Pietro Manzini, University of Bologna

1. Even if it has caused great displeasure to enthusiasts of the current mounting populism, it is now clear that the UK Parliament will play a crucial role in the so-called Brexit procedure. As it is already well known, the UK Supreme Court has confirmed that an act of British Parliament is required to authorise ministers to give notice of the decision of the country to withdraw from the European Union, pursuant to Article 50 (1) TEU (see, in this regard, UK Supreme Court, 24 January 2017, Judgment (on the application of Miller and another) (Respondents) v. Secretary of State for Exiting the European Union (Appellant)). At the time I write the bill to give such authorization has been passed in House of Commons and the final word of the House of Lords is expected at the beginning of March.

From the EU perspective, the recognition of the central role of the Parliament amounts to an increased possibility that, after giving notice of withdrawal, the UK may change its mind on Brexit. In the Miller case (quoted above) both the UK High Court in the first instance and the UK Supreme Court on appeal started from the premise – common to the applicant and the respondent – that the notification pursuant to paragraph 1 of Article 50 is irrevocable. However, on the one hand, the alleged irrevocability of the notification does not preclude any future “change of one’s mind” based on domestic law, and on the other, the only institution that may finally decide on this issue is the Court of Justice, and not national institutions. This is why I believe, following opinion already expressed (see, Miglio), that the Supreme Court should have referred the question of revocability of the notification to the Court of justice. Even if, as a matter of judicial policy and of political expediency, the reason why it did not is perfectly understandable.

Amid the political turmoil triggered by the Brexit referendum, the parliamentary majority may well change in UK during the two years’ time set out by Article 50 to conclude the exit procedure; or, after considering the results of the negotiation with the EU, Parliament may decide to have a second referendum on the UK’s membership of the EU, which may well have the opposite result to the one held last year.

Therefore a decisive question may actually arise: can the withdrawal notice be revoked, or does it mean forever?

Article 50 provides that a Member State may decide to withdraw from the Union by notifying the European Council of its intention. However, it contains nothing on the issue of the State’s right to revoke the notice before the withdrawal actually takes place. If the revocation is admissible, the State concerned would remain in the EU, since paragraph 3 of Article 50 declares that the EU treaties shall continue to apply to the withdrawing State until the entry into force of the withdrawal agreement or, failing that, the expiry of the two-year period provided in the same provision.

In this respect it must be clear that a revocation of the notice should not create particular problems if it is agreed to by the State in question and the European Council with a unanimous decision. To this purpose two authors have argued that the negotiation provided for in paragraph 2 of Article 50 could end up, instead of an agreement to leave the EU, with an agreement not to leave the EU, or, as an alternative, with a common decision to extend the negotiation indefinitely, and, eventually, the adoption of a Protocol stating that the notification of withdrawal has been revoked. I do not find these arguments wholly convincing. Indeed, the first argument contradicts the terms of Article 50, paragraph 2, which provides that the aim of the negotiation is «setting out the arrangements for the withdrawal», and it does not foresee that the negotiation can be concluded without withdrawal. Differently, the second is contrary to the intention of Article 50, paragraph 3, which is to allow parties to extend the negotiations in order to reach a better or a more comprehensive withdrawal agreement. Accordingly, only on a basis of an evident misinterpretation of Article 50, it is possible to argue that the European Council and the State in question may decide to indefinitely extend the negotiations with the consequence that the withdrawal never happens.

Alternatively, it is my own belief that the revocation of the withdrawal notification agreed by the European Council and the State concerned could be admitted, since it is possible to qualify it as an agreement on the interpretation and application of the EU Treaty, in accordance with Article 31, paragraph 3, of the Vienna Convention on the Law of Treaties. This solution would be in accordance with international customary law, and – contrary to those mentioned above – would not be in conflict with the terms and the purpose of Article 50, since it would simply clarify its scope, filling a lacuna of its provisions.

