Brexit II? The legal issues of revoking the notification to leave the EU but then notifying to leave again

Ronan McCrea, Professor of Constitutional and European Law, University College London

The CJEU ruling in the Wightman case (discussed here) has provided important additional information in relation to the Brexit process. The Court accepted arguments first made by Piet Eeckhout and Eleni Frantziou that notification of an intention to withdraw from the EU could be withdrawn unilaterally by the UK without the need for permission of the remaining EU 27.

This is important and may strengthen the hand of those arguing that the UK government should hold a second referendum with remaining the EU as an option on the ballot paper. However, the ruling of the Court of Justice is disappointing. It fails to provide meaningful guidance in relation to key questions in relation to revocation of Article 50.

The test set out by the Court for unilateral revocation is as follows:

-          revocation must take place before the Withdrawal Agreement comes into force (or if there is no such agreement, within the two-year period from notification set down in Article 50),

-          the revocation must take the form of notification in writing to the European Council following a to revoke that was taken in accordance with the constitutional requirements of the Member State in question,

-          revocation must be ‘unequivocal and unconditional’ which means ‘the purpose of that revocation is to confirm the EU membership of the Member State concerned’.

Notably absent from this test is the condition suggested by the Advocate General that the revocation must be sincere and done in good faith which as Oliver Garner noted could be very difficult to apply in practice.

This is useful information from the Court. We now know that, if the UK authorities change their mind on Brexit, proceed to reverse the decision to leave in accordance with the requirements of the British constitutional order and give written notification of this decision to the European Council before March (or later, in the event that Brexit Day is extended by joint agreement of the UK and the EU27), the UK can remain in the EU on its existing terms.

This is important as there had been suggestions that if the UK wanted to remain in the EU in this way, it may have needed the consent of other Member States who may have taken advantage of the UK’s weak bargaining position to remove benefits, such as the rebate on payments to the Union, that the UK currently enjoys.

It is not entirely certain whether a valid revocation would require a referendum under EU law. The Court’s ruling refers three times to the decision to revoke taking place through ‘a democratic process’ but in the operative part of the ruling the reference to democratic processes is dropped and only ‘constitutional requirements’ are referred to.

Some may argue that it would be undemocratic to reverse the Brexit decision without a referendum. However, representative democracy is a valid form of democratic decision-making. In addition, under the British constitutional system, Parliament remains the supreme legislative authority so if Parliament decided to reverse Brexit, such a decision would be in accordance with the constitutional requirements of the UK and is unlikely to be second-guessed by the CJEU (however unlikely it is in political terms, that this situation would arise).

Either way, thanks to the CJEU ruling we now know that if all of these hurdles are overcome and the UK manages to take a definitive decision to reverse Brexit before the end of March 2019, it cannot be stopped from remaining in the EU on its existing membership terms.

While that is somewhat useful information, it is not particularly useful because it is unlikely in the extreme that the UK will manage to take a definitive decision to reverse Brexit in the next couple of months.

For one thing, the UCL Constitution Unit has shown that a new referendum will need at least 22 weeks to organize. This means that even if the UK Government decides it wants to hold a referendum to reverse Brexit it is likely to need to obtain the agreement of the other 27 Member States to extend the two-year period under Article 50 to give it time to organize this vote. Whether the Member States would agree to such an extension unless the UK government promised to campaign for remain is very uncertain.

Perhaps more seriously, the Court’s judgment leaves us largely in the dark as to what would happen in the not unlikely scenario that the UK seeks to revoke without having definitively decided to remain in the EU. Indeed, it was this issue of the potential for a right of unilateral revocation to be abused that was the core of the arguments made by the Commission and Council in Wightman. And on this issue, the Court has provided little useful guidance.

Consider the following scenarios (none less likely than the UK definitively deciding to reverse Brexit before March 2019):

1)      The UK government seeks to revoke Article 50 in early March 2019 but makes it clear that this revocation will need subsequently to be ratified by a referendum of the British people.

2)      The UK government revokes Article 50 and declares that the UK will be staying in the Union. However, three months later, it declares that due to changed political circumstances, it is triggering Article 50 once more.

3)      The UK revokes Article 50 saying that it is bringing the current exit procedure to a close but makes it clear that it will retrigger in the near future to begin a new process (in which the UK is better prepared and has a clearer idea of what it wants).

In all three scenarios, we are faced with a situation where the revocation is legally unequivocal but is (or is later revealed to be) politically equivocal. This would seem to be an abuse of the right of revocation and raises the prospect of the UK being allowed to circumvent the strict two-year time limit set out in Article 50. As it goes against the requirement set down by the Court in Wightman that revocation be unequivocal it should, in theory, be capable of being challenged. But working out how such a challenge would work is very difficult.

As the act of revocation is unilateral, so there is no act of acceptance on the part of the Council that could be challenged. Perhaps a Member State or EU institution could apply to the General Court to annul the revocation by the UK or (given that the giving of a notice of revocation is a sovereign act) to get a declaration that the act of revocation did not produce consequences within the EU legal order on grounds that it failed to satisfy the Wightman criteria. If the courts did not intervene then an equivocal revocation would stand, something that seems to go against in the Wightman judgment. It would also seem to defeat the temporal limits on the withdrawal process enshrined in Article 50.

If the EU courts did intervene and annul the revocation then they would face a nightmare in seeking to give effect to that ruling. In scenario 2, the Court would be faced with retrospectively annulling a revocation meaning that the in theory, the UK would actually have left the Union back in March 2019 without realizing it.

Even in relation to scenarios 1 and 3, it is highly likely that any ruling from the EU courts would be given either after or very close to the end of the two-year time limit in Article 50. This means that the ruling could have the equally chaotic effect of pushing the UK out of the Union in a chaotic no-deal fashion either immediately or almost immediately. All of these outcomes would bring absolute chaos not to mention giving Brexiters the chance to blame their bete noire, the Court of Justice, for the chaos of a no deal exit.

The EU courts could, as they have done before, seek to limit the chaotic consequences of their ruling by refusing to give immediate effect to the annulment of the revocation. That would also involve the courts in a political nightmare by requiring them to decide how long the UK would have before being chaotically ejected from the Union.

By giving a unilateral right to withdraw notification without any possibility for the Council to exercise its political judgement over its sincerity or equivocal nature, the Court of Justice has left itself as the only institution with a role in limiting potential abuse of this right. This leads it into adjudicating on inherently political matters such as the likely future actions of the UK authorities or deciding how much additional time the UK should be given to avoid a chaotic exit.

