpretation provides a ‘EU interest’ perspective, which neither corresponds with the exit-friendly perspective of international law, nor with the exit-hostile perspective that presents exit as a completely automatic consequence of triggering the Article.  The power balance created by Article 50 implies that the UK has no right to revoke but will depend on the willingness of the other Member States to allow such revocation.  I will conclude by addressing the consequences for the Miller case.

The procedural route to the EU interpretation of Article 50

The most likely route to the EU interpretation of Article 50 TEU is the preliminary reference procedure (Article 267TFEU).  Its basic principles have been extensively rehearsed in the recent commentary on the Miller case.  Article 267TFEU allows national courts to ask the CJEU for clarification of EU law whenever doubt arises on the correct interpretation of EU law.   Moreover, if the national court is acting at last instance, it is obliged to make a preliminary reference if 1) it considers that there is indeed lack of clarity in the interpretation of EU law and 2) a decision on the correct interpretation is necessary to decide the case.  This leaves a degree of discretion to the national court on whether to refer or not (even when acting at last instance).   As was expected, the High Court did not make a preliminary reference as it was not acting as the court of last resort and it probably preferred avoiding the political turmoil and delay that would come with involving the CJEU.

However, now the Government has decided to appeal against the decision in the Supreme Court a referral to the CJEU becomes more likely, as the Supreme Court is acting in last resort.   It is difficult to argue that Article 50 does not raise a question of interpretation.  The Supreme Court is faced now with two High Courts going in  opposite directions on the question of revocability.  While the High Court in London has taken the non-revocability argued by the parties at face value, the High Court in Belfast seems to suggest the opposite in McCord. The Court did not make an explicit statement on revocability but argued that: ‘On the day after the notice has been given, the law will in fact be the same as it was the day before it was given. The rights of individual citizens will not have changed – though it is, of course, true that in due course the body of EU law as it applies in the United Kingdom will, very likely, become the subject of change. But at the point when this occurs the process necessarily will be one controlled by parliamentary legislation, as this is the mechanism for changing the law in the United Kingdom.’ (McCord 105).  This paragraph can be interpreted in two ways.  Maybe the Court intended to say that even in case of non-revocability, exit triggered by Article 50 would not automatically affect statutory rights even at the end of the negotiation(failed or succeeded) because it would be Parliament putting that outcome into law.  This interpretation, though, is not correct as exit would automatically affect statutory rights (see below).  Alternatively, the paragraph suggests that Parliament has still the option to revoke withdrawal between triggering and the two year deadline.  In the latter case, Miller and McCord take the opposite view on revocability.  In Ferreira da Silva (C-160/14) the CJEU said the Portuguese Supreme Court should have referred to the CJEU as lower courts had been giving  conflicting interpretations of EU law (although the CJEU may have been led particularly by the fact that courts in several countries had given divergent opinions) (Daniel Sarmiento).  Moreover, it would be difficult to argue that it is not necessary to answer this interpretative question to resolve the dispute.  The argument of non-revocability was key for the claimants and the Court to conclude that Parliament needs to be involved prior to triggering Article 50.   If Article 50 were revocable, its triggering would not lead to the automatic loss of statutory rights and Parliamentary involvement would not be required at this stage.

The fact that both the claimants and the defendant agreed on the triggering of withdrawal under Article 50 not being revocable does not imply that the correct interpretation of the Article is not essential to resolve the dispute.   It is up to the Court to provide the correct interpretation of the law necessary to resolve the dispute, and investigate and inquire (if needed with reference to the CJEU) on the correct legal interpretation, rather than rely on a (potentially) wrong interpretation of the law on which the parties agree.

The two key conditions for requiring a referral to the CJEU are thus met.  I do not agree with some arguments put forward in commentaries on the Miller case that claim that the Supreme Court should not refer.

George Peretz argues that the second condition, namely ‘interpretation is needed to decide on the case’, is not met, because even if notification is not revocable, triggering the Article could still lead to something else than not withdrawing, such as postponing the deadline of negotiation, or Treaty reform leading to the UK staying in.  This is not convincing. Postponing the deadline of negotiation does not change the fate of a withdrawal procedure.  Regarding the possibility of Treaty reform; such a procedural option is not provided in Article 50.  Legally it is as ephemeral as a changing of the wind in Downing Street.  Both are external events that might impact the desire to ask revocation of withdrawal, but whether that is legally possible is precisely the question to be addressed by the Court.  Moreover, even if one accepts that Treaty reform is a third option under Article 50, besides exit agreement or automatic exit after two years, it would still imply that triggering Article 50 automatically leads to the situation in which Parliament has no control on rights protected under domestic law as it would be up to all national Parliaments to ratify such a new Treaty.  Hence, only if Article 50 is revocable one could be sure Parliament can keep control over all these statutory rights. Answering the question about revocability or not thus remains essential to decide the case.

Another critique of the Miller case might at first sight be more convincing to argue that a referral is not needed to decide the case.  John Finnis, Mark Elliot & Hayley Hooper , and David Feldman have argued that triggering Article 50 does not affect statutory rights but only EU rights, although made possible by Statute.  In that view, the European Commnities Act 1972 (ECA) has given Government the power to act internationally to adopt these EU rights, and thus is also considered in a position to remove them.  If that is the case, the question of revocability becomes irrelevant. However, Jeff King and Nick Barber have convincingly argued against this view.  The fact that the ECA gives the executive the power to participate in the definition of EU rights, does not imply it has the unilateral power to remove them.  The executive’s power has to be seen in the context of the general scheme of rights provided in the ECA, which includes, among others, reference to adoption of EU rights via EU procedures and giving EU regulations direct effect.

