Does UK Law Require a Referendum on the EU Withdrawal Agreement?

Donal Coffey

Most people will throw their hands in the air when confronted with the possibility of another EU referendum in the United Kingdom. This article does not argue in favour or against such a referendum, but asks the following question: From a legal point of view, does a withdrawal agreement concluded between the United Kingdom and the European Union require popular approval by referendum in the UK?

Does UK Law Require a Referendum on the EU Withdrawal Agreement?
Prime Minister David Cameron, UK Parliament, CC-BY-NC-2.0

The basis for this question is the European Union Act 2011, offered by David Cameron to Eurosceptics in his Conservative parliamentary party in order to prevent further transfers of power to the EU. This act is currently in force but is due to be repealed under Schedule 9 of the European Union (Withdrawal) Bill which, having passed the House of Commons, is currently debated in the House of Lords.

European Union Act 2011

A treaty falling within the scope of act cannot be ratified unless a statement is laid before Parliament within certain time frames (for our purposes, within two months), the treaty is approved by statute and either the ’referendum condition‘ or the ’exemption condition‘ applies. Under the referendum condition, a treaty cannot be ratified unless a referendum has been held in the UK (or the UK and Gibraltar, if Gibraltar is also affected), and a majority of those voting in the referendum approve ratification. Under the exemption condition, the treaty does not fall within Section 4 and a referendum is not required.

In that regard, the 2011 act establishes a two-part test for whether a referendum must be held. First, the treaty must ‘amend or replace’ the Treaty on European Union (TEU) or the Treaty on the Functioning of the European Union (TFEU). In the recital to the act, it refers to ‘treaties relating to the European Union’. Second, the treaty must meet one of the referendum conditions set out in Section 4 of the act. Exemption conditions include treaties enabling the accession of a new Member State.

For our purposes, the relevant parts of Section 4 are conditions 1(i) and 1(j):

(i) the conferring on an EU institution or body of power to impose a requirement or obligation on the United Kingdom, or the removal of any limitation on any such power of an EU institution or body;

(j) the conferring on an EU institution or body of new or extended power to impose sanctions on the United Kingdom.

It is therefore necessary to consider whether or not a withdrawal agreement concluded under Article 50 would satisfy these two conditions. For the purposes of this piece, it is assumed that the withdrawal agreement is concluded between the UK and the EU under Article 50 and deals with the conditions of the UK’s withdrawal from the EU.

Article 50 and ‘Amend or Replace’

The first question is whether or not concluding a withdrawal agreement constitutes an amendment of the TEU or TFEU. The operative sections of the TEU relating to the withdrawal agreements are contained in the now-famous Article 50:

2. A Member State which decides to withdraw shall notify the European Council of its intention. In the light of the guidelines provided by the European Council, the Union shall negotiate and conclude an agreement with that State, setting out the arrangements for its withdrawal, taking account of the framework for its future relationship with the Union.…

3. The Treaties shall cease to apply to the State in question from the date of entry into force of the withdrawal agreement or, failing that, two years after the notification referred to in paragraph 2, unless the European Council, in agreement with the Member State concerned, unanimously decides to extend this period.

These should be read in conjunction with Article 52, which states, in relevant part:

The Treaties shall apply to… the United Kingdom of Great Britain and Northern Ireland.

There are three possible interpretations of how these provisions should interact. First, the machinery outlined under Article 50 TEU might serve to automatically amend Article 52 TEU, without the need for any intervening legal action. Second, Article 52 might itself need to be amended through the withdrawal agreement. Third, Article 52 might need to be amended by a separate treaty consequent on or contemporaneous with the withdrawal agreement.

Of these possibilities, the most difficult to read harmoniously with the provisions of Article 50 is the third. The UK’s triggering of Article 50 and subsequent departure from the EU are sufficient grounds for the remaining Member States to amend Article 52 directly. Failure to conclude a separate treaty would not mean that Article 52, and therefore EU law, would continue to apply to the UK after Brexit. This does not mean that the European Union might not convene an intergovernmental conference to amend the treaties around the UK’s departure. It simply means that this is not legally required under Article 50 and, from the UK’s point of view, the treaties will cease to apply from the date of entry into force of the withdrawal agreement.

The second interpretation appears to be hinted at in the European Parliament February 2016 briefing on Article 50:

Unlike the accession of new Member States to the EU, the withdrawal of a Member State does not require ratification by the remaining Member States – Article 50(1) TEU mentions (in a declaratory way) only the decision of the withdrawing state, in accordance with its constitutional requirements. However, any treaty changes or international agreements (such as a free trade agreement) that might be necessary as a consequence of the withdrawal agreement would need to be ratified by the remaining Member States in accordance with Article 48 TEU. At the very least, Article 52 TEU on the territorial scope of the treaties, which lists the Member States, would need to be amended, and protocols concerning the withdrawing Member State revised or repealed.