Furthermore, we simply don’t know how the Court would carry out the role it has granted itself. By requiring that revocation be unequivocal and bring the process of leaving to an end, the Court has implied that it would exercise some power of review over acts of revocation but has refused to provide any guidance about how this power would be exercised. Given how much focus the Council and Commission placed on the issue of abuse of the right to revoke, the failure to give guidance on this issue is very disappointing.

The UK government had asked the Court to refuse to give a ruling on the grounds that the question referred by the Scottish court was theoretical as the UK government has given no indication that it wants to withdraw its notification. Once the Court decided to give a ruling it really should have attempted to give guidance that was as useful as possible.

Unfortunately, it ended up giving a ruling that only really applies to the extraordinarily unlikely scenario of the UK definitively renouncing Brexit before March 2019 (or a later date in the event of an extension) and that gives very little useful guidance in relation to the key issue that preoccupied the Commission and Council, the potential for misuse of revocation in order to circumvent the temporal limitations imposed by Article 50.

All of this means that in most likely scenarios, political actors find themselves acting without much idea of what the applicable law is. This increases the likelihood that we will find ourselves before the Court of Justice again with CJEU judges struggling to rule on inherently political matters such as the likely future action of the UK government. Worse, next time the Court rules the clock will have moved ever closer to midnight in Brexit terms producing an even more fraught atmosphere and even more likelihood that the Brexiters’ favourite judicial bogeyman will become a scapegoat for the chaos of a no deal Brexit.

In these circumstances the Court may well regret that it did not take the chance to rule that inherently political matters such as the assessment of the sincerity of the revocation of Article 50 were not left to political bodies such as the European Council.

Barnard & Peers: chapter 27

Photo credit: WiredUK

Devolved powers and EU withdrawal: Scotching the worst impacts of Brexit?

Alan S. Reid, Senior Lecturer in Law, Sheffield Hallam University. The author welcomes comments on the blog at a.s.reid@shu.ac.uk.

The modern concept of Devolution in the United Kingdom was devised, created and implemented in the general context and expectation of continued UK membership of the European Union. Scotland, Wales and Northern Ireland were all granted various levels of legislative and executive autonomy under a set of constitutional statutes of the UK Parliament (Scotland Act 1998 (SA 1998), Government of Wales Act 1998 (GOWA 1998)(now the Government of Wales Act 2006 (GOWA 2006)) and Northern Ireland Act 1998 (NIA 1998)). These laws created a certain level of asymmetry in the scope and range of powers exercisable by the nations of the United Kingdom, however they all shared one common feature - devolved legislative and executive powers could only be exercised in a way that is not incompatible with the UK's international obligations, particularly as regards the law of the European Union (Sections 29(2)(d) and 57(2) SA1998, Sections 58A and 108A(2)(e) GOWA 2006 and Sections 6(2)(d) and 24(1)(b) NIA 1998).

Background to Devolution

Scotland's devolution arrangements were always predicated upon a reserved model of devolution. Such a  system provides that the Scottish Parliament is free to legislate on any residual matter, other than those matters that are explicitly retained by the UK Parliament. The matters that are explicitly retained by the UK are set out in section 29(2) of the 1998 Act. This provides that the Scottish Parliament cannot pass laws that apply extraterritorially, relate to reserved matters, attempt to change the terms of certain Acts of the UK Parliament (including the European Communities Act 1972), are incompatible with the European Convention of Human Rights or EU law or are an attempt to remove the Lord Advocate as the head of criminal prosecutions and the investigation of deaths.

Notwithstanding the relative clarity of these terms, it has been judicially noted that there is of course some inevitable degree of overlap between the concepts (Martin v HM Advocate 2010 SLT 412, Lord Hope of Craighead at para 11 and Imperial Tobacco Ltd, Petitioner [2010] CSOH 134). Thus, the system of devolution itself sets out a detailed process for determining whether a proposed Scottish law is within or outwith the legislative competence of the Scottish Parliament. More generally, the notion of devolution presupposes that the UK Parliament, as the only sovereign Parliament for the whole of the UK, is competent to legislate in any matter, devolved or reserved (Section 28(7) SA 1998).

The Devolution settlement for Scotland and Wales, at inception, was always subject to the, admittedly theoretical, possibility that it could be simply reversed by a subsequent Act of the UK Parliament. However, both the Scotland Act 2016 and Wales Act 2017 provide for a more permanent arrangement, whereby the Scottish Parliament, Scottish Government, Welsh Assembly and Welsh Government are now treated as permanent features of the constitutional landscape of the United Kingdom (Section 63A SA 1998). This strengthening of the centrality of devolution to the fundamental tenets of the UK constitution also resulted in significant changes to the interplay between the Scottish Parliament and the UK Parliament. The Scotland Act 2016 now provides that the Sewel Convention, the constitutional convention that the UK Parliament will not normally legislate for Scotland in devolved matters without the consent of the Scottish Parliament, is now statutorily recognised (see Section 28(8) SA 1998).

In the two decades of Scottish devolution, hundreds of Sewel conventions have been granted by the Scottish Parliament.

The slim UK vote to leave the European Union on 23rd June 2016, poses significant issues and implications for Scottish devolution. The Scottish population overwhelmingly voted to remain in the EU (62%), however the UK Government is committed to leaving the EU on terms that mean the UK no longer accepts the jurisdiction of the ECJ, leaves the EU Single Market and the EU Customs Union and ends payments to the EU budget. The Scottish Government was therefore placed in an impossible position - it had to respect and defend the democratic will of the Scottish people to stay in the EU and its economic and legal structures, within the confines of the overall desire of the entire UK to leave the EU.

The Scottish Government, from the immediate aftermath of the EU Referendum, was committed to engaging with the UK Government in reaching an acceptable compromise on the detailed terms of the UK's exit from the European Union. However, it later became clear that the terms of the UK's proposed Great Repeal Bill, and thereafter the EU Withdrawal Bill, were unacceptable to the Scottish Government. As a consequence, the Scottish Government intimated that it was likely the Scottish Parliament would be minded to refuse a Legislative Consent Motion (LCM) authorising the UK Parliament to enact legislation providing for the UK's exit from the EU. Such a refusal would have limited practical legal effect, since the UK Parliament has unlimited power to legislate and in particular, the Scottish Government has no power to enforce any refusal of an LCM against the UK Parliament (Miller v S. of S. for Exiting the EU [2017] UKSC 5], discussed here). Nevertheless, the refusal would have significant political effect. 