2. The unilateral revocation of the notice of withdrawal, decided by the State concerned, is admittedly far more troublesome. In this case, by definition, no interpretative agreement between the parties is concluded, and therefore in the light of the terms and purposes of Article 50, the most straightforward solution would seem to deny the possibility of revocation. This would respect the principle according to which ubi lex voluit dixit, ubi noluit tacuit, which in these circumstances maintains its heuristic value.

Nevertheless, three lines of reasoning have been put forward in order to uphold the possibility to revoke the notice of withdrawal, even unilaterally. The first refers to international treaty law; the second is grounded on a parallelism between the withdrawing procedure laid down in paragraph 3, of Article 50, and an ordinary contractual negotiation; and the third is based on a teleological interpretation of paragraph 1 of Article 50.

Anticipating my conclusions, I would say that I find the first two lines of reasoning unconvincing, while the third one seems more workable, even if it needs a number of qualifications.

The first, and weakest, line of reasoning relies on the possibility of revocation provided for by Article 68 of the Vienna Convention on the Law of the Treaties, which states that the notification of the withdrawal (or other instruments of international law) «may be revoked at any time before it takes effect». It is argued (Craig 2016a, Miglio 2016) that this provision, considered as a customary rule and, thus, applicable to the European Union, would support the idea that a unilateral revocation of the decision to leave the Union is admissible within the scope of Article 50.

There are a number of arguments against this opinion.

Firstly, the provision concerning the withdrawal provided for in the Vienna Convention shall apply only where the treaty does not specifically regulate the withdrawal. This can be derived from Article 56 of the same Vienna Convention, which states that a treaty that contains no provision regarding its termination and which does not provide for denunciation or withdrawal is not subject to denunciation or withdrawal, unless it is established that the parties intended to allow these possibilities, or a right of denunciation or withdrawal may be implied by the nature of the treaty. Therefore, it seems rather inappropriate to interpret Article 50, which is a specific provision on withdrawal, by using a rule applicable only where such kind of provisions actually does not exist in the treaty. Moreover, the reference to the Vienna Convention could be used in the opposite way: it can be asserted that Article 68 demonstrates that when States intend to provide the possibility to revoke withdrawal, they do so explicitly. By consequence, in the event that this possibility is not provided for in the applicable provision, it is not acceptable to derive it by interpretation. Obviously, the fact that Article 68 can be used to reach opposite conclusions greatly weakens its interpretative value.

The above point notwithstanding, the reference to the rules on revocability contained in the Vienna Convention appears rather unfeasible in the context of the European Union. It should be noted that these rules do not have any elements in common with provisions of Article 50 (in this regard I share the view already expressed by Munari). The aim sought to be achieved by the withdrawal procedure of the Vienna Convention – more specifically Articles 65 to 68 – is primarily to resolve possible disputes between parties about the interpretation or application of the treaties, and not to ensure, for each Member State, the unilateral possibility to withdraw from the Union, which constitutes the fundamental rationale of Article 50. This can be derived, for instance, by the fact that, in accordance with Article 65 of the Vienna Convention, when a State intends to withdraw from a treaty, it must notify the other parties of its claim and the notification must indicate the reasons thereof. If, after a period, which cannot be less than three months after the receipt of notification, no party has raised any objection, the State making the notification may withdraw. However, if an objection has been raised by any other State, the parties must seek a solution through the means indicated by Article 33 of the United Nations Charter. Article 66 of the Convention provides, in this regard, that if no solution is reached within a period of twelve months following the date on which the objection was raised, other procedures shall be undertaken in order to resolve the dispute, in particular judicial settlement, arbitration or conciliation. It is in this context that Article 68 states that the notification may be revoked: this possibility is based on the positive resolution of the dispute triggered by the withdrawal notice and by the subsequent objection raised by one or more parties of the treaty.

