TODAY:

Brexit. A bridge to nowhere?


The Brexit transition period: analysis and annotation.


Last week saw significant developments in the Brexit talks. On Monday March 19th, the EU27 and the UK agreed on a large part of the proposed withdrawal agreement, most notably the details of a transition period (lasting from Brexit day in March 2019 to the end of 2020) which the UK was particularly concerned to agree. (There were earlier drafts of the entire agreement on February 28 and March 15). Subsequently, on Friday March 23rd, the EU27 decided that there was sufficient progress in the talks to move to adopt its guidelines for the future relationship with the UK, which will take the form (for now) of a declaration linked to the withdrawal agreement. Ultimately, that future relationship will be regulated by separate treaties ratified after Brexit day.

For now, though, I will focus on the core issue of the transition period, since it attracted a lot of controversy during the week, in particular as regards fisheries and trade issues. I’ve previously annotated a first draft of the transition period rules, as proposed by the Commission, and then updated that in light of the UK response to that draft. These followed on from the negotiation guidelines on the transition period adopted by the EU Council in January, which I annotated here.

Overall, the UK has obtained some concessions during these talks: explicit powers to hold treaty talks with non-EU countries; the possibility to abstain on controversial EU foreign policy measures which it would have vetoed as a Member State; some greater consultation rights; an attempt to consider the UK still part of treaties with non-EU states for the time being; and protection against any grab of fisheries catches in UK waters. However, the UK government had to concede on issues which it had placed much stress on: taking fisheries catch issues out of the transition period and not extending all acquired rights to EU27 citizens who came to the UK during this period.

It seems likely that the UK government was concerned about the economic impact of not agreeing a transition period a year before Brexit day; in the event, that prospect has been averted. But then again, the transition period is only one part of the withdrawal agreement, and some controversial issues in the rest of that agreement have yet to be agreed; the agreement would also then have to be ratified. So we can’t be certain that the transition period will apply in practice yet.

For those concerned about the transition period making the UK a “vassal state”, it should be noted that some Leaver supporters had long advocated an interim period after EU membership based on being a member of the EEA, which would have given the UK more rights to consultation over EU legislation, less coverage by new EU laws, the possibility of refusing to apply new laws and a different court to apply them in a less mandatory way. So such critics might wish to ask why the UK government did not pursue such an approach to the transition period – or explain why they have never bothered to present a serious alternative proposal of their own (besides, obviously, not continuing with Brexit at all).

The annex to this blog post is a detailed annotation and analysis of the agreed rules on the transition period in the withdrawal agreement. Before that though, I look in detail at two other issues: how the specific provisions on the transition period fit into the agreement as a whole; and what happens if the two sides want to extend the transition period, given that it might be difficult to agree all aspects of the UK/EU future relationship by the agreed deadline of end-2020.


Structure of the withdrawal agreement

The recent deal on the transition period, which also includes agreement on a large part of the rest of the withdrawal agreement, can only be understood fully in the context of the rest of that agreement, which I’ll summarise here.

Part One of the withdrawal agreement (Articles 1-7) sets out the “Common Provisions”. These deal with basic issues like definitions and territorial scope. This Part is agreed except for the rules on interpretation of the agreement (Article 4).

Part Two of the withdrawal agreement (Articles 8-35) deals with citizens’ rights. It is entirely agreed. It will mostly not apply until the end of the transition period, and as part of the recent agreement, the UK government has accepted that EU27 citizens arriving in the UK during the entire period will retain the same rights as those who arrived before Brexit day.

Part Three of the agreement sets out “separation provisions” (Articles 36-120). These deal with issues like what happens to European Arrest Warrants pending at the end of the transition period. This part will largely apply as from the end of that period. It has been partly agreed, as follows: issues of goods on the market, public procurement, and Euratom: mostly agreed; pending criminal and police proceedings, data protection: mostly not agreed; civil litigation: conflict of laws points agreed, but jurisdiction and mutual recognition points not yet agreed; intellectual property rights: mostly agreed; pending ECJ cases and administrative proceedings (like competition investigations): not agreed; and other institutional issues (like privileges and immunities): agreed.

Next, Part Four sets out the rules on the transition period as such (Articles 121-126). These provisions are annotated below. As noted already, this part is fully agreed.

Part Five concerns the financial settlement (Articles 127-150). It incorporates the earlier agreement that the UK is takes part in the EU’s financial rules until the end of the transition period. (The end-2020 cut-off date of the transition period in Article 121 matches the end of the current EU budget cycle). This part is also fully agreed.