The Scottish Parliament refused to grant the LCM and the Scottish Government then pressed ahead with plans to create its own withdrawal legislation, in order to avoid a legal vacuum in the devolved statute book after the UK's exit from the EU. The Scottish Government put forward its alternative legislative proposal, the UK Withdrawal from the European Union (Legal Continuity) Bill 2018, on the 27th of February 2018. It was presented as emergency legislation before the Scottish Parliament and as such, the UK Government was not notified of the proposal, as is the normal convention.

As the law currently stands, Acts of the Scottish Parliament, as laws of a devolved legislature, do not enjoy an automatic presumption of legality, in contradistinction to the position of Acts of the UK Parliament (Imperial Tobacco Ltd v Lord Advocate 2013 SC (UKSC) 153 as per Lord Hope of Craighead). As such, legislative proposals of the Scottish Parliament are vulnerable to being attacked as being outwith the competence of the Scottish Parliament under two main lines of attack.

Firstly, Bills of the Scottish Parliament can be attacked as being effectively unconstitutional during the four week gestation period of a bill between its approval in the Scottish Parliament and being given Royal Assent (Section 32 SA 1998). The Advocate General for Scotland, the Lord Advocate or the Attorney General are entitled to challenge the legality of a Bill of the Scottish Parliament before the Supreme Court (Section 33 SA 1998).

Secondly, actual Acts of the Scottish Parliament are similarly susceptible to attack as being outwith the legislative competence of the Scottish Parliament before the Supreme Court (Section 29 SA 1998; AXA v Lord Advocate [2011] UKSC 46). 

The UK Withdrawal from the European Union (Legal Continuity) Bill 2018 was criticised by no less a figure than the Presiding Officer of the Scottish Parliament itself. The Bill attempts to remove the limitation on legislative power set out in section 29 of the Scotland Act 1998 as regards EU law. The Bill also purports to bring EU law into the devolved statute book by creating Retained (Devolved) EU law and gives the Scottish Minsters powers to amend such law where this is necessary to make the law coherent and consistent. The Bill also attempts to require that UK Ministers do not change certain aspects of Retained EU law without the consent of the Scottish Ministers.

When the UK leaves the EU, currently scheduled as the 29th of March 2019 under the UK Parliament's EU (Withdrawal) Act 2018 (Section 20), the provisions of the Scotland Act 1998 limiting the power of the Scottish Parliament and Government to make law and perform executive functions respectively, would be rendered nugatory. However, the Presiding Officer considered that this was a step too far: the Scottish Parliament will be fully subject to the limitation set out in section 29 until Exit Day. The Attorney General and Advocate General for Scotland completely agreed with this view and accordingly utilised their right to petition the Supreme Court for a judgment on the legality of the Scottish Parliament's Continuity Bill (Reference on UK Withdrawal from the European Union (Legal Continuity) Bill 2018).

Conversely, the Lord Advocate takes the diametrically opposite view, alongside the Attorney General for Northern Ireland and the Counsel General to the Welsh Government. These three legal officers consider that the Bill was squarely within the competence of the Scottish Parliament, notwithstanding the reservations expressed by the Parliament's Presiding Officer.

The Supreme Court, in a seven judge configuration, including of course the two Scottish justices (Lord Reed and Lord Hodge) heard the case on the 24th and 25th of July 2018 and delivered its judgment on the 13th of December 2018.

The challenge is without precedent: it is the first time, in the two decades of operation of the Scottish Parliament that both the Presiding Officer has questioned the competence of a Bill passed by the Scottish Parliament and the UK Legal Officers have exercised their legal right to send a question on the competence of a Scottish Parliament Bill to the Supreme Court.

The novelty of the case has thrown up some interesting legal questions.

First, it has to be remembered that this action before the Supreme Court is unique. It is not analogous to the many legal challenges that have been taken to the Supreme Court about the legality of devolved legislation (eg AXA v Lord Advocate [2011] UKSC 46) as opposed to bills. In the AXA case, Lord Hope at p. 142, stated that;

“in principle Acts of the Scottish Parliament are amenable to the supervisory jurisdiction of the Court of Session at common law".

This judgment is authority for the proposition that Acts of the Scottish Parliament are vulnerable to challenge under the limits to Scottish Parliament law making set out in section 29 of the Scotland Act 1998 and judicial review principles of the common law. However, he then went on to qualify this statement by noting that not all of the generally available judicial review grounds of challenge would be available to challenge the provisions of a public body such as Acts made by the Scottish Parliament since the Scottish Parliament was a democratically elected body with law making and tax raising powers and as such its laws would enjoy the highest legal authority.  Thus, challenges to laws of the Scottish Parliament could not ordinarily be taken on grounds of irrationality, unreasonableness or arbitrariness. Indeed, Lord Hope stated that the common law grounds of challenge could only ever be contemplated in extreme circumstances, such as where a law purported to violate the rule of law by closing off judicial review or the courts' ability to protect an individual's rights.

Although factually and legally correct, and intensely interesting, the Lord Advocate swiftly dispensed with this line of argumentation. These options for review are only appropriate for challenges taken against laws made by the Scottish Parliament. Legislative proposals before the Scottish Parliament can only be challenged in the 4 week window of opportunity according to the exacting terms of section 33 of the Scotland Act 1998. As such, it is argued that the question before the Supreme Court is not a nuanced one akin to cases seeking annulment of laws based on rule of law conceptions such as irrationality and arbitrariness in the exercise of powers stated in that law. Rather, it was submitted by the Lord Advocate that the question was simpler and binary in nature: is the Bill within the legislative competence of the Scottish Parliament, by not encroaching upon matters properly addressed by the UK Parliament or not?

If the question is indeed binary as suggested, then the substantive question turns on the notion of international relations and EU law as international law. Again, here the question may not be as complicated as argued by the UK Law Officers. The UK Law Officers argue that the Bill is not within legislative competence since the Bill cuts across fundamental sovereign powers only exercisable by the UK Parliament, namely the exercise of international relations. Succour for this line of argumentation comes from the Miller case, whereby the Supreme Court conformed that the triggering of Article 50 TEU fell squarely within the legal power of the sovereign Parliament of the UK (Miller v S. of S. for Exiting the EU [2017] UKSC 5). However, again here there is possibly a misunderstanding of the scope and nature of the powers purportedly being exercised in the Scottish Parliament's Bill. Of course, the contested Bill is concerned with EU law and the Scotland Act clearly prohibits the Scottish Parliament from acting in the international plane or to do anything interfering with the sovereign power of the UK Government and UK Parliament to both conduct relations with (and within) the EU institutions and to change UK law, notably the European Communities Act 1972, to comply with changed EU law obligations (such as additional Treaty obligations).

