Professor Steve Peers, University of Essex
Introduction
Is initiating the Article 50 withdrawal process from the EU irreversible, like jumping off a cliff? Or is it more like going shopping, where you can change your mind along the way, in the parking lot, or even inside the store? Perhaps this change of heart might be due to discovering the promised sale doesn’t exist, the management is under investigation, and the supposed extensive international food section is just a couple of dusty packages.
The Court of Justice will soon provide an answer. Currently, we have a non-binding opinion from an Advocate-General. While not binding, the Court will carefully consider the opinion, so discussing the issues it raises is worthwhile.
The Opinion
The opinion dismisses the UK government’s claim that the case is hypothetical because they haven’t decided to withdraw the Article 50 notification and have no plans to do so. The ruling’s outcome could be relevant when Parliament votes on the withdrawal agreement or decides what to do if it’s rejected.
The opinion sides with the petitioners—primarily UK politicians—stating the notification is, in principle, unilaterally revocable. It notes Article 50 doesn’t address revocation, leaving three possibilities: no revocation, unconditional revocation, or conditional revocation. The opinion examines general international law, Article 50’s wording, and agreed revocation.
Regarding international law, it’s unclear if revoking a treaty withdrawal notification is established customary practice, despite examples. However, the Vienna Convention on the Law of Treaties (VCLT) expressly allows it. The opinion concludes the VCLT doesn’t apply to EU withdrawal, as the EU and some members aren’t party to it. However, the VCLT might help interpret Article 50.
Article 50’s literal interpretation doesn’t settle the matter, requiring examination of its context. The Advocate-General sees it starting with a national phase solely for the withdrawing state to decide on withdrawal, contingent only on its constitutional requirements. This implies a unilateral power to revoke that decision, reflecting the state’s sovereignty. Unilateral powers persist into the negotiation phase, beginning with notifying the European Council and ending in two years, extendable by unanimous consent. Like other legal areas, absent a prohibition or alternate rule, unilateral declarations of intent can be retracted until accepted by the addressee, either through action or a contract.
The opinion highlights Article 50(2) TEU referring to notifying an ‘intention’ to withdraw, not the withdrawal itself, which occurs later. ‘Intention’ isn’t final and can change; notifying a third party might create expectations but doesn’t obligate maintaining that intention.
The opinion also mentions a withdrawal decision being annulled if a national authority (usually the highest court) finds it unconstitutionally adopted. In this scenario, the notifying state must communicate the revocation, as the initial decision lacked a crucial prerequisite.
Similarly, actions aligned with constitutional requirements (e.g., a referendum, a significant parliamentary vote, a general election yielding an opposing majority) could reverse an initial decision by nullifying its judicial and constitutional basis. Again, the state must notify the European Council of this change. Such situations have international precedents and align with Article 68 of the VCLT.
Furthermore, the opinion argues that forcing a Member State to leave under such conditions would defy common sense. Accepting revocation respects a national parliament’s role in a Member State’s identity. Rejecting revocation amounts to a forced EU exit.
Applying to rejoin contradicts this interpretation, as spending the Article 50 period negotiating future membership is illogical. As the CJEU noted, a Member State remains so during the Article 50 period.
The opinion then refuted the Commission and Council’s counterarguments. The Advocate-General believes a unilateral element remains after notification, as the withdrawing state can refuse to negotiate a withdrawal agreement. The power to extend the negotiation doesn’t imply, conversely, no power to withdraw the notification exists. The Council’s cost argument is unconvincing, as the financial implications of withdrawal for the EU and its citizens would far exceed the minimal costs of revocation.
Unilateral revocability, the opinion argues, strengthens the Treaties’ “ever closer union” clause, Member States’ national identities, and EU citizens’ rights. The historical context of Article 50’s drafting supports this view.
However, conditions apply. A formal revocation notification, mirroring the withdrawal notification, is necessary. National constitutional requirements, determined by each Member State, must be met. The UK’s requirement of parliamentary authorization to leave logically implies parliamentary approval for revocation. No justification for revocation is needed.
A time limit exists: revocation is possible only within the two-year negotiation period following the withdrawal notification. Once the withdrawal agreement is concluded, implying both parties’ agreement, revocation is impossible. This contradicts the opinion’s earlier mention of extending the Article 50 timeframe.
Article 4(3) TEU’s principles of good faith and sincere cooperation set a limit. The EU institutions worried a Member State might revoke and halt unfavorable negotiations, then resubmit the withdrawal notification, triggering a new two-year period and circumventing the time limits. The Advocate-General disagrees: the potential for abuse doesn’t negate the right’s existence. A way to prevent abuse is needed. The EU law principle of “abuse of rights” could apply if a Member State abuses successive notifications and revocations to improve withdrawal terms.
There’s no indication of planned abuse, and it could only occur with a second notification, not a unilateral revocation. The Advocate-General finds numerous “tactical revocations” unlikely, as national “constitutional requirements” act as a filter, potentially requiring elections, court decisions, or referendums.
Finally, the opinion allows for agreed revocation if the departing Member State requests it and the European Council unanimously agrees.
Comments
Firstly, the opinion convincingly argues that the case is not hypothetical. It rightly cites case law presuming the relevance of national court referrals, rebuttable only in limited instances. The status of MPs as petitioners strengthens the argument for addressing these questions. Otherwise, a catch-22 arises where MPs hesitate to vote on holding a referendum without knowing if unilateral revocation is possible.
The main argument for unilateral revocability, particularly emphasizing the word “intention” and Member States’ sovereign decision-making power, is compelling. Regarding safeguards against abuse, national constitutional requirements may not always filter effectively. Explicit limits—perhaps refusing to consider a notification within a year or two of the previous one, or resuming an Article 50 process from where it left off—would be preferable.
The opinion’s vagueness on the time limit is unfortunate. Is notification possible only within the two-year timeframe, or also during an extension? The opinion is contradictory, which the Court’s judgment (if aligned with the opinion) should avoid. The possibility of a delayed timeframe in the withdrawal agreement also warrants mention. Logically, the opinion suggests revocation should be possible within the extended period.
Revocation with consent doesn’t fit neatly within Article 50’s wording, as it lacks a consent process. A Member State likely wouldn’t choose this option if unilateral revocation is possible. However, it might apply if the Member State’s existing membership is being renegotiated, a scenario only Tony Blair seems to envision for the UK.
Overall, if the Court follows suit, this opinion is a complete victory for unilateral revocation proponents. Whether it impacts the political landscape is uncertain. Another referendum (which Remain would still need to win) remains unlikely. Nevertheless, unilateral revocability could still be relevant for other Member States considering leaving the EU.
Barnard & Peers: chapter 27
Photo credit: The Irish Times