Part Six sets out “Final Provisions” (Articles 151-168). This part is partly agreed. For instance, the rules on ECJ jurisdiction over EU citizens (which applies, for the most part, for eight years after the end of the transition period) is agreed (Article 151). So is the clause on a new authority to help guarantee EU citizens’ rights, which might be shut down at the end of that same eight-year period (Article 152). The UK and EU have agreed to continued ECJ jurisdiction over EU budget law rules referred to in the financial settlement part, but not yet to that Court’s continued jurisdiction over "separation rules" (Article 153). The existence of a Joint Committee to monitor the agreement and take some decisions to implement it is agreed (Articles 157-159), but the rules on dispute settlement (Articles 162-165) – some of which would apply during the transition period and some of which would apply from after that point – are not.

There will also be Protocols on Irish border issues and UK bases in Cyprus, both of which would apply from the end of the transition period; the latter has not yet been proposed. The parties have agreed that there will be a "backstop" text on Irish border issues, which will apply if no other solution can be found. This is a controversial issue, but the text has been partially agreed: the Common Travel Area (fully agreed); equality rights (agreed as a policy); movement of goods issues (not agreed); but electricity market and state aids clauses agreed as a policy; a vague reference to other North/South cooperation is agreed in principle; and some institutional issues are agreed.

Overall, then, a substantial proportion of the withdrawal agreement is already agreed in principle. The main issues outstanding – in increasing order of political difficulty – are parts of the separation provisions, the general rules on dispute settlement, and Irish border issues.

Note that I have separately annotated the withdrawal agreement provisions on: the common provisions and citizens’ rights; the final provisions (mainly focussing on dispute settlement and the role of the CJEU); and the Irish border. (These annotations concern earlier drafts of the agreement, although there are not vast changes in the March 19th version; I hope to prepare an updated annotation of the entire text eventually).

Extending the transition period?

As noted above, it’s widely believed that at least for some issues, it may prove necessary to extend the transition period, since negotiations on the future EU/UK relationship might not be complete by the end of 2020. Leaving aside the politics of the issue, is that legally possible?

First of all, a procedural point: it would arguably be possible to settle this (and other legal questions about the withdrawal agreement) in advance, by using the special power to ask the ECJ to interpret a planned agreement between the EU and a non-EU country in advance (see Article 218(11) TFEU). While the UK is not now a non-EU country, it will be when the withdrawal agreement applies, and while Article 50 TEU does not refer to Article 218(11), such a quibble is excessively pedantic: the ECJ has always interpreted its jurisdiction on this point widely, and the point of asking it questions on the withdrawal agreement would be consistent with the established purpose of Article 218, which is to avoid in advance complications which may arise from legal problems which may arise in a future agreement between the EU and a non-EU country. This reasoning applies a fortiori if one of the questions is whether the withdrawal agreement can even be amended after Brexit day, in part or whole: if it can’t, then the complications which may arise will be greater than ever.

Although the Court has to give its rulings in such cases before a treaty enters into force (the withdrawal agreement is set to come into force on 30 March 2019: Article 168), and its judgments usually take over a year to decide, it can fast-track rulings into four to six months if necessary. However, since any adverse ruling by the ECJ would mean that the draft withdrawal agreement would have to be amended to address the Court’s concerns, and in any event there would need to be time for ratification of the agreement after the Court’s judgment, the clock is ticking if the Court is going to be asked to rule on the draft.

Moving to the substantive issue, can the withdrawal agreement be amended after Brexit day, and if so, how? The drafting of Article 50 TEU is not explicit on this point, but it seems to set out special rules only relating to a state in the process of withdrawal – not a state which has already left. So Article 50 can’t apply after Brexit day to amend the withdrawal agreement. At first sight, though, other EU treaty provisions could be the basis of an EU power to amend the withdrawal agreement after Brexit day, with the mutual agreement of the UK. This might lead to a more difficult process of amendment, since Article 50 provides for the EU only as a party to the withdrawal agreement, not its Member States, as well as qualified majority voting, rather than unanimous voting; and there is a good chance that requirements of unanimous voting and national ratification would apply after Brexit day.

But at least it would be possible – or would it? Some have argued not, for two reasons. The starting point is the rule in the Vienna Convention on the Law of Treaties (Article 39), which says simply that the parties to a treaty can amend it using the general rules on negotiating a treaty, “except in so far as the treaty may otherwise provide.” The withdrawal agreement does not explicitly rule out amendments, or explicitly provide for them either, except for a few amendment powers for the Joint Committee which it sets up to amend it. Does that mean a contrario that the parties can’t amend it otherwise? Such an interpretation is not EU practice: for instance, the Association Council set up to implement the EU/Turkey Association Agreement has very broad powers, yet he EU and Turkey have also agreed Protocols to that treaty in practice.