However, the Scotland Act is also concerned with a completely separate, internal aspect of EU law: the Scottish Parliament and Government cannot make law or perform executive functions in violation of EU law. This provision is actually designed to protect the UK from the legal consequences of devolved institutions acting, to borrow a delictual and tortious phrase, on a frolic of their own which will impute liability upwards to the UK. This limitation on power set out in section 29 of the Scotland Act 1998 is vital and core to the very essence of the Scottish Parliament's function: an overwhelming majority of legislative activity in the Scottish Parliament is directed at complying with EU law and giving effect to EU law. At the inception of devolution at the tail end of the last century, the Scottish Parliament and Scottish Government were given power over any areas of competence not reserved to the UK Government and Parliament.  Most of these devolved powers were actually powers exercised at the EU level, on a shared basis with the member States - agriculture, fisheries, environmental protection, consumer protection etc.

Therefore, the Scottish Parliament was directly tasked with complying with EU law, but this 'internal' aspect of EU law in no way impugns the 'external' aspect of EU law, exercisable by the UK Government and Parliament. Relatedly, it has already been judicially recognised that the devolution institutions will experience a significant Brexit dividend - an entire swathe of restrictions on the competence of the devolved institutions will be completely expunged on Brexit day (see Miller). This 'internal' aspect of EU law directly concerns the Scottish Parliament and it would be nonsensical if the Scottish Parliament could not take measures to clarify the scope of how this fundamental shift in its competences is to operate.

The Lord Advocate also dispenses with the argument made by the UK Law Officers concerning the future contingency nature of the provisions of the Bill. The UK Law Officers argue that the terms of the Bill, as drafted would cause the Scottish Parliament to be in breach of its obligation not to undermine EU law as set out in section 29 of the Scotland Act 1998 by simple dint of the fact that the Scottish Parliament must comply with EU law right up to 11pm on the 29th of March 2019. The UK Law Officers set much store on the contingent nature of the power given by the Bill to the Scottish Ministers to bring into force such provisions once the Bill became law. This argument appears weak. Commencement orders are a perfectly natural feature of the UK legal landscape, both devolved and reserved. It is perfectly normal and legal for a legislature to anticipate future changes and to legislate accordingly. Further, if the said contingency never materialises, then it is simply the case that that legal power is never exercised and can be replaced or repealed.

A further objection to the proposed Bill was taken by the UK Law Officers to the effect that the Bill could not alter the terms of the Scotland Act 1998 as regards the restriction of legislative competence on the grounds of EU law compliance (section 29(2)(d) SA 1998). The Lord Advocate tersely rejected this argument as well.  The system of devolution generally provides that Acts of the Scottish Parliament can be used to modify Acts of the UK Parliament, where that UK law relates to devolved competences. This significant power is a well-established feature of the devolved landscape. Nevertheless, legally and practically there has to be limitations upon this power, otherwise the devolved Parliament would, at a stroke, have been transformed into an omnipotent, fully sovereign Parliament: An anathema to the very nature of devolution.

Thus, the Scotland Act 1998 does generally provide that an Act of the Scottish Parliament cannot be used to modify certain central tenets of the constitutional devolution framework as provided for in the Scotland Act 1998. To put it more simply, the Scottish Parliament cannot pass an Act of Parliament giving itself more power. However, the Scotland Act 1998 does allow for some modifications to take place (Schedule 4 SA 1998). For present purposes, para. 7 of Schedule 4 of the Act is central. This part of the 1998 Act provides that an Act of the Scottish Parliament can be passed to take out 'spent' provisions. The key argument facing the Supreme Court is the notion of when the references to the EU limitations on the legislative competence of the Scottish Parliament become spent. If the UK Law Officers are right, the Scottish Parliament cannot act to expunge the references to EU law limitations on Scottish Parliament competence until the UK leaves. Form the Lord Advocate's perspective this view is unduly restrictive. It binds the hands of the Scottish Parliament until Brexit Day and, given the inherent uncertainty over the exact legal implications of the UK's departure from the UK on the 29th of March 2019, the Scottish Parliament cannot act prospectively to arrange its affairs as regards Scots law (the devolved statute book) to manage and ensure an orderly, legally certain departure.

The final aspect of the case which is intriguing is brought about by the mere effluxion of time. At the time of passing the Bill in the Scottish Parliament, the UK Parliament had yet to pass its own legislative proposal for an orderly Brexit. But soon after the legal objection was made to the Scottish Bill, the UK Parliament's EU (Withdrawal) Bill received Royal Assent. This law provides for retained EU law to be created for the utility of the entire UK. However, of course, the Scottish Parliament objected to the terms of that law, refused a Legislative Consent Motion and proceeded to continue to attempt to enact its own continuity law.

A direct consequence of that fact was that the Scotland Act 1998 was then amended to add the terms of the EU (Withdrawal) Act 2018 into the list of laws in Schedule 4 that are unamendable by the Scottish Parliament. Thus, the practical result is that if the Scottish Parliament were to now try and introduce a Bill along the lines of the impugned Continuity Bill, that Bill would be clearly outwith the legislative competence of the Scottish Parliament. Thus, the Lord Advocate has craved that the Supreme Court consider the question of legislative competence as of the date of passing the Bill and not at a later date in time. If the Supreme Court were to look at the case as from a later date in time, then the reference has been of purely academic interest.

However, even if the Supreme Court looks back in time to the point of passing the Bill, the answer also becomes academic since the unamended Bill, from a practical point of view cannot be passed since it has been rendered otiose. The best the Lord Advocate can hope for now is that the Bill is characterised as being in legislative competence and that certain, now unlawful provisions of the Bill, can be expunged from it, allowing the legislative proposal (as amended) to be turned into an Act of the Scottish Parliament.

