29 March 2019

Protecting the EU from a Kill Switch: Why EU Law Does Not Require EP Elections in the UK

According to the EU, postponing Brexit beyond May 23 legally requires UK elections for the European Parliament. If no elections are held, the argument goes, the new European Parliament would not be legally constituted. All acts adopted by that Parliament could then be open to legal challenge. Imagine the chaos.

Several legal arguments support this view. Chief amongst them is Article 14(3) TEU, which requires ‘direct universal suffrage in a free and secret ballot’ for European Parliament elections. More generally, elections should not be trifled with, especially not in a Union based on the principle of democracy. The UK, it seems, should therefore choose: leave before May 23, or hold elections for the European Parliament. 

Yet, on closer inspection, this conclusion is not as legally convincing as it appears. The failure of one Member State to participate in elections for the European Parliament does not necessarily undermine the legality of the entire European Parliament. Luckily so, one may add. For the current EU position effectively gives every Member State an EU kill switch. Simply by not holding European Parliament elections, any Member State could shut down the entire EU. After all, without a functional European Parliament, there would be no new Commission, no EU budget, and no EU legislative acts in general. Imagine Victor Orban, Marine Le Pen, Luigi Di Maio,Thierry Baudet or future equivalents wielding such a kill switch. The problem would no longer be Brexit, Frexit or Nexit, but EUxit.

There are at least five arguments why EU law does not oblige the UK to hold elections for the European Parliament, at least whilst Brexit is on-going. 

First, we should not just look at the European Parliament but at the whole EU system for democratic representation. As Article 10 TEU expressly recognizes, EU citizens are also indirectly represented in the European Council and the Council of Ministers, and by national parliaments acting on the EU level. The European Parliament is therefore only one channel of democratic legitimacy in the EU. 

Interestingly, these other representative institutions are not blocked when a Member State chooses not to participate. Say for example that Italy would refuse to participate in the Council. This would in no way block the functioning of the Council, nor would we doubt the legality of any Council decisions taken by a sufficient majority. Quite the contrary: the Italian absence would not even contribute towards a blocking minority. Of course, we faced precisely this situation when De Gaulle opted for his famous ‘empty chair policy’. This blocked EU decision-making where unanimity was required but did not threaten the legality of any acts adopted during this time.  

So why would the same rule not apply to the European Parliament? Why would the European Parliament, different from the Council and European Council, become illegitimate when one Member State voluntarily chooses not to participate? Consider also that the quorum for plenary decisions by the European Parliament is only one third of its component members. Only 251 of the 751 MEP’s therefore have to be present to adopt binding laws. This quorum, moreover, is only applied where at least 40 MEP’s request the President to check if the quorum is met. If fewer than 40 MEP’s object, therefore, the European Parliament can adopt acts with even fewer than 251 MEP’s, which can easily mean that multiple Member States have no MEP voting. 

Yet what about Article 14(3) TEU and democracy, which require ‘direct universal suffrage in a free and secret ballot’ for European Parliament elections? So even if the European Parliament could function without the UK holding European elections, would this not violate the individual rightof each UK citizen to vote for the European Parliament in an election? 

This is where the second argument becomes relevant: EU law only provides a rightto the UK and its citizens to participate in EU elections, but not an obligation. Here some close reading of the EU Treaties is in order. To begin with, Article 14(2) TEU only states that the European Parliament shall ‘be composed of representatives of the Union’s citizens’. Notice that it does not require all Member States’ citizens to be represented per Member State. Quite the contrary: formally all MEPs are required to represent the interests of all EU citizens anyway. What is more, Article 14(3) TEU only says that ‘[t]he members of the European Parliament shall be elected for a term of five years by direct universal suffrage in a free and secret ballot’. Note that it again does not say that all Member States must participate. The provision only states that the European Parliament may not be elected in a different way. Lastly, Article 20(4)(b) TFEU further supports the conclusion that there is no obligation to join in European Parliament elections. This provision holds that all EU citizens have ‘the right to vote and to stand as candidates in elections to the European Parliament (…)’. In other words, participating in EU elections is a right, not an obligation. If correct, this also means that EU citizens may choose not to use this right by not participating in EU elections. Take the hypothetical example where all French citizens would choose to boycott the European Parliament elections, donning purple jackets instead. This would leave France without validly elected MEPs. But would it undermine the legality of the European Parliament for five years?