Article 50 does not provide for any of this. The notice of withdrawal does not need to state the reasons for the State’s decision; no Member State can object to the withdrawal and no procedure is set out to resolve the possible dispute. Accordingly, it is clearly inappropriate to refer to the possibility to revoke the notification provided for in the Vienna Convention in order to “read” an analogous possibility in Article 50 TEU, which, as noted, says nothing on the point.

3. A second line of reasoning used to admit a unilateral revocation of the notice of withdrawal within the framework of Article 50 relies on a possible parallelism between the withdrawal procedure laid down in particular in paragraph 3 thereof and the course of an ordinary contractual negotiation. The argument runs as follows: in any negotiation any party may decide to withdraw until the agreement or the contract is concluded; analogously, before the conclusion of the withdrawal agreement or the expiry of the two-year timeframe provided for in paragraph 3, the State in question must be permitted to leave the negotiations by revoking its notice; since paragraph 3 establishes that the treaties are applicable to the State until the entry into force of the withdrawal agreement or until the two-year time period has elapsed, the “disappearance” of the notification implies that State has no intention of leaving the Union anymore (see on that Craig 2016b).

Even though the author specifies that his argument should not be interpreted as a way to read into Article 50 something that is not there, I think that, in substance, this is precisely the consequence of this kind of reasoning. To be sure, paragraph 3 merely deals with the moment when the treaties cease to apply to the withdrawing Member State. Neither the text nor the goal of this provision offers the possibility to derive, even implicitly, the right to revoke the notification.

Moreover, it is to be noted that the legal framework arising out of the notification of withdrawal is very different from the background of ordinary contract negotiations. In this kind of negotiations the possibility for each party to interrupt the process and to avoid the conclusion of the contract is perfectly understandable and legitimate.

Firstly, contract negotiations are characterized by the will of the parties to reach a satisfactory result for all. Therefore, it is logical to admit that, in the event this is not possible, any party can refuse to continue. Differently, the notification pursuant to Article 50 does not express the will to conclude an agreement suitable for both parties, but it declares the intention to unilaterally interrupt an agreement already in force. This obvious consideration already clarifies that it is rather simplistic to argue that, in the context of Article 50, general principles governing contractual negotiation, such as the right to change one’s mind, apply.

Secondly, the negotiation for the conclusion of an ordinary contract or agreement does not have, as a prerequisite, either a formal and public decision of one of the parties to reach a specific aim, or the possibility for such a party to unilaterally reach the aim pursued with its decision. Conversely, these are the elements characterizing the withdrawal procedure of Article 50, which, as we have seen, foresees the possibility for the Member State to unilaterally decide to leave the Union and the possibility that the withdrawal from the Union is automatically achieved after two years from the notification, irrespective of the conclusion of an agreement between the State concerned and the Union. These elements generate in the Union and its Member States a strong expectation as regards the outcome of the withdrawal procedure, which does not normally exist in the framework of the ordinary negotiation of agreements or contracts. Within these contexts, indeed, except in special circumstances, no one could have a legitimate expectation that the contract will be effectively concluded. In addition, it should be noted that under such interpretation of Article 50, the right to “change its mind” through the revocation of the notification would be only in the hands of the withdrawing Member State and, by contrast, the Union would not benefit from it, since it could not in any way unilaterally block the withdrawal procedure.

Therefore, only through a very creative interpretation of paragraph 3 of Article 50 and on the basis of an artificial parallelism between the scope of the contract negotiation and the legal framework set up by the withdrawal procedure, it could be concluded that the unilateral revocation of the notice of withdrawal is admissible.

4. Lastly, a third line of reasoning has been proposed in favour of the unilateral revocability of the notification. This reasoning seems more convincing than the others, however, caution is needed.