This morning’s judgment is a clear vindication of the Lord Advocate’s position. The court was unanimous that, at the point in time when the Bill was passed in the Scottish Parliament, the entire Bill was within the legislative competence of the Scottish Parliament, with the sole exception of section 17 of the Bill. However, time marched onwards such that, at the time of this judgment, some of the legal space about to be occupied by the Scottish Bill was now occupied by the law of the sovereign Parliament of the United Kingdom. As such, the court therefore qualified its decision and recognised the clear legal reality pertaining at the date of the judgment - certain provisions of the Bill were now rendered otiose and legally impermissible as a direct result of the coming into force of the terms of the UK Parliament’s EU (Withdrawal) Act 2018. Thus, it is in part a pyrrhic victory for the Lord Advocate. The Bill, as presented at the time of progressing to Royal Assent was predominantly within the legislative competence of the Scottish Parliament. However, large sections of the Bill now fall away since they are in direct conflict with the terms of the 2018 Act and are thus outwith legislative competence of the Scottish Parliament.  

In these uncertain times, it is absolutely certain that there will now be interesting conversations regularly taking place between HM Government and the Scottish Government to resolve this legal conflict.

Barnard & Peers: chapter 27

Photo credit: Daily Record

Anonymity in CJEU cases: privacy at the expense of transparency?

Peter Oliver, Barrister, Monckton Chambers; Visiting Professor, Université Libre de Bruxelles

Juliet famously asked: “Romeo, Romeo wherefore art thou Romeo?”  And then adds: “What's in a name? That which we call a rose by any other word would smell as sweet.”  Having an unfortunate name (like Montague, if you have the misfortune to fall in love with a Capulet) can be challenging.  But what about having no name?  That can be decidedly problematic too.  That is why both the Strasbourg and the Luxembourg courts have recognised a person’s fundamental right to the use of their own name.

The long-standing practice of the Court of Justice has been to replace the names of asylum seekers, children and other individuals in sensitive litigation by initials, while using the actual names of natural persons in other contexts.  Then, in late June, the Court announced out of the blue that this long-standing policy was to partially be reversed: since 1 July, the default position is that, where preliminary references are made in cases to which individuals are party, their names will be replaced by initials which do not correspond to their actual initials. 

Needless to say, the impetus behind this change was the Court’s own case law on the fundamental right to data protection, the recent entry into force of the GDPR and the proposed (and subsequently adopted) Regulation to replace Regulation 45/2001 on the processing of personal data by the Union’s institutions. 

Of course, data protection is not merely commendable, but absolutely indispensable in the digital age.  But this fundamental right, which is enshrined in Article 8 of the Charter, is not absolute, as is plain from the language of that provision itself and Article 52(1) of the Charter.

Consequently, this right must be balanced against an individual’s right to the use of his or her first name and surname.   It is settled law that, because this is an essential element of a person’s identity, it falls within the right to private and family life under Article 8 ECHR.  Over twenty-five years ago, in his seminal Opinion in Konstantinidis, AG Jacobs introduced the same principle into EU law; and it now undoubtedly falls within Article 7 of the Charter.

In addition, the freedom of expression enshrined in Article 11 of the Charter must be borne in mind.  Article 85 of the GDPR specifically requires the Member States to balance data protection with the “freedom of expression and information”, including processing for the purposes of journalism.

As one would expect, the Court’s announcement makes it clear that the Court “may” depart from this new rule “in the event of an express request from a party or if the particular circumstances of the case so justify”.   Unfortunately, this phrase is so nebulous as to give no guidance as to the circumstances in which the Court will consider that the “particular circumstances” will justify this.

In the well-known case of Wightman (discussed here), the Court has set out the full names of all the applicants: that is scarcely surprising since they are all politicians or activists.   Frequently, however, litigants who do not – initially, at least – fall into either category are proud to have their names attached to a leading case.  Will individuals always be informed of their right to have their names used in the proceedings in time to inform the Court of their wish before it is too late? 

Rumour has it that the Court is contemplating informing the senior judiciary of the Member States of each new preliminary reference more speedily than in the past so as to deter other national courts from posing the same preliminary questions.  That seems an excellent idea, but it should not work to the detriment of the fundamental rights of litigants.  In any case, there is nothing to preclude the Court from withholding their names when first making the existence of a new preliminary reference known, and revealing them shortly afterwards if it is appropriate to do so.

The Court was able to effect this radical and abrupt change of policy without seeking any amendment of the Rules of Procedure, which would have required the approval of the Council by virtue of Article 253 TFEU.  The stage was already set for the Court to do so by Article 95 of the Rules of Procedure, which reads:

1. Where anonymity has been granted by the referring court or tribunal, the Court shall respect that anonymity in the proceedings pending before it.

2. At the request of the referring court or tribunal, at the duly reasoned request of a party to the main proceedings or of its own motion, the Court may also, if it considers it necessary, render anonymous one or more persons or entities concerned by the case.

All the Court had to do was to announce that it was altering its practice as to when it would anonymise parties’ names “of its own motion” in future.  What is more, it is not clear whether, in keeping with normal rules of good governance, the Court took to the trouble to consult any other institutions or bodies such as the other EU institutions, the Member States or the CCBE before taking this important step.

Article 95 applies only to preliminary rulings.  However, it is supplemented by Article 190(3), which provides:

Article 95 shall apply, mutatis mutandis, to the procedure before the Court of Justice on an appeal against decisions of the General Court.

Nevertheless, as already mentioned, according to the Court’s announcement, its new policy is confined to preliminary rulings.  Why?   The most plausible explanation is that it did not wish to interfere with the decisions of the General Court in this regard, or realised the futility of doing so. (Article 66 of the General Court’s Rules of Procedure broadly corresponds to Article 95(2) of the Rules of Procedure of the Court of Justice.  Since the Court changed its practice in June, the GC has not followed suit so far.)

What is striking is the disparity between the Court’s apparent willingness to accept the GC’s judgement on matters of anonymity and its determination to impose its will with respect to preliminary references, regardless of the practice of the referring court.  The Court’s new practice is puzzling on two counts: first, it appears to run counter to the Court’s own mantra that preliminary rulings involve cooperation between itself and national courts; and imposing anonymity makes no sense when the national courts are unwilling to follow suit once a case is remitted to them after the Court’s preliminary ruling has been delivered. 

The Court has no power to prohibit the courts of the Member States from continuing to use a litigant’s real name, nor is the Court in any way empowered to stop journalists or others from doing so. This is graphically illustrated by the high profile case known as Celmer in Ireland, which became LM in Luxembourg (on this judgment, see discussion here) but went back to the Irish High Court as Celmer (see for instance the High Court’s final ruling in the case); and Artur Celmer’s name been mentioned frequently in the Irish press (see here and here).