Of course, this argument raises the further question of how a Member State, or a Member People, could legally decide not to participate in EU elections. This is particularly tricky because voting for the European Parliament is a right of the individual citizen, and not of the Member State, as the judgments in Matthews and Eman confirm. So, could the UK take away the individual rights of UK citizens that do want to participate in European Parliament elections? 

Obviously, this brings us to dangerous terrain. Withholding the right to vote is as dark and problematic as it gets in terms of democracy, and should not be allowed lightly. Yet if there would be one way to justify such a restriction of voting rights then it would seem a popular referendum to leave the EU, by direct universal suffrage and in a free and secret ballot, would be it. UK citizens voted to leave the EU, which logically entails losing the right to vote for the European Parliament. If the democratically-elected UK government, in executing this wish of the UK voters, then considers it necessary to delay exit without European elections, so be it. Consider in this regard also the hypothetical situation where the UK would have submitted its notice to leave on 29 May 2017 instead of 29 March. The UK would then still have been an EU Member State on 23 May 2019, without the need for an extension. Could UK citizens then still claim an individual right to vote in European Parliament elections, even if an exit would just be a week away (absent an extension or revocation)?

The argument that, in the exceptional circumstances of Brexit, it may be allowed not to hold European Parliament elections is further strengthened by a third argument. This argument isbased on the unique legal context of Article 50 TEU. Joining and leaving the EU are special moments of transition where some rules should and have been applied flexibly. Now of course leaving is not the same as joining, and there are important differences between Articles 49 and 50 TEU. Yet it is fair to assume that at least the same level of flexibility should be available in the treatment of a Member State that is leaving as in the mirror scenario of a State that is actually becoming a full-blown Member. And in the context of accession, multiple fundamental principles of EU law have been temporarily deviated from, including the right to vote. 

In the accession of Bulgaria and Romania, for example, it was agreed that Romanian and Bulgarian citizens would not have full free movement rights. Even though they had just become full EU citizens, Bulgarians and Romanians were not treated equally to other EU citizens. This deviated from the foundational EU principle of non-discrimination and equal treatment, which is at least as fundamental as the right to vote for European Parliament elections.

The accession of Croatia provides an even more striking example. Article 19 of the Accession Treaty makes an explicit derogation to Article 14(2) and (3) TEU. Derogating from primary law, Article 19 first changes the number of seats in the European Parliament. Subsequently, it provides for ad hoc elections in Croatia, even before accession, to fill these extra seats. If, however, the accession of Croatia would be less than six months before the next European Parliament elections, Croatian MEPs could be ‘designated by the national Parliament of Croatia’. Apparently, this would not have undermined the legitimacy of the European Parliament. So why would it be illegitimate if the UK, for a limited time and in thecontext of withdrawal, did not participate in European Parliament elections? Or if the UK Parliament, for example, prolonged the tenure of current UK MEPs by ‘designating’ them instead of holding elections? Such an approach, furthermore, would also fit with the principle of subsidiarity and the fact that it is up to Member States to determine the specific system for elections to the European Parliament, albeit of course within the limits of EU law

fourth argument concerns the substance of democracy. Would it not be incompatible with the EU’s commitment to substantive democracy that UK citizens would be bound to decisions from a European Parliament they have not voted for? In short, no. To begin with, the UK would have democratically chosen not to participate in the European Parliament elections, whilst it would still be represented in the European Council and the Council. More importantly, there are multiple states that are substantively bound by EU law without having any direct representation in the European Parliament (or any other EU institution for that matter). Take the examples of Norway or Switzerland. Substantially, both countries are legally obligated to transpose EU law, in which they had no say, into their national laws. Or staying closer to Brexit, take the example of the transition period as agreed in the EU-UK withdrawal agreement. During transition, which could last between a year and a half to almost four years, the UK would be fully bound by EU law without having any representation in EU institutions, including the European Parliament. Apparently, this would not violate the principle of democracy, so why would it if the UK only lost its representation in the European Parliament for a limited time in the run-up to its withdrawal? 