Hence, when Section 2(1) ECA states that ‘all such rights, powers, liabilities, obligations and restrictions from time to time created or arising by or under the Treaties’ shall be given legal effect in the UK,  the expression ‘from time to time’ means the executive can participate at EU level in defining that variation but it is contemplated within the framework of on-going membership and the extensive treaty provisions governing how such rights are varied.  King and Barber convincingly argue that the title and content of the ECA evince a commitment to membership, and that it is not simply a conduit for international law rights, giving Government the discretion to commit or withdraw from EU rights ‘from time to time’ as it pleases.  Moreover, besides the ECA, the European Parliament Elections Act 2002 is on its own enough of an argument to show that statutory rights are at stake. Hence, as domestic statutory rights are at stake, the right interpretation of revocability is required to decide on the case.

Finally, some authors have contested directly the competence of the CJEU to give an interpretation of the revocability of withdrawal under Article 50, or of the national courts to make such a referral.

According to Mike Wienbracke a preliminary reference should not be made because withdrawal in accordance with a Member State’s ‘own constitutional requirements’ (Article 50(1) TEU would not be a matter for interpretation via Article 267TFEU.  However, reference to ‘own constitutional requirements’ does not undo that this is a Treaty text and therefore under the competence of the CJEU to interpret. As Lang rightly argues, it has to be referred to the CJEU, even if the latter may immediately send it back to the national court to interpret what ‘own constitutional requirements’ means under national law.  Wienbracke’s reading also lifts paragraph one of Article 50 out of its context.  It suggests that withdrawal is only a matter of national law, which it is not.  It is also a matter of EU law, and the procedure for it has been set out in the entirety of Article 50.   It is for the CJEU to judge on the issue of revocability within the context of the procedure as set out in Article 50 and the Treaties more broadly.  For that reason, referring to the CJEU is not simply a question of procedural correctness, but the Court may well have more to say on the issue than a simple statement that it is up to the national court to interpret ‘own constitutional requirements’.

The argumentation of Mikolaj Barczentiwicz is similarly flawed.  He argues that the Supreme Court cannot refer to the CJEU because the European Communities Act 1972 never granted national courts the power to ask for a preliminary ruling from the EU Court on matters of withdrawal since it can never have been Parliament’s intention to bind the UK to those EU rules when the very question is whether it should be bound by those rules.  Like Wienbracke, Barczentiwicz sees withdrawal as an entirely unilateral process under national law, while Article 50 makes very clear this is also a negotiated procedure set out at EU level.  The interpretation of the latter is up to the CJEU and cannot be left to the judiciary of a Member State which has doubts it still wants to be bound by those rules.  Where would Barczentiwicz draw the line if we follow his reasoning? Potentially all aspects of the implementation of EU law during the withdrawal process could become disputed and fall under the sole interpretation of national courts.  Such a ‘pick and choose’ approach cannot be defended. Article 50 provides for a negotiated withdrawal process, but it assumes that the withdrawing country remains a full member of the EU until the moment of exit (with all that it implies including accepting the competence of the CJEU).  If a country wants to unilaterally withdraw under international law, that is possible. But if it a wants a negotiated withdrawal, it cannot consider itself to be a partial Member on its own terms during the withdrawal process.

Of course, the fact that the legal conditions for referral to the CJEU are fulfilled does not guarantee that the Supreme Court will do so, given the political sensitivity and risk of delaying the triggering of Article 50.  However, if the Government is changing its position on revocability, it becomes legally more plausible for the Supreme Court to refer.   Delay also should not be a major concern as the CJEU is likely to act via expedited procedure (Daniel Sarmiento).  Moreover, despite the political sensitivity, there is also a good political argument to refer to the CJEU.   As mentioned above, if it comes to the stage where the UK would intend to revoke its withdrawal, it is important to know whether the EU Institutions accept that an Article 50 notification can be withdrawn.  To know whether EU Institutions are in the position to decide on that, it is better to find out the opinion of the CJEU beforehand, rather than building the political strategy on the assumption that revocation is possible while running the risk it is not.

Whatever the outcome in the Supreme Court, the parties do not have a right to appeal against it in the CJEU.  It is also highly unlikely that the European Commission would take the UK to court under the infringement procedure if the Supreme Court failed to ask for a preliminary ruling.  The Commission has once started an infringement procedure for this reason (against Sweden), although the issue was settled before it came to Court.  However, in general the Commission is reluctant to start an infringement procedure against a Member State for failure by its judiciary to respect EU law, as this is politically sensitive (but see C-129/00 Commission v. Italy [2003] ECR I-14637). It is even more unlikely it would do so against a departing Member State.   In theory there is still the option that if the Supreme Court fails to ask for a preliminary reference, one of the parties concerned may try to claim damages in a lower national court (Kobler), which in turn could still ask for a preliminary ruling.  This seems though unlikely, in particular, if the two parties stick to the same interpretation of Article 50 in the Supreme Court.

It is not excluded that the question of revocability of withdrawal might still turn up in another ‘Brexit case’ before a national court, which could ask for a preliminary reference on (several aspects of) Article 50.  EU citizens in the UK, for instance, or British citizens elsewhere in the EU may go to court to contest losing their citizenship rights.   However, as far as the question of revocability is concerned, lower UK courts may be extra reluctant to ask for a preliminary reference if, on top of the hostile political climate, the Supreme Court would not have considered such a referral necessary in the Miller case.  Such reluctance, though, would not apply for cases brought (potentially by British citizens) in national courts elsewhere in the EU. Yet, in any case, time is ticking for such cases still to emerge within the timespan that revocability of withdrawal remains politically relevant.