Following the Court’s announcement, it published a new recommendation to national courts on preliminary references in which it called on national courts to remove the names of natural persons from those references; but that is purely aspirational, on the Court’s part except where anonymisation is required by the GDPR or Article 8 of the Charter. 

Needless to say, this absurd situation does little to enhance the Court’s standing or authority.  Of course, in the common law tradition, parties’ actual names are used in the overwhelming majority of court proceedings.  But, even after Brexit, Ireland is not likely to be the only Member State which is at odds with the Court’s new approach, since the use of parties’ names is very widespread in several national legal systems.

What is more, the Court’s website is not up to its new task of finding random initials – which is highly problematic for those users who lack the gift of memorising large quantities of case numbers.  My search for “LM” yielded just one reference – to an eponymous staff case.  Only after a further five or ten minutes of frantic searching did I unearth what I will continue to call the Celmer case.  Had it not been a very recent case, my efforts would no doubt have taken longer, as I would have had to sift through more cases.

In future cases, names such as Bosman, Coman and Klohn - which enable lawyers to find cases on the website in a trice – will simply disappear. This will make research into the case law far harder, and could even make it more likely that relevant judgments are omitted from submissions to the Court itself.  What is more, students of EU law will inevitably suffer, as will their professors.

Adding key words to case names, as the Court has promised to do in its announcement in certain cases, will help but it will not be as effective as using the litigant’s actual name, unless the key words are very rarely found in the case law – which is not always feasible.  In the subsequent cases, LM is referred to as Minister for Justice and Equality (Deficiencies in the System of Justice); the letters “LM” have disappeared.  Unless you remember the precise name of the Ministry or search for “deficiencies”, you are likely to lose a good deal of time.  Unlike Google, the Court’s website does not look for words with a similar spelling or meaning to those requested.

In short, this new practice will make the Court’s case law less accessible to practitioners, academics and students alike.

At the very least, the Court should respect the practice of the referring court in each case so as to avoid a repetition of the Celmer debacle, unless there are quite exceptionally strong reasons to do otherwise.  No doubt, other mechanisms can be devised to mitigate the effects of this hastily introduced new policy, including an improvement of the Court’s website.

Barnard & Peers: chapter 10

Photo credit: Inforrm’s blog

The Global Compact for Migration: cracks in unity of EU representation

Pauline Melin, PhD, Lecturer in European Law (Maastricht University) and Researcher at the Institute for European Law (KULeuven)

Today, the UN Global Compact for Safe, Orderly and Regular Migration is meant to be approved by an intergovernmental conference in Marrakesh with the purpose to be adopted on the 19th December 2018 as a Resolution of the UN General Assembly in New York. The debate over the adoption the Global Compact for Migration has been politically polluted. Since the end of October 2018 and the withdrawal of Austria, a Member State who had until then positively contributed to the negotiations Global Compact through the Council Presidency, no week has passed without a new announcement of an EU Member State withdrawing or doubting its support for the Global Compact for Migration. That vague of lack of support coming from the EU Member States is in contradiction with the (supposedly) common position presented by the Union delegation during the negotiation process. Those cracks in unity of representation of the EU on the international scene raise questions concerning the role of the EU in the negotiations of the Global Compact and the consequences for its Member States in light of the principle of sincere cooperation.

The role of the EU in the negotiation of the Global Compact for Migration

The process eventually leading to the adoption of the Global Compact started in September 2016 with the New York Declaration whereby 193 Heads of State and Government recognized the need for developing an international cooperation on migration.

Throughout its different drafts (the Zero Draft on 5th February 2018, the Zero Draft Plus on 5th March 2018 and the Final Draft on 11th July 2018), the wording of the Global Compact concerning its legal nature stayed consistent. The Global Compact is intended to present “a non-legally binding, cooperative framework that builds on the commitments agreed upon by Member States in the New York Declaration for Refugees and Migrants.” (emphasis added, points 5,6, and 7 of the Zero Draft, Zero Draft Plus and Final Draft respectively).

From the UN side, the negotiations of the Global Compact were meant to be as inclusive as possible with the participation of all its State Members, all members of specialized agencies that have an observer status with the General Assembly, intergovernmental organizations and other entities having received a standing invitation. The EU, as a regional group, was granted standing status in order to participate in the negotiations and the conclusion of the Global Compact.

From an EU perspective, given the fact that the Global Compact for Migration is a non-legally binding international instrument, the procedure for negotiating and concluding an international agreement enshrined in Article 218 TFEU was not applicable (C-233/02 France v. Commission, para. 40). The EU participated in the negotiation of the Global Compact through the delivery of Union delegation statements. According to the European Commission, the Union delegation statements were “EU coordinated statements” constituting an “unified EU approach”. That unified EU approach must however be nuanced considering that, since March 2018, Hungary has proposed a very different approach than the one defended by the Union delegation. Furthermore, the exclusion of Hungary from the EU coordinated statements can be seen from the Union delegation’s statements themselves which mention “on behalf of 27 Member States”.

The (lack of) common position

The fact that the Global Compact is a non-legally binding international instrument does not entail that the principles of conferral, institutional balance or sincere cooperation should not be respected. In case C-660/13 Council v. Commission on the Swiss MoU, the Court found that the Commission needs a Council Decision authorizing it to sign the negotiated text of a non-binding instrument before it can approve it on behalf of the Union. In the context of the negotiation of the Global Compact for Migration, the Commission in fact relied on this case in order to justify its proposal for Council Decisions (since withdrawn) authorizing it to sign the Global Compact in Marrakesh on behalf of the Union.

However, as opposed to the Council v. Commission case, the Commission here did not seem to have a clear negotiating mandate from the Council. In Council v. Commission, there were some Council Conclusions adopted in 2012 authorizing the start of the negotiation between the Swiss Confederation and the EU whereby the content of the negotiation as well as the designation of the European Commission as the negotiator were set out. The European Commission then considered that because the negotiated text was similar in content to the negotiated mandate, it could sign the negotiated text on the basis of Article 17(1) TFEU (para.35). However, the Court disagreed and held that neither Article 17(1) TEU nor the negotiating mandate found in the Council Conclusions were sufficient bases for authorizing the Commission to sign the negotiated text without a Council Decision on the matter.