Although this might appear to be consistent with the third interpretation above, the briefing appears to suggest a process involving the European Parliament, European Council and the remaining Member States which is more consistent with the second option. In these circumstances, the withdrawal agreement itself would stipulate the changes to be made to the treaties, including, for example, the levels of capitalisation of the European Investment Bank. As it would clearly necessitate treaty adjustments, the withdrawal agreement itself could amend the Treaty on European Union.

This seems the most elegant legal method of constructing the withdrawal agreement, although treaty change would involve the Article 48 TEU procedure, which is not explicitly mentioned under Article 50 and could introduce further political complexity to the matter. It does, however, have the virtue of consistency with Article 7 TFEU: ‘The Union shall ensure consistency between its policies and activities, taking all of its objectives into account and in accordance with the principle of conferral of powers.’ It is also unlikely that Article 52 will remain unamended indefinitely, and it seems preferable to change this sooner rather than later.

The first interpretation is also possible: Article 50 is self-executing and serves to automatically amend Article 52. Indeed, the default two-year time limit is only sustainable on such a theory of Article 50 – it is specifically not based on an intervening act. The conclusion of a withdrawal agreement, however, changes the application of Article 50(3) from the two-year default limit to the withdrawal agreement. In this way, the first interpretation also means that a withdrawal agreement amends the TEU.

The amendment of Article 52 TEU is dependent on the interaction of Article 50(3) and the withdrawal agreement: absent a withdrawal agreement, amending Article 52 takes place in a different fashion. It can be argued in contrast that Article 50 stipulates the end irrespective of any agreement (on withdrawal) and therefore the agreement itself does not amend Article 52. Under this argument, the amendment of Article 52 is carried out entirely by virtue of Article 50 and, as the withdrawal agreement does not amend Article 50, it does not amend the TEU and thus, returning to the question of UK law, falls outside the 2011 act.

This argument suffers from three main deficiencies. First, it is likely that the withdrawal agreement will proceed on the basis of the second interpretation outlined above, as it would be more consistent with Article 7 TFEU. Second, it is not obvious that one can separate out the withdrawal agreement from the operation of Article 50, as the two are interlinked, but this is a matter for European law to decide. Third, if the withdrawal agreement does not ‘amend’ the TEU or TFEU, it can still be argued that it ‘replaces’ the TEU and TFEU.

The reason for this third argument may be found again in the 2011 act. According to Section 18, the status of EU law, including the treaties, ‘falls to be recognised and available in law in the United Kingdom’ only by virtue of UK law. So, if the withdrawal agreement changes the relationship between the United Kingdom and the European Union institutions, as it must if there is to be a continuation of the acquis and institutional framework during the transition period, then it ‘replaces’ the TEU and TFEU as a matter of British law. This does not apply in a ‘no-deal’ scenario as, in those circumstances, there is no treaty to trigger Section 2 of the 2011 act.

Under the analysis here, therefore, it is plausible that the withdrawal agreement could (1) contain clauses that amend the treaties, (2) be viewed as forming an integral part of the Article 50(3) amendment procedure as a matter of European law such that it ‘amends’ the TEU and/or (3) ‘replace’ the TEU and TFEU as a matter of UK constitutional law by virtue of the status of European law under the 2011 Act. These considerations do not apply if there is no deal.

The Referendum Condition

Let us return then to the European Union Act 2011. The act provides for a referendum if a treaty ‘amends or replaces’ the TEU or TFEU and satisfies one of the conditions laid out in Section 4. Based on the analysis above, it seems that the withdrawal agreement will satisfy the first condition, but will it satisfy the second? This is unclear at present, but it appears unlikely that the UK will be able to avoid new enforcement mechanisms with Brexit that are differ from EU membership.

The Commission position paper on the transitional agreement contains a new mechanism for the EU to suspend certain benefits from participating in the internal market for the UK ‘where it considers that referring the matter to the Court of Justice of the European Union would not bring in appropriate time the necessary remedies.’ Recent internal EU guidelines includes a list of ‘available sanctions’ under three models: CETA, Ukraine DCFTA, and the EEA. Each includes new sanctions that are not currently available. Moreover, the statement by Brexit Secretary David Davis in December 2017 suggesting that the UK government did not regard itself as bound by the Phase 1 Brexit agreement makes it more rather than less likely that such provisions will be enshrined in the withdrawal agreement. While it is not clear what model will ultimately be adopted, it seems inevitable that some new sanctioning mechanism will be included.

These provisions would meet the condition in Section 4(1)(j): ‘the conferring on an EU institution or body of new or extended power to impose sanctions on the United Kingdom’. It might be argued that the 2016 EU referendum disposed of the requirement under the 2011 act. However, that referendum merely provided the basis to notify the European Union of the intention to withdraw under Article 50. It did not provide for any particular form of withdrawal and therefore does not cover the withdrawal agreement.