fifth and last argumentconcerns the proper interpretation of EU law. When confronted with multiple possible interpretations of EU law, the Court of Justice of the EU (CJEU) of course protects fundamental rights. Yet most of all the CJEU protects the effectiveness, or effet utile, of the EU and its legal order. As seen in inter alia Opinion 2/13Kadi, or Achmea, the CJEU will go to great lengths to protect the autonomy and stability of the EU legal order. Confronted with one interpretation which temporarily offers greater protection to the democratic rights of UK citizens but grants an EU kill switch to each Member State, or another interpretation which limits the democratic rights of UK citizens, albeit for a limited time and in a limited fashion, but which safeguards the future functioning of the EU from fatal sabotage, the meta-principle of effectiveness would argue for the latter. Such an interpretation also follows from the commonsensical rule that, when interpreting constitutional-like instruments such as the EU treaties, one should choose an interpretation that does not undermine the long-term viability, stability, and functioning of the polity created. Lastly, an interpretation that would allow the UK not to organize elections would also fit with the strong focus on sovereignty and self-determination in the process of withdrawal as stressed by the full CJEU in Wightman.   

Clearly the above arguments do not provide a definitive answer, and counter arguments exist. Yet, with Brexit we are in legal and political terra incognita. Moreover, thanks to political brinkmanship, we are far beyond the realm of neat, happy, elegant, win-win or even relatively decent solutions. 

As a result, we have the obligation to look for the least horrible solution. So yes: not holding European Parliament elections in the UK, even if the UK remains a Member State after 23 May, is ugly and clearly regrettable. But it is not so obvious that it would be illegal or threaten the legality of the European Parliament. And faced with two possible legal interpretations, why not choose the one that does not hand an EU kill switch to populists? Granted, this is a normative as well as a legal position, but if Article 50 TEU provided populists with the nuclear option of leaving the EU, let us not give them the hydrogen bomb of paralyzing the entire EU by not holding EU elections. No Brexit mess is worth that much.  

SUGGESTED CITATION  Cuyvers, Armin: Protecting the EU from a Kill Switch: Why EU Law Does Not Require EP Elections in the UK, VerfBlog, 2019/3/29, https://verfassungsblog.de/protecting-the-eu-from-a-kill-switch-why-eu-law-does-not-require-ep-elections-in-the-uk/, DOI: 10.17176/20190517-144339-0.


  1. Tim Corthaut Fri 29 Mar 2019 at 11:34 - Reply

    I disagree with most points.

    1) While I agree that the European Parliament could validly convene if no representatives of a MS are present, it does not follow that a MS could opt not to organise EP elections. It is not because the EP, like the other institutions could survive such an anomaly that the anomaly becomes the norm. Moreover, if the UK is still a Member, this means that the other MS will still have to hold EP elections according to the present division of seats.

    2) The second argument obfuscates the individual political right to vote in the EP elections. Surely, citizens could decide to stay home en masse, but the decision to take away the franchise from the citizens does not lie with the UK political institutions, but solely with the EU pouvoir constituant.

    3) That is precisely the reason the third argument is unconvincing as well. The examples given are all instances where the divergent modalities for the EP elections were laid down in the accession treaties. They were not the result of the political whim of the local authorities who just didn’t feel like it.The hypothetical of a Brexit on 29 May 2019 is just that – precisely to avoid this type of grey zone, the European Council has set the date before the EP elections, and for the UK that extra week should not matter (especially not since, apart from its political rights, economic integration continues in that scenario for the time being as part of the Withdrawal Agreement).

    4) The argument only holds if we accept that the refusal to organise EP elections is a decision that is for the UK institutions – democratic or not – to take. As explained under 2) I disagree.

    5) I do agree that the ECJ does not need to opt for the kill switch interpretation. But that does not mean that in doing so it can immediately also give the UK – and the other MS, as the correct composition of the EP depends on whether EP elections take place with or without the UK still being a member – a pass as to the way EP election are organised. At the very least the Commission could bring infringement procedures against the UK and win. Also, if a UK court, confronted with a claim of even a single UK voter (which, by the way does not even have to be a UK national, but could also be an EU citizen from another MS who would also be disenfranchised), were to be referring questions as to whether the UK had to hold EP elections, the ECJ could only answer in the affirmative.