The second argument: given that the Article 50 powers expire on Brexit day, should it necessarily follow that the EU is prohibited from using other powers to amend the withdrawal agreement after Brexit day? In the absence of express wording to that effect, this argument that the agreement must be legally “frozen” is untenable. The text of the agreement already specifies that other treaties can replace it as regards foreign policy and the Irish border, and it would be odd if the parties could not choose to amend it otherwise even on a minor point – for instance, to take a different (or supplementary) approach to how to deal with European Arrest Warrants pending at the end of the transition period, or to add rules on how to deal with issues that were left out of the withdrawal agreement because the parties forgot about them or couldn’t agree on them at the time. For example, the draft agreement has no “separation” rule on what happens to the transfer of responsibility for asylum applications which are pending at the end of the transition period.

To some extent, the dynamism of the agreement itself inherently raises the possibility that new legal issues will arise which weren’t foreseen when it was signed – given that the UK will be bound by new EU laws which are adopted during the transition period, and even earlier laws which are adopted between the signature and entry into force of the withdrawal agreement. (There might be quite a few of the latter, since Brexit day coincides with the end of a European Parliament term, when a large batch of outstanding legislation is usually agreed). The post-signature case law of national, EU and international courts might raise issues that the parties might choose to take into account too.

However, this brings to another problem. Any prospect of providing in the withdrawal agreement itself for an indefinite renewal of the transition/implementation period could raise legal issues as to whether this falls beyond the scope of Article 50 TEU, given that (on the EU’s interpretation, at least) Article 50 cannot serve as the basis for a permanent future relationship with the EU. Amending the withdrawal agreement post-Brexit to the same effect might arguably run into the same problem.

Could this issue be avoided entirely, by simply adopting a new EU/UK treaty post-Brexit, which simply includes the transition period rules in the withdrawal agreement? Those rules could, for instance, constitute the first phase of the future EU/UK relationship in a post-Brexit association or partnership agreement, being replaced by more permanent rules whenever those are agreed. There’s an argument that this would be “going too far” as compared to what the EU treaties permit as regards a permanent relationship with a non-EU state, but it seems little more significant than the EU/Turkey agreement discussed above, which provides for Turkey to fully join the customs union, internal market and common agricultural policy.

Whether either one of the parties could denounce the withdrawal agreement is a different question – which could be particularly relevant for EU27 citizens living in the UK, or those concerned about the Irish border. Denunciation of treaties is regulated by Article 56 of the Vienna Convention, which provides that:


  1. A treaty which contains no provision regarding its termination and which does not provide for denunciation or withdrawal is not subject to denunciation or withdrawal unless:

(a) It is established that the parties intended to admit the possibility of denunciation or withdrawal; or

(b) A right of denunciation or withdrawal may be implied by the nature of the treaty.


It is hard to see how either of these criteria is satisfied here. There’s no express clause on denunciation and the parties contemplate the agreement lasting some time, not only eight years after the end of the transition period (Articles 151 and 152) but for the lifetime of the citizens protected by Part Two (Article 35). That same clause points to the nature of the treaty being indefinite – otherwise the citizens’ rights protected by Part Two would be undermined. Equally there is no final time limit on the Irish Protocol, with only the possibility of replacing it by an alternative set of rules.

Finally, it’s worth noting that Article 50 expressly provides for the alternative of extending EU membership – although the UK government does not seem interested in that idea. While that would give the UK a full role in the EU while talks continue, it would not avoid the EU side’s belief that there is a legal problem negotiating fully on the future relationship as long as the UK is a Member State. There is also the possibility that the withdrawal agreement has a different date of entry into force than the two-year default set out in Article 50; but again the UK government may not be enthused about staying in the EU longer.

Both solutions would raise the political difficulty of the UK participating in the next elections to the European Parliament. While some argue that the UK could simply be exempted from this, I doubt the legal validity of excluding any current EU Member State from participation in EU elections; and such a decision would surely be vigorously litigated by those who wanted to vote.


Finally, during the transition period any attempt to rejoin the EU would be subject to the normal rules on new Member States’ accession, as noted in Article 50(5) TEU. These could well be fast-tracked for a country which already applies most EU law (as in the example of Iceland), but the question will arise whether the UK’s opt-outs and budget rebate will still remain. The issue here is whether they will automatically expire on membership, or whetehr the Treaties and legislation will have to be revised to remove them. (The exemption from Schengen should necessarily stay in any event, as it’s necessarily linked to the Irish exemption on the same issue, due to the Common Travel Area).🔷


Barnard & Peers: chapter 27






Embed from Getty Images


(This piece was first published on EU Law Analysis.)


(Cover: Pixabay.)


Author image
Professor of EU, Human Rights & World Trade Law, University of Essex. Latest book: European Union Law (edited with @CSBarnard24, 2nd ed, OUP).
London, UK Website

About us | Our Writers | Disclaimer | Terms & Conditions | Privacy Policy | Submissions | Code of Conduct


No part of this publication may be reproduced or used without the express permission of the publisher.