One can wonder whether, if not through preliminary reference, the CJEU may come to pronounce itself on Article 50 via the annulment procedure (Article 263 TFEU). This seems very hypothetical, particularly in the current political constellation. If the UK were to change its mind, the revocation of the withdrawal could be accepted (implicitly or explicitly), or refused (explicitly) by the EU. A refusal by the EU might in theory open the possibility for the UK to go to the CJEU to ask an annulment of that decision arguing that Article 50 permits a Member State to revoke its withdrawal.  Yet, if the current political constellation makes it unlikely that the UK would want to revoke its withdrawal, it seems pure science fiction to imagine that it would want to insist on such an attempt via the CJEU if the EU has made it politically clear that it considers itself better off with the UK out of the EU. Vice versa, one may wonder whether a decision of the EU to accept the UK’s desire to revoke its withdrawal could lead to one or several Member States to contest this via an annulment procedure. Given the silence of Article 50 on revocability it is not clear which act the EU would use to adopt a decision on accepting revocation. The most likely option would be by a resolution or decision of the European Council.  Such a resolution or decision would constitute a reviewable act for annulment. However, given that Article 50(3) requires unanimity of the European Council to extend the period of negotiation of an exit agreement beyond two years, it is likely that other European Council decisions explicitly amending the ‘natural course’ of Article 50 would also be taken by unanimity; implying that no Member State would contest the decision in Court. In the unlikely case that such a decision was not taken by unanimity, it remains politically very unlikely that one or more Member States (and even less any  EU Institution) would go to Court to openly contest a decision taken by the majority of  Member States to keep the UK in the EU.   The situation is less clear if the EU accepts the revocation in a more informal manner by not explicitly adopting any act giving effect to its decision. In that case there would be no reviewable act for annulment.

Hence, if the Supreme Court in Miller does not ask for a preliminary reference, the chance that the CJEU can still pronounce itself on the issue of revocability in due course is low.  However, even in that case it is important to reflect on what answer the CJEU is likely to give on the question of revocability of withdrawal.   If the UK would wish to revoke its withdrawal, the possibility to do so or not will ultimately depend on how the EU interprets Article 50.  In the absence of a CJEU ruling, it will be the legal services of the EU Institutions (and most likely of the Council in particular) which will provide that interpretation.  It is this ‘EU interpretation’ of Article 50, provided by the CJEU or the legal services of the EU institutions,  which will ultimately decide on whether a decision to revoke withdrawal can be accepted by the rest of the EU.   Yet, one should add that this is only likely to happen if the Supreme Court accepts in the Miller case that withdrawal is revocable.  If the Supreme Court declares that Article 50 is not revocable, it is rather difficult to imagine that the political actors would still ask for revocation of withdrawal.

The law to interpret Article 50: EU rather than national or international law

That Article 50 TEU states that ‘any Member State may decide to withdraw from the Union in accordance with its own constitutional requirements’ does not imply that the withdrawal process is just an issue of national law.   Article 50 describes the withdrawal process as a negotiated process between the EU and the departing Member State.  The CJEU (and EU legal services) will explain that process with reference to the entirety of EU law and the context of Article 50 and objectives of the Treaties in particular.

Similarly, the CJEU’s stress on the autonomy of the EU legal order means that the Court and EU legal services will prefer such a teleological interpretation rather than handing over the issue to the application of international law.

It is worth reminding here that while a withdrawing Member State is negotiating its exit  under Article 50 it still is a Member State and thus fully subject to EU law.  It is true that under Article 50, the withdrawing Member State has a special status as far as the negotiation process is concerned since it is not allowed to participate in the

European Council and the Council when its withdrawal is discussed.   However, this does not turn the withdrawing state into a non-Member, not subject to EU law.  Neither does it imply that the negotiation becomes one of international law instead of EU law.

Equally, one should not be misled by the reference in Article 50 to Article 218(3) TFEU.  Article 218 TFEU sets out the procedure for negotiations between the Union and third countries or international organisations.   However, Article 50TEU only refers to the third paragraph of Article 218 TFEU, which states:

The Commission, […] , shall submit recommendations to the Council, which shall adopt a decision authorising the opening of negotiations and, depending on the subject of the agreement envisaged, nominating the Union negotiator or the head of the Union’s negotiating team.

The reference to Article 218(3) does not turn the withdrawal negotiation into a negotiation with a third country or international organisation. It simply provides a ‘shortcut’ for one procedural step of the specific procedure for the withdrawal of a Member State, that is otherwise entirely set out in Article 50.

This does not imply that international law may have no role at all in the interpretation of Article 50.  The CJEU has confirmed that the EU must respect international law in the exercise of its power, meaning respecting the international agreements to which it has signed up, but also complying with the rules of customary international law (Poulsen, Racke, Firma Brita).  However, in the case of Article 50TEU we are not dealing with the external action of the EU, and most importantly, what is at stake is not an issue of interpretation of an international agreement signed by the EU with a third State but one of interpretation of the EU Treaties itself.    In that case, the CJEU is far more reluctant to refer to international law, as the Court stresses the autonomy of the EU legal order.  In Kadi it stated that international law (in that case the UN Charter) has primacy over secondary EU acts but not over the EU Treaties. In particular, international agreements can neither “affect the allocation of powers fixed by the Treaties Kadi, para 282), nor can they “have the effect of prejudicing the constitutional principles of the EC Treaty” (Kadi, para. 285).

Hence, the CJEU’s interpretation of Article 50 will be primarily inspired by the EU Treaties and will above all ensure the balances set out therein.

In relation to the question of revocability of withdrawal, it has been suggested that Article 68 of the Vienna Convention of the Law of Treaties (VCLT) should be applied, which states that a notice to withdraw ‘pursuant to the provisions of a treaty’ (art.67(2)) may be revoked at any time.  However, the EU is not a signatory to the VCLT.   It applies the VCLT only to the extent that the stipulations of the Convention express general international customary law (Firma Brita).