It seems likely therefore that the withdrawal agreement would meet the second part of the test outlined in the European Union Act 2011 and a referendum is required under the 2011 act. Before moving on to consider the Withdrawal Bill, it is worth noting a possible argument that Mark Elliot advanced against the interpretation outlined here:

What the act means by ‘replace’ must be understood by reference to the underlying purpose of the act. And it is arguable that its purpose is to regulate the exercise of the UK Government’s treaty-making powers vis-à-vis the EU in the context of the UK being a member of the bloc. Indeed, the act exists in order to provide a direct democratic check upon the Government’s management of the terms of the UK’s membership of the EU, the idea being that there is a brake – in the form of the referendum requirement – upon any contemplated transfer of competence to the EU. On this view, the act is concerned with treaties that shape the UK’s membership of the EU, not with treaties that are concerned with the termination of that membership or which provide for a post-membership relationship between the EU and the UK. In this way, it can be argued that a treaty providing for the UK’s departure from the EU would be outside the scope of the EU Act 2011, as would a treaty staking out a relationship between the UK and the EU that did not entail membership.

This reading seems to accord with the motivation of the proponents of the 2011 act. The tests outlined in the act itself are, however, what constitute current law. It would have been possible for the act to have excluded certain types of treaties – such as those associated around withdrawal – and indeed as indicated the act already does have exemptions, such as for accession treaties. The fact that withdrawal treaties were not included in this list is significant.

A referendum under the 2011 act would not be a rerun of the 2016 referendum – Remain or Leave – it would instead be on the withdrawal agreement and a straightforward Yes/No choice. This would obviously place the conduct of the UK government in the Brexit negotiation process squarely before the British people (and Gibraltar) to vote on.

Parliamentary Procedure and Exiting the EU

The above reading of the European Union Act should be tempered by an acknowledgement that it is due to be repealed by the EU Withdrawal Bill. The fact that it is due to be repealed indicates that the drafters of the Withdrawal Bill felt that there was at least a possibility that the 2011 act could be triggered.

This is where the importance of the House of Lords becomes apparent. The fact that the 2011 act remains in force today means that the position of the Lords is considerably strengthened if it wishes to see another referendum on the withdrawal agreement, particularly where there is no electoral mandate for the government’s position. The status quo, under the reasoning above, is that a referendum is required. Failure to agree on the Withdrawal Bill will simply maintain this legal position.

EU chief negotiator Michel Barnier has indicated that agreement on the Withdrawal Bill needs to be reached by October 2018. Therefore, if the 2011 act is still in force until the beginning of 2019, a referendum would need to be held – a minister must lay a statement before Parliament less than two months after the agreement is reached.

Moreover, the Lords could propose an amendment to the Withdrawal Bill to include a referendum on the withdrawal agreement on the grounds that, as the law in the United Kingdom currently stands, the people are entitled to one more vote (under the 2011 act). It is unlikely that the Commons would agree to such an amendment, but if the Withdrawal Bill fails to pass, then the 2011 act will continue in force.

Alternatively, the threat of a referendum on the withdrawal agreement by the House of Lords might prompt the government to attempt to craft a cross-party consensus on the outlines of the withdrawal agreement. In this case, the government need only to convince a majority in the Lords to back the Withdrawal Bill, repeal the 2011 act and the withdrawal agreement would merely require parliamentary approval.

A further complication is the fact that, under the UK’s devolved structure, a legislative consent motion is required from the devolved legislatures. Neither the Scottish Government nor the Welsh Government are prepared to recommend that their respective parliaments consent to the Withdrawal Bill at present. This presents another legislative hurdle which might serve to prevent the passage of the Withdrawal Bill. If it does so for long enough, then the European Union Act 2011 will remain UK law when the agreement is reached and will require a referendum. In the case of the Scottish Government, this may not be an unwelcome turn of events.

What would be the outcome if the withdrawal agreement is voted down? One possibility is that the two-year default date then applies. A more likely outcome is that the UK government would seek an Article 50 extension, given that there was no mandate for the agreement which it had negotiated.

It is not clear that it would necessarily be in the European Union’s interests to grant such an extension, but if the United Kingdom indicated a wish for a closer relationship with the Union, and the alternative was a ‘no-deal’ Brexit, the EU27 may well be convinced that such an extension would be in their own interests. A renegotiation might then proceed on different lines – perhaps membership of the European Economic Area.


Francesca GelminiDonal Coffey
Max Planck Institute for European Legal History

Dr Donal Coffey is Senior Research Fellow at the Max Planck Institute for European Legal History in Frankfurt am Main. His research focuses on the legal history in the period between the First and Second World Wars and he is Managing Editor of the American Journal of Legal History.
.


Shortlink: edin.ac/2EqqTtW | Republication guidance

Please note that this article represents the view of the author(s) alone and not European Futures, the Edinburgh Europa Institute or the University of Edinburgh.

Creative Commons License This article is published under a Creative Commons (Attribution-NonCommercial-NoDerivatives 4.0 International) License.

Leave a Reply

Your email address will not be published. Required fields are marked *