In the case of the negotiation of the Global Compact for Migration, there is in fact no Council Decision authorizing the start of the negotiation. In order to justify its negotiating position on behalf of the Union, the European Commission relies on two documents: the European Council Conclusions on Migration from October 2016 and the European Consensus on Development from 2017. In the European Council Conclusions on Migration, it is simply mentioned that the European Council welcomes the New York Declaration. In the European Consensus on Development, it is stated that the EU and its Member States will actively support the elaboration of the UN Global Compacts on Migration and Refugees. While the two documents could be considered as an indication that there is a willingness from the European Council to have a common position in the negotiations of the Global Compacts for Migration and Refugees, there is nothing in the documents that either identify the European Commission as the negotiator on behalf of the Union and its Member States nor that indicate what the content of that common position would be. With no official document proving the common position of the EU and its Member States, one might wonder whether the Member States were obliged to abstain from contradicting the Union delegations’ statements.

The point of departure of the principle of sincere cooperation

From the PFOS case, it seems that the principle of sincere cooperation whereby the Member States should support the position expressed by the EU or, at least, abstain from contradicting it starts from the moment there is an established common position (para. 89). In the PFOS case, Sweden had submitted a proposal to list PFOS in Annex A of the Stockholm Convention while a Commission proposal for a Council Decision authorizing the Commission to submit on behalf of the Union and its Member States a list of chemicals to be added in the Annexes of the Stockholm Convention did not comprise PFOS within the list. Firstly, the Court recalled that the principle of sincere cooperation flows from the requirement of unity in international representation of the Union (para.73). Then the Court reminded that the submission by the Commission of a Council proposal for the authorization to start the negotiation of a multilateral agreement represent the point of departure for a concerted action even though the proposal has not been adopted by the Council (para. 74). The Court added that the establishment of a common position through a Council Decision was not a prerequisite for its existence as long as “the content of that position can be established to the requisite legal standards” (para.77). In the PFOS case, the Court found that the fact that Council’s conclusions and minutes of the meeting of the Council’s Working Party on International Environmental Issues were sufficient evidences of an established common position which the Member States should support or, at least, abstain from contradicting (para.89). In reference to the “requisite legal standards”, the Court referred another case whereby the Council and the Commission had concluded an arrangement setting up a coordination procedure between the Commission and the Member States to decide on the exercise of responsibilities or on statements in the context of the UN Food and Agriculture Organization (‘FAO’).

In the context of the drafting process of the Global Compact for Migration, one may consider the General Arrangements for EU Statements in multilateral organisations as being the relevant requisite legal standards. In the General Arrangements, point 2 indicates that “[G]iven the sensitivity of representation and potential expectations of third parties, it is essential that, in conformity with current practice, the preparation of statements relating to the sensitive area of competences of the EU and its Member States should remain internal and consensual.” (emphasis added). Hence, while the need for a Council Decision is not required, there is a need to have an internal discussion and a consensus about the common position. Considering the fact that internal coordination for setting a common position takes place behind closed doors, it is not possible to ascertain whether the internal coordination on the Global Compact for Migration actually took place.

Conclusion

While there might have been internal coordination and an agreed common position, the lack of evidence of it becomes a problem when Member States start expressing opinions contradicting the Union’s position. The uneasiness of the European Commission faced by rebellious Member States can be felt through the timing of its proposals for Council Decisions authorizing it to sign the Global Compact for Migration on behalf of the Union. The European Commission adopted those proposals in March 2018 whereas the Final Draft of the Global Compact for Migration was only agreed in July 2018. As a result of the lack of proof of an agreed common position, it is difficult for the European Commission to ensure that the Member States respect their obligations under the principle of sincere cooperation. However, there should be a strong interest for the European Commission to avoid cracks in unity of representation as it ultimately diminishes the credibility of the EU as a global actor. Given the growing relevance of non-binding international instruments, it is unlikely that the cracks in unity of representation will remain anecdotal to the approval of the Global Compact. It is therefore time to think about a clear legal framework for the negotiation and conclusion of non-binding international instruments.

Barnard & Peers: chapter 24, chapter 26

Photo credit: Steve Peers

Brexit: you can U-turn if you want to. The CJEU judgment in Wightman

Professor Steve Peers, University of Essex

Today’s Full Court judgment in the Wightman case confirms that the UK can unilaterally withdraw its notification that it intends to leave the EU, on the most generous possible terms. It broadly follows last week’s non-binding opinion from an Advocate-General, discussed here. (See also the discussion here of the national court background to the proceedings, and the discussion here of the arguments for and against unilateral revocability. The EU courts have also ruled on a challenge to the withdrawal agreement negotiations in Shindler, and on the UK’s current status as a Member State in RO: see discussion here and here).

The judgment

First of all, the Court rejects the UK government’s argument that the case is hypothetical, noting that some of the litigants are MPs who will be voting imminently on the proposed withdrawal agreement (unless that vote is delayed). It displays its usual deference to national courts’ decision to ask the CJEU questions about EU law, which leads to a presumption of relevance.

On the merits, the Court takes its usual view that EU law should be interpreted taking account of its wording and objectives, but in light of its context and the provisions of EU law as a whole. On the wording, the Court notes that Article 50 TEU (the Treaty provision on the withdrawal process) is silent either way about the revocation of a notification of the intention to withdraw from the EU.  But the Court points out that Article 50 refers to notifying an ‘intention’ to withdraw: ‘An intention is, by its nature, neither definitive nor irrevocable’.

Observing that the decision to withdraw its unilateral, in accordance with a Member State’s ‘own constitutional requirements’, the Court rules that

the Member State is not required to take its decision [to withdraw] in concert with the other Member States or with the EU institutions. The decision to withdraw is for that Member State alone to take, in accordance with its constitutional requirements, and therefore depends solely on its sovereign choice.

As for the objectives of Article 50, the Court characterised it as having two objectives: ‘first, enshrining the sovereign right of a Member State to withdraw from the European Union and, secondly, establishing a procedure to enable such a withdrawal to take place in an orderly fashion’. It then located the issue of revocation as part of the first of these objectives: linking revocation with the sovereign decision to withdraw, and clarifying the timing of the right of revocation:

…the sovereign nature of the right of withdrawal enshrined in Article 50(1) TEU supports the conclusion that the Member State concerned has a right to revoke the notification of its intention to withdraw from the European Union, for as long as a withdrawal agreement concluded between the European Union and that Member State has not entered into force or, if no such agreement has been concluded, for as long as the two-year period laid down in Article 50(3) TEU, possibly extended in accordance with that provision, has not expired.