Moreover, in the Firma Brita case, the CJEU applied the VCLT via customary international law to its external action, and not to the interpretation of the EU Treaties.    While the EU Treaties are indeed treaties under international law, which might suggest the application of the VCLT (at least if expressing customary international law), the CJEU’s stress on the autonomy of the EU legal order implies it will be very reluctant to refer to international law (and VCLT) for their interpretation.

In the case at hand, it has not been established that the relevant article on revocability of withdrawal (Article 68 VCLT) is a norm of international customary law.  Moreover, the VCLT has not been drafted having in mind its application to international organisations, as it originally only applied to States as subjects of international law; and therefore does not provide any particular procedure for withdrawal (and revocation thereof) from an international organisation.  Hence, the CJEU would be faced on the one hand, with an article of the VCLT (to which the EU is no signatory), which is not established as a norm of customary international law,   and part of a treaty originally not assumed to apply to international organisations, and on the other hand, a comparatively detailed procedure set out within the own EU legal order to regulate withdrawal of a Member State.  The CJEU will be strongly inclined to refer only to the latter and any interpretation of Article 50 will be informed by the Treaties rather than reference to an uncertain international norm.

One should not forget that the EU is not a standard international organisation, due to its organisational features, broad scope of competences and ability to create directly enforceable rights for its citizens.   The withdrawal of a Member State from the EU is therefore not equivalent to the withdrawal of a country from an international agreement creating rights and obligations between States.

It is the Court’s duty to ensure the integrity of EU law, and the realisation of the objectives of the Treaties.  This includes safeguarding the rights of European citizens and ensuring the balance of powers set out in the Treaties (between the EU Institutions, and between the EU and the Member States).   As I explain below, Article 50 sets out a particular balance of power, and it is this teleological and EU interest interpretation of Article 50 that will prevail, rather than an interpretation of primary EU law with reference to an uncertain norm of customary international law.

The EU interest reading of Article 50: more exit-hostile than exit-friendly

The exit-friendly interpretation

The exit-friendly interpretation of Article 50 sees the article mainly as the confirmation of a unilateral right of a Member State to withdraw combined with a procedure that facilitates a negotiated outcome.  According to Hillion and Syrpis, Article 50 has a specific function in relation to the integration process in that it ‘bolsters the normative basis for a negotiated withdrawal’ and ‘points towards a strong post-withdrawal engagement by the Union with the former Member State’.  In that friendly context, a withdrawing Member State would also have the option to revoke unilaterally its withdrawal notification. Several arguments have been given that would support such revocability:

1. Article 50 does not impede such revocability explicitly

2. The history of the drafting process of Article 50 shows that proposed amendments to condition and limit the right of a Member State to withdraw were not withheld (although there is no trace that anything was said about the issue of revocability).

3. One could argue that the decision to revoke withdrawal notification can be taken on the basis of Article 50(1) in accordance with national constitutional requirements. However, the link between the first and second paragraph of Article 50 clearly suggests that the decision to withdraw from the Union ‘in accordance with its own constitutional requirements’ is taken before notification of withdrawal to the EU. Once such notification is given the process becomes one under EU law.  I therefore disagree with Paul Craig who suggests that going ahead with withdrawal would be illegal in case revocation is given since such revocation would, in respect of national law, make invalid the notification of withdrawal (Paul Craig). Rather, the original notification stands in respect of national constitutional requirements, while the revocation is invalid under EU law.

4. Finally, it is said that non-revocability could lead to untenable situations in case a withdrawing country changes its mind, faced, for instance with dramatic negative economic consequences of exit, and/or as a result of elections or a new referendum. Moreover, if that were the case, the country would then have to go through accession negotiation, including ratification by all Member States.   This situation would indeed be far from desirable, but this does not turn revocability into the ‘natural textual reading’ of Article 50 as Craig seems to suggest.    Rather, one needs to assess this desirability within a teleological interpretation.  Can it be argued in light of the objectives of the Treaties that such effect cannot be intended in Article 50?   I agree with Craig that in the light of the objectives of European integration it would make little sense that a withdrawal could not be stopped if the withdrawing State and EU desire so.  However, I do not agree that implies the withdrawing State can unilaterally decide to revoke its withdrawal.  The spirit of the article suggests otherwise. Article 50 is not entirely exit-friendly and written in such a way that the withdrawing State would be able to steer the withdrawal process as it believes is most convenient.

The exit-hostile interpretation

In the Miller case, non-revocability is taken at face value.  Digging deeper, the key argument for non-revocability lies in those aspects of the Article which make it a hostile rather than friendly environment for withdrawal.

Nick Barber, Tom Hickman and Jeff King have argued very rightly that Article 50 has the intention to turn the bargaining power during the negotiations in favour of the remaining EU States.  Once a Member State has decided its withdrawal and notified the European Council, the 2 year period of paragraph 3 is triggered, which leads to either a negotiated agreement, or, failing that, automatic exit, unless the European Council, by unanimity decides to extend the negotiation period.  For the withdrawing Member State there is much more at stake than for the other Member States. Failing to reach an agreement means that the exiting country will fall back on standard WTO rules, while the cost for the remaining Member States of an exit without agreement will be shared between 27 Member States. The threat of an automatic non-negotiated exit therefore favours the bargaining position of the rest of the EU.

If one accepts revocability on the initiative of the withdrawing State, the Member State could abuse Article 50 and overcome the two year limitation by starting negotiations and breaking them off if inconvenient, while given a new notice for withdrawal later.  The problem though is not simply one of abuse.   Accepting the unilateral revocability of the withdrawal procedure undermines the balance of power  intended by Article 50.   Even if a withdrawing State honestly wants to revoke its withdrawal and has no intention to restart the procedure, accepting revocability takes away the threat of automatic exit that Article 50 clearly intends.