The argument goes as follows. Paragraph 1 of Article 50 is the only provision that allows the State to decide withdrawal unilaterally; to be specific, it confers to each Member State the right to withdraw from the EU, provided that this decision is taken in accordance with its own constitutional requirements. In the event that the decision to leave is followed by a decision to remain in the EU, the notice of withdrawal would not reflect the constitutional arrangements of the State, and therefore it would not be in compliance with paragraph 1 of Article 50 (this assumption is developed by Eeckhout and Frantziou 2016, but it is also suggested by Craig 2016b). Accordingly, the State in question should be permitted to revoke the notification. It may be added that this interpretation of the provision concerned is supported by a teleological approach, since the Treaties intend to promote European integration and to create an ever closer union among the peoples of Europe, and thus their rules should be interpreted in accordance with these objectives.

As mentioned before, caution is called for when looking at this argument. On the one hand, the State concerned may abuse the right of revocation (Eeckhout and Frantziou 2016), i.e. it could trigger the withdrawing procedure in order to negotiate preferential conditions for its membership with the Union, and it could decide to revoke or not revoke notification according to the outcome of the negotiations. This approach could be replicated by other Member States with disruptive consequences for the Union.

On the other hand, it should be considered that, as it is evident from Article 50, paragraphs 2-4, once the withdrawal procedure is triggered in accordance with paragraph 1, the membership of the State ceases to be a simple matter of domestic law and it becomes an issue covered by Union law. This law is inspired, inter alia, by the principle of good faith in the relations among States from which the corollaries of legal certainty and legitimate expectation arise. In the light of these corollaries, Member States shall apply and interpret the treaties to ensure the stability and foreseeability of the relations with the other parties as much as possible; they shall respect their commitments once they have made them in a way that generates legitimate expectations in the other Member States. The European Parliament has implicitly, but nonetheless clearly, taken a similar stance in its Resolution on the UK’s decision to leave the Union. In the days following Britain’s vote, the EU Parliament invited the UK to notify the withdrawal pursuant to Article 50 as soon as possible «in order to prevent damaging uncertainty for everyone and to protect the Union’s integrity». It is now clear that if damaging uncertainties are likely to be generated by a mere delay in the withdrawal notification, they would be caused a fortiori by a revocation of the notice of withdrawal, once this notice has been formally submitted to the Union. There is no doubt that a withdrawal decision taken consistently with the constitutional requirements of a Member State raises for the Union and for the other Member States the legitimate expectation that such a decision has been seriously and carefully taken and, therefore, it will be maintained.

In light of the above, the unilateral revocation of the withdrawal could be admitted, on the condition that abuses are avoided and that the principles of legitimate expectations and of legal certainty are respected. According to the opinion of those that have supported the argument in question, in order to protect the first instance, the revocation decision must be taken in good faith. I am ready to accept this position as long as the meaning of revocation in “good faith” can be reasonably clarified. To this regard, I think that four conditions must be met: a) the revocation decision shall be taken in compliance with domestic constitutional requirements; b) in order to avoid the risk of abuses, no prior revocation can be attributed to the State concerned; c) the revocation should be unconditional, i.e. the State in question shall return to its original position in the Union; d) lastly, the revocation shall be made within a reasonable time, namely not close to the end of the withdrawal negotiations when the results are already foreseeable.

As regards the principles of legal certainty and legitimate expectations, it is worth recalling that the European Court of Justice, in Cargill v. Commission, ruled that any Community institution has the right to withdraw a measure tainted with illegality within a reasonable period with retroactive effect. However, that right may be restricted by the need to fulfil the legitimate expectations of a beneficiary of the measure, who has been led to rely on the lawfulness thereof. These conditions should apply, mutatis mutandis, to the revocation of the withdrawal notice; accordingly, the revocation should be timely and it should provide the restoration of the legal position formerly enjoyed by those who relied on the fact that the withdrawal decision was permanent.

To sum up, the values of European integration and of an ever closer union among the peoples of Europe could lead to an interpretation of paragraph 1 of Article 50 that allows the unilateral revocation of the withdrawal notice, despite this provision having a different scope and objective. Let’s be honest: it would be amusingly ironic that the only way the UK could return to the EU family would be to invoke the main reason it left, namely the principle of an ever closer union.