It is clear that the unilateral right to revoke the notification still exists if the Article 50 period is extended (which must be unanimously agreed by the withdrawing Member State and the EU27). The Advocate-General’s opinion had conversely been unclear on this; it’s a crucial point since an extension would likely be necessary if another referendum on Brexit were held in the UK.

What rules apply to unilateral revocation? Since Article 50 is silent, the Court says that the same rules apply to withdrawal as applied to the original notification: ‘it may be decided upon unilaterally, in accordance with the constitutional requirements of the Member State concerned.’ The Court also confirms that a revocation would mean that the UK retains its current status as an EU Member State, as the revocation would reflect ‘a sovereign decision by that State to retain its status as a Member State of the European Union, a status which is not suspended or altered by that notification’ (following the CJEU’s previous ruling in RO), ‘subject only to the provisions of Article 50(4) TEU’ (which says that a departing Member State does not participate in EU decision-making concerning the withdrawal agreement). Revocation ‘is fundamentally different’ from a request for extension of the Article 50 time period, which entails unanimous consent of the EU27, rejecting the analogy with the second objective of Article 50 which the EU Commission and Council wanted the Court to make.

As for the context of Article 50, the Court stressed the Treaty objectives of an ‘ever closer union among the peoples of Europe’, EU citizenship, and the values of liberty and democracy, noting that ‘the European Union is composed of States which have freely and voluntarily committed themselves to those values’, and that ‘any withdrawal of a Member State from the European Union is liable to have a considerable impact on the rights of all Union citizens, including, inter alia, their right to free movement, as regards both nationals of the Member State concerned and nationals of other Member States’. Therefore, ‘given that a State cannot be forced to accede to the European Union against its will, neither can it be forced to withdraw from the European Union against its will’, which would be the case if a ‘Member State could be forced to leave the European Union despite its wish — as expressed through its democratic process in accordance with its constitutional requirements — to reverse its decision to withdraw and, accordingly, to remain a Member of the European Union’.

The Court also looks at the process of drafting the earlier version of Article 50, during which various proposed amendments were rejected, ‘on the ground, expressly set out in the comments on the draft, that the voluntary and unilateral nature of the withdrawal decision should be ensured’. Moreover, the Court’s findings were ‘corroborated’ by the Vienna Convention on the Law of Treaties, ‘which was taken into account’ when drafting the earlier version of Article 50. That Convention states ‘in clear and unconditional terms, that a notification of withdrawal’ from a treaty ‘may be revoked at any time before it takes effect’.

Next, the Court rejects the argument of the Council and the Commission that revocation would need unanimous consent, as this ‘would transform a unilateral sovereign right into a conditional right subject to an approval procedure’, which ‘would be incompatible with the principle…that a Member State cannot be forced to leave the European Union against its will’.

Finally, the Court sets out the conditions for revocation; it:

…must, first, be submitted in writing to the European Council and, secondly, be unequivocal and unconditional, that is to say that the purpose of that revocation is to confirm the EU membership of the Member State concerned under terms that are unchanged as regards its status as a Member State, and that revocation brings the withdrawal procedure to an end.

The Court does not otherwise address the argument of the Council and Commission, which was discussed by the Advocate-General, that there must be a ‘good faith’ element to revocation.

Comments

The Court’s judgment may ultimately have no impact on the political likelihood of whether the UK reverses Brexit – which I continue to think is very unlikely. It does lower the potential barriers to a U-turn by the UK, but this may only strengthen the resolve of Brexit supporters, rather than change their mind.  

Let’s look at the judgment from a legal perspective. Even more so than the Advocate-General’s opinion, this ruling strongly supports unilateral revocation of a notification on the easiest possible terms. The admissibility of the case is unsurprising in light of the prior case law deferring to national courts. As the Court says, Article 50 is silent on the issue and that does not point necessarily to a resolution to the issue, but the Court was right to point out that there’s an explicit reference to an intention to withdraw in the wording of Article 50.

The Court’s ruling on the unilateral nature of the initial notification to withdraw confirms the recent EU General Court judgment in Shindler on the nature of the UK’s initial decision, which was not ‘approved’ or filtered by the EU institutions upon receipt of the notification. The parallel which today’s judgment draws between notification and revocation suggests that if Shindler is upheld on appeal on this point, as it logically should be, there could be no review by EU institutions of the revocation, provided that it meets the very minimal requirements set out by the Court (and discussed further below).  

It seems that the key to the logic of the judgment is the Court’s characterisation of Article 50 of having two objectives, and then categorising revocation of notification as an aspect of the first objective – the process of deciding to withdraw – instead of an aspect of the second objective – the orderly withdrawal process. It followed from this that the Court drew an analogy with the unilateral nature of the decision to withdraw, rather than the bilateral nature of the withdrawal agreement negotiation process, and in particular the unanimous requirement to extend that process.

As for the link to international law, it contradicts the usual autonomy of EU law from international law that the Court refers to itself at the outset of the judgment, but the Court justifies that because the drafters of what became Article 50 took the Vienna Convention into account. It is, in any event, only a secondary part of the Court’s reasoning.

Finally, on the conditions for revocation, the submission in writing is straightforward enough: the EU institutions could surely work out whether a revocation was genuine or not, in light of the publicity that would obviously accompany it. The requirement of an ‘unequivocal and unconditional’ revocation, ending the withdrawal process on unchanged terms, suggests that the notification of revocation must confirm that the UK is not intending to renegotiate its membership or to send another notification shortly afterward. Implicitly if the UK revocation arguably did either of those things then the legal question would arise of what the European Council could do about it. It could either refuse to accept the notification, with the result that the UK might then challenge that decision; or it could simply decide to cross that bridge when it came to it, either refusing to renegotiate membership or (more problematically) to accept a fresh notification of withdrawal if that followed shortly after a revocation of the first notice (again, that decision could then be challenged).  

But the Court, unlike the Advocate-General, makes no mention of the domestic process leading up to revocation, noting only that it must be in accordance with the UK’s own constitutional requirements. In Shindler, the General Court said that these requirements were not for the EU institutions to judge, but the UK’s national courts and political institutions, with the proviso that a national Court might ask the CJEU if a particular requirement was compatible with EU law.

Overall, then, legally the road is clear for a U-turn if the UK wants to – but that is irrelevant as long as the lady is not for turning. Whether she changes her mind – or someone else takes the wheel and does so instead – remains to be seen.

Barnard & Peers: chapter 27

Photo credit: Millenium Post