Barber, Hickman and King concluded on the non-revocability of withdrawal (by focusing on the power balance in Article 50), in order to stress (pre-Miller case) the importance of reflecting on the national constitutional requirements prior to triggering Article 50.   By turning their focus on the national level, they did not fully engage with the broader EU law context of Article 50.    The EU interpretation of Article 50 sits in fact between the exit-friendly and exit-hostile interpretation.

The EU (interest) interpretation

As argued above, the CJEU (and the EU legal services) will give an EU interpretation of Article 50, which can also be understood as an ‘EU interest interpretation’.  Article 50 will be interpreted in the light of the objectives of the Treaties and the checks and balances set out therein.

From the drafting history of Article 50 it is clear that the article was introduced (first in the EU Constitutional Treaty, and subsequently by the Lisbon Treaty) with the aim to indicate clearly that a Member State has a right to withdraw from the Union and to provide a procedure that leads preferably to a negotiated outcome.  However, this exit-friendly intention is combined with a more exit-hostile procedure.  The article makes clear that the withdrawing country shall no longer take part in the deliberations and decisions of the Council and European Council on this topic.  It is the exiting country against the remaining countries gathered in the Council and European Council.  Most importantly, the ‘guillotine clause’, threatening automatic exit if no agreement is reached by the deadline, is clearly intended to favour the remaining Member States and the defence of the general interest of the Union. The article is clearly written in such a way that while a negotiated solution is the preferred outcome, it is up to the remaining Member States to pull the strings and decide which solution is best for the EU once the procedure has been triggered at EU level.

Hence, once a Member State has decided internally (respecting its own constitutional requirements) that it intends to withdraw, its notification of withdrawal to the European Council triggers Article 50 and the checks and balances set out therein to ensure the general interest of the EU.  There is no unilateral way back once the issue has been put on the European negotiation table.  At the same time, it would be entirely against the spirit of the Treaties and of Article 50, if the EU could not stop the withdrawing process if the outgoing State has changed its mind and wants to remain.  Put differently, the power balance in Article 50 and the objective of the Treaties imply that while the withdrawing Member State has no right to stop the withdrawal process on its own, the other Member States cannot be obliged to continue with the withdrawal if both they and the withdrawing Member State no longer desire to do so.

A withdrawing Member State can ask to revoke its withdrawal but it is up to the other Member States to decide whether they accept that request.

The question then raises which procedural threshold is required to adopt such a decision.

The third paragraph of Article 50 requires that a decision on extending the negotiation period beyond two years has to be adopted by the European Council by unanimity.   Given the general spirit of the Treaties and Article 50 to ensure the general interest of the EU, one could deduce that an even more important decision on the fate of Article 50 negotiation than extending the deadline, namely, a decision to revoke the procedure, has to be approved equally by unanimity in the European Council.  One can argue that the natural course of Article 50 leads either to agreement or automatic exit after two years, and that any decision changing that natural course, such as extending the deadline or accepting revocation requires unanimity.

However, there is an alternative and in my view preferable interpretation in which the decision to accept revocation would require a qualified majority vote (QMV) in the Council and consent in the European Parliament.  This is the requirement for an exit agreement.  The power balance in Article 50 is created by the ‘threat’ that in an absence of an exit agreement, there will be an automatic exit after two year.

Hence, the natural course of Article 50 is either an exit without agreement, or an agreement based on qualified majority in the Council and consent by the EP.

One can argue that such an agreement could be the exit deal but could equally be an agreement for the country to remain.   This reasoning respects the power balance created in Article 50, and allows the remaining Member States to accept a request to remain, in light of the broader objectives and interests of the Treaties.  I believe this interpretation, i.e. requiring QMV + consent, is preferable over the interpretation that requires unanimity.   What is at stake here is a final decision over the fate of a negotiation; it is thus more similar to an exit agreement than to a mere decision to offer extension of the negotiation period.

The QMV and EP consent might ironically enough be easier to reach than unanimity in the European Council.  It should be relatively easy to find a simple majority in the normally pro-integration oriented EP, whereas unanimity in the European Council implies one Member State with a grudge would be enough to lead to automatic exit.

Implications of the EU interest interpretation for the Miller case

The consequence of the EU interest interpretation is that once Article 50 is triggered, the decision to revoke is out of the control of Parliament.  Does that imply that statutory rights are affected without Parliament having a say on it? Although triggering itself does not immediately lead to the loss of those rights, it sets in action a process in which Parliament is not guaranteed it can prevent  the loss of those rights.  As mentioned above, and convincingly argued by King and Barber, the rights at stake are statutory rights.   Politically it is unlikely that the remaining Member States would not accept the UK’s request to revoke withdrawal and thus to remain in the Union, although that risk is higher if the requirement is unanimity in the European Council rather than QMV+ EP consent. In any case, there is no legal guarantee that Parliament can avoid the loss of statutory rights once Article 50 has been triggered.

Hence, the implication of the EU (interest) interpretation of Article 50 for the Miller case is that Parliament should be given a say prior to triggering Article 50, given that subsequently it is not in a position to avoid loss of statutory rights.  However, if the UK, by way of the Government or Parliament subsequently intends to revoke its withdrawal notification, this might still be accepted by the EU.

The author would like to thank Sara Drake and Keith Syrett for comments on an earlier draft. The usual disclaimer applies.

Stijn Smismans is Professor of Law and Director of the Centre for European Law and Governance, School of Law and Politics, Cardiff University.

(Suggested citation: S. Smismans, ‘About the Revocability of Withdrawal: Why the EU (Law) Interpretation of Article 50 Matters’, U.K. Const. L. Blog (29th Nov 2016) (available at

26 comments on “Stijn Smismans: About the Revocability of Withdrawal: Why the EU (Law) Interpretation of Article 50 Matters

  1. Chris V
    November 29, 2016

    Very interesting.

    If the Crown argues it has power to trigger article 50 on the basis that Parliament can revoke the withdrawal notice, and so no rights will be lost except with the approval of Parliament, is it not arguable that notice is conditional. Is conditional notice in compliance with the requirement that it meets the member states “own constitutional requirements”.

    This situation is different from one there one administration serves notice and then, subsequently following an election, a different administration wishes to withdraw. How, if at all, would the “good faith” of the member state matter?

    Politically I see no prospect of the crown at the appeal trying to use revocability since this would demonstrate lack of good faith and raise doubts about notice meeting our own constitutional requirements, once we establish what they are! Then the question remains: would the SC feel obliged to ask the CJEU. Would this not turn on whether they considered revocability as essential to a determination? If they had concluded that the crown had no prerogative, regardless of revocability, then the question should not need to asked.

    I raise these as questions, not firm opinions.

    • Chris V
      November 29, 2016

      To add to my earlier comments. A notice under article 50 by means of royal prerogative premissed on the power of Parliament to reject the resulting negotiations and thereby to withdraw the notice as not being valid strikes me as not a notice served within the proper interpretation of article 50. That is to say it is not intended as a final notice but simply as a negotiating strategy hoping for the best. In those circumstances, revocation would be a decision for member states.

      • paulobrennan
        November 29, 2016

        Well said, Chris V. If Article 50 can be unilaterally revoked (which I assume is the case for the purpose of the rest of this paragraph), then any decision by the Government could be overruled by Parliament – the decision to leave the EU would be conditional on Parliament not overruling the Government within the two year notice period. In such circumstances it can only be known that the UK has actually decided to leave the UK once the condition is satisfied, on the expiry of the two year period. One might say that in these circumstances the decision is conditional , but it is more accurate to say that no decision has been taken at all. The decision would only occur as a consequence of Parliament forbearing to countermand the Government’s decision over the two year period – Parliament’s failure to act would be an integral element of the decision. As the notice of withdrawal can only be served after the decision to leave is made and the decision to leave cannot be completely made until after a two year notice period has expired, the notice period would never start running and so neither could the notice period come to an end: The decision could never take effect. Even if it were permissible under the terms of the UK’s constitution for the decision to leave the EU to be taken by the Government under Royal Prerogative, the Government’s decision would not constitute a decision within the meaning of Article 50(1) which is capable of triggering the obligation to serve notice under Article 50(2). The Government might have decided to leave the EU or serve an Article 50 notice (which are in fact different decisions), but unless and until the decision to leave the EU is made by the supreme authority under the United Kingdom’s constitution, Parliament, the UK will not in fact have made a decision to leave the EU. As argued by the Respondents in their statement of case, a step equivalent to the ratification of the UK’s accession to the EEC, that is an Act of Parliament, must be taken in order for the decision to leave to be sufficient for the purposes of Article 50.

        Consequently, if the Supreme Court is contemplating allowing the Government’s appeal (which, for what it’s worth, is highly unlikely in my view), it should first refer the following questions to the CJEU: Is an Article 50(2) notice revocable and, if it is revocable, would a decision by the Government that can be overruled by Parliament constitute a decision by the UK within the meaning of Article 50(1)? A failure to do so would leave the UK exposed after any service of a notice purportedly under Article 50(2) on the basis of a mere decision under the Crown Prerogative to further litigation under the terms of the EU Treaties over the question of whether or not the notice is actually valid. A half-competent Government would take all steps necessary to ensure that any such uncertainty could not arise.

        Needless to say there is a difference between a decision being made by a body such as the Government which can be overruled by Parliament, and a decision made by Parliament which Parliament may subsequently decide to change of its own accord. In the latter case the revocability of Article 50 is not an issue, since the ability of Parliament to change its own decision does vitiate the validity or completeness of the original decision. In practical terms too a change of mind by Parliament is, appropriately in the circumstances, a far more onerous proposition than a change of mind by Government.

  2. Mike Fearon
    November 29, 2016

    Whilst not wishing to detract in any way from the sound reasoning and careful argument of this contribution, I do sometimes wonder whether anyone has read the excellent (and I mean that literally) piece by Phillip Allott.

    To quote:

    “Article 50 says that ‘a Member State may decide…’ and ‘[that] Member State shall notify…’ (italics added). The first is a legal power. The second is a legal obligation. It is normal international practice for the executive branch of the national government to express externally a decision to exercise a power or to carry out a duty under a treaty. The process for enabling the executive branch to do so is a matter of national constitutional law ”

    “A notification to the European Council of an intention to withdraw is a political act, subject to the constitutional conventions mentioned in (f) above. It does not ‘trigger’ anything. It fulfils a UK treaty obligation, enabling the European Council to begin the formal process of drawing up negotiating guidelines.”

    If the UK has made a decision to withdraw from the EU, it is obliged to notify the European Council of its intention to withdraw. It is academic, if the academics will pardon that expression, whether the notification is authorised by Parliament in a formal process. Whilst it is interesting to examine the legal possibilities of a future change of mind, the U.K. will be (deservedly) a laughing stock throughout Europe if it asks the CJEU whether it will be able to change its mind about a decision it has made (assuming for the moment that it has) before it can decide whether it has made it. I despair.

    The political problem is greater than the legal issues here. A state which decides to withdraw does so, at least in theory, because it believes that the benefits of withdrawal outweigh the benefits of remaining a member of the EU. The political difficulty arises because a large proportion of the population, perhaps an overwhelming proportion in relation to those with power and influence, are not of that view. Hence the concerns illustrated quite innocently in this contribution about the balance of power lying with the remaining states in the negotiation, and the frequently highlighted “cliff edge” of withdrawal without an agreement that nothing will change very much.

    The Divisional Court, was mistaken, in my opinion, to take the view that “nothing turns” on the issue of whether the case is about whether the decision has been made in accordance with the constitutional requirements of the U.K., or whether the Crown may notify the intention of withdrawal without Parliamentary authorisation. If the decision has been made, and that has been accepted by the litigants (ms Miller in public says it has) the case is frivolous, and much of this debate is sterile. I do not argue that the decision has been made or not been made, but this is the essence of the political dilemma, which the Supreme Court should not be called upon to resolve.

    • Chris V
      November 29, 2016

      I suppose the argument could be characterised a “who” not “whether” but that misses an important constitutional point which is that the crown cannot exercise it’s prerogative in such an important matter. The substantive issue, IMO, is the involvement of Parliament in debate before notification and involvement during negotiations as discussed in other posts on this site. The argument has been pitched as the crown as the people’s champion against Parliament. This strikes me as bizarre as well get another example of an over strong executive seeking to bypass Parliament.

    • paulobrennan
      November 30, 2016

      I agree, MIke Fearon, that the distinction between the decision to leave the EU and the decision to serve the Article 50 has been rather neglected in this case and the commentary surrounding it. I disagree with your conclusion, however. The decision as to whether or not to leave is not, as you appear to suggest, merely a political decision taking place remote from the UK’s legal and constitutional framework, it is a decision with the most profound legal and constitutional consequences. It is the decision to leave which triggers the operation of Article 50, not the notice served under Article 50(2) once Article 50 is activated.

      At the first instance hearing the decision was described by the Attorney General as a complex affair taking the form of various statements made by David Cameron, Theresa May and others both before and after the referendum. Incredibly, he contended that these steps, which include a promise by Mrs May in announcing her candidacy for the position of Prime Minster (sic) and, after she became Prime Minister, repeated statements by Mrs May and other ministers in unspecified fora that “the Government will deliver the departure of the United Kingdom from the European Union” constitute a decision taken by the Government in accordance with the provisions of Article 50(1). It is extraordinary that the Government should seek to claim that a decision of this magnitude taken in such a manner satisfies the constitutional requirements of the UK, when, as Patrick Green QC put it at the hearing, the decision comes to the Court “in a manner which the decision of a minister on a licence to make pipes would not”, with not even a copy of the decision available, let alone any other information as to how the decision was made or by whom. As Green goes on to observe, “the Government fails to distinguish between a Government intention to bring forward a bill and lay it before Parliament and achieve a policy objective on the one hand, and the Government having the power to take the decision itself without Parliament’s involvement at all.” The Government may have decided as a matter of policy to implement the referendum result, of that there is little doubt, but, as in the case of any other manifesto commitment or policy decision, the Government has to secure the passage of the requisite legislation through Parliament in order to deliver its policy. It is for Parliament to decide whether or not the Government is to be permitted to do so.

      You rightly say that if the decision is made (validly), the Government is obliged to serve notice under Article 50(2). Although there is no strict timeline within which to serve the notice, it is implicit that the notice must be served within a reasonable period of time. One would imagine that 6 months, let alone 9, is far more time than might reasonably be required to send a notice. In arguing the decision to leave the UK has already been taken, the Government is arguing that it is in wilful breach of Article 50(2). A breach of the EU Treaties by the Government is of course a breach by it of the law as provided for by ECA section 2(1). I am not a litigator and I wonder whether it is appropriate or permissible for the Government to be praying its own breach of the law in aid of its case.

      • Rodger Harris
        November 30, 2016

        I have written before on the distinction and congruently interdependent functions of form and substance.
        There is as yet no “decision” that can be pinned down within the British Constitutional system with any precision, saving by reference to the unrepealed Government Prerogative. To argue that there has to be something different owing to the nature of the Treaty is begging the question.
        On the basis of our tectonically unwrit and therefore unread, suppositions as to the decision, by the “unwashed” EU Council of Ministers relying upon the Referendum result are unfounded, and will remain so until a Notice is actually filed by HMG. without some assistance from form, convention or a procedure there is as yet no article 50 decision to be notified and therefore served. Herr Schaüble has distinguished himself as defining our domestic constitutional law as to decision making solely by reference to the will of the British people.
        The irony of such an arrogation of otherwise unconferred Kompetenz will not be lost on this forum.

      • Mike Fearon
        November 30, 2016

        Personally, paulobrennan I tend to agree with you that some form of parliamentary confirmation is required. The government’s argument as I understand it, is that Parliament resolved to make the decision by entrusting it to the electorate, and constitutional requirements are therefore met. I am not sure whether that argument will be put to the SC, and it may or not be successful if it is.

        I have suggested elsewhere on this blog that Parliamentary authorisation does not require a Bill or an Act, but could in practice be pursued through an early day motion. It is a speedier process, and the result in practice is going to be the same. Parliament will confirm the government’s interpretation. In reality the posturing is going to evaporate if or when Parliament has to decide. The legal debate is fascinating and educational, but ultimately irrelevant.

        The contribution by Chris V is useful. Decide means Decide just the same as Brexit means Brexit. Parliament, and the litigants might as well face up to it now, the alternatives are a negotiated agreement, which Parliament is content to ratify, or no agreement. It is not practical for Parliament to negotiate the agreement, although it can and will have an input to the process through established mechanisms.

        The government would be well advised to say quite simply that its negotiating strategy is to present the UK as a strong independent state which is confident of its ability to trade with the world. It does not need to be a supplicant, it is an important trading partner to the remaining countries. Dragging this through the Supreme Court, even worse the CJEU, does not give the impression of confidence and strength that a negotiating team needs. This is not the fault of the courts or the lawyers, the blame lies firmly at the collective door of Parliament.

      • paulobrennan
        December 2, 2016

        (In reply to Mike Fearon’s second comment.).

        I don’t see how Parliament can be said to be at fault, unless you are saying that it is at fault in failing to legislate for a legally binding outcome to the referendum. In any case the fault is without doubt that of David Cameron and the 80 or so Conservative MPs who campaigned for Brexit, for their failure to make adequate preparations for one of the two possible outcomes. These failings have been compounded by Mrs May’s subsequent inept leadership.

        It would appear that the Government has been badly served by its legal advisers, though my own theory is (which is nothing more than supposition) is that Mrs May has singled out the Attorney General as the fall guy, by insisting that he leads the Government’s case personally. I find it hard to believe that the Government’s legal team gives them anything other than an outside chance of success on appeal.

        Throughout, the Government has made an elementary error in failing to distinguish between making a decision to adopt a particular policy objective and taking the steps necessary to realise that objective. The Government’s failure to defend the members of the Divisional Court against unfounded allegations of bias by the Daily Mail and others shows that it is interested more in protecting its own reputation than upholding the rule of law. It’s failure failure to stand up for the UK’s independent judiciary has not gone unremarked on the Continent. In the circumstances one might expect, and I certainly hope, that the Supreme Court’s ruling will be critical indeed of the Government’s conduct.

        I think though I am veering into political commentary and away from legal analysis, so I will leave it there.

  3. Rodger Harris
    November 29, 2016

    There appears to have been no research quoted on the mechanisms adopted to place article 50 in the current Treaty during the negotiations.
    Without my spending time on this point which I can ill afford, Stijn might wish to refer to those, first, as the point raised by Arel Sari as to the withdrawing state being subject to International as opposed to the EU International cocktail proposed is valid, and has only been answered by a bald EU style assertion.
    Only certain competences have been expressly conferred on the EU. Withdrawal being a fundamental International right at an international law level, the United Kingdom may not have surrendered its unilateral competence to withdraw in the manner in which Stijn asserts.
    A little more argued and researched comment might be usefully made, particularly as both Lord Kerr, part of the negotiating team and also The President of the Council have stated that an article 50 notice served on the Council is revocable,largely for the reasons which Sarel has pointed out.

  4. paulobrennan
    November 30, 2016

    FFS The “not” still got lost! It should read: “In the latter case the revocability of Article 50 is not an issue, since the ability of Parliament to change its own decision does NOT vitiate the validity or completeness of the original decision”. I hope readers think my legal observations are slightly less comical than my typing.

  5. Chris V
    December 2, 2016

    Re Mike Fearon’s comment: The contribution by Chris V is useful. Decide means Decide just the same as Brexit means Brexit.

    I fear neither is clear as to meaning in any useful way. My view is that Parliament must decide and that royal prerogative is not available. A decision and what Brexit means will then become clear. No point in notification if the strategy is not agreed. How one expresses the strategy seems to me to call for more than Mike suggests. We can all agree that the best deal for Britain is an unarguable point, but sadly begs the question.

    These points take us beyond legal argument and the scope of this blog.

    • Mike Fearon
      December 5, 2016

      Perhaps I should make it clearer that my view is not that the legal debate is irrelevant to the process, which is of course the focus of this debate in the blog. It is that the debate will prove to be irrelevant to the outcome, which may be viewed as outside the scope of the blog. I hope the point may add contextually to the debate, and that participants will not be offended by its introduction.

  6. Pingback: Robert Craig: Miller: An Index of Reports and Commentary | UK Constitutional Law Association

  7. Peter
    March 29, 2017

    We are left with would it be lawful for the UK and the EU to agree to terminate the Article 50 negotiations and agree that the notification is deemed withdrawn. The answer seems to be yes. That being so politics and public opinion througout Europe remains in play and there is much uncertainty stll.

  8. Peter
    March 29, 2017

    And a final observation, as far as final appellate Courts go, and in this old lawyers opinion, the CJEU is one of the most politicised in the world. Political sensitivity, as Mr Smismans rather gently describes it, is rarely viewed in jurisprudential theory or practice as a sound or healthy virtue in any court, least of all a court of last resort. One would hope as the EU grows with or without the UK, at look easy that will change. A court that listens to the siren voices of today’s mob will not command respect for long nor will the institutions it serves thrive.

  9. Pingback: Shadow-boxing against a “no deal” Brexit is counter-productive

  10. Pingback: It's in the EU’s interest to press ahead with Brexit negotiations | Wikipedia Editors

  11. Pingback: It's in the EU’s interest to press ahead with Brexit negotiations -

  12. Pingback: It's in the EU's interest to press ahead with Brexit negotiations - GoterNews

  13. Pingback: It’s in the EU’s interest to press ahead with Brexit negotiations | PolicyBristol Hub

  14. 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.

  15. Pingback: Can the United Kingdom unilaterally revoke its Article 50 notification to withdraw from the EU? Wightman v Secretary of State for DexEU [2018] CSIH 62 | EuroReads

  16. 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

  17. Pingback: “Does the principle of the autonomy of the EU legal order allow for a Member State to revoke the notification of its withdrawal from the EU?” | EuroReads

  18. Pingback: Can an Art. 50 TEU withdrawal notice be revoked? How Advocate General Bordona offered a legal Trojan horse to Union law | 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: