The arduous process of withdrawing the United Kingdom (UK) from the European Union (EU) began shortly after the Brexit referendum in June 2016. The “leave” vote plunged the British government into an existential crisis and led to dramatic special summits at the EU level. Looking at the British exit vote against the history of EU-related referendums reveals its peculiarities and pitfalls. The resulting mandate left the government some room for manoeuvre and initially offered plenty of options for its negotiations with Brussels. However, the government in London was not able to translate the tight “No” vote into a viable negotiating position. The EU insisted on negotiating exclusively with Her Majesty’s Government. Paradoxically, referendums in member states illustrate the European multi-level system’s dependence on states. Member states and the EU can both learn lessons by observing the British government’s handling of the exit referendum.
Since 1972, there have been 60 EU-related referendums held at the national level in Europe, but only in two of them – both in the UK – was leaving the EU an option: in 1975 and in 2016. While elsewhere accession, treaty ratification, or individual policy issues (such as the introduction of the euro) were put to a vote, the exit referendum is a British specialty. The Brexit decision of June 2016 is thus, on the one hand, part of a longer trend of EU politicisation, in which European issues increasingly have become the subject of public debate. On the other hand, the exit referendum represents a special case in the history of EU-related referendums. By interpreting the outcome extremely narrowly, the British government has further amplified the explosive power of the Brexit referendum. Their interpretation provides a starting point for explaining the drama of the exit negotiations.
Referendums on EU Issues
In the first 20 years (1951/52–1971) of European integration, referendums did not play any role, and in the following 20 years (1972–1991), they only played a subordinate role. However, in the course of EU enlargement, constitutionalisation, and politicisation in the 1990s and 2000s, national referendums became a common instrument for legitimising European political reforms and decisions: 27 of the 60 EU-related referendums were decisions on membership (i.e. votes on membership application accession or withdrawal), 18 were for the ratification of European treaties, and the remaining 15 concerned individual EU-related issues. The large majority – 43 of the 60 EU-related referendums – had outcomes that confirmed the positions of the national governments and allowed the EU to proceed as planned.
In those cases where the majority of the population voted for an EU-sceptical option – thus opposing an uninterrupted course of EU reform or decision-making processes – both the national government and the EU were compelled to provide responses. In failed referendums on accession to the EU or the European Economic Area (Norway, Switzerland), alternative, sometimes complex arrangements had to be found, and rejected EU treaty reforms made renegotiations and clarifications necessary. EU referendums have thus become a relevant factor in EU policy.
National Referendums in the European Multi-level System
Referendums are not held at the EU level. As a union of states and citizens, the EU is not a self-contained electoral body. In contrast to the elections to the European Parliament, no EU-wide referendums have been held simultaneously on the same topic at the level of the 28 member states. Not even the dates of the 10 announced national referendums for the ratification of the 2005 EU Constitutional Treaty were coordinated. In this respect, referendums remain territorially limited to the member states and are largely under their control. National provisions based on constitutional and political traditions apply, and there are no sets of rules for national referendums that are agreed upon or recommended at the EU level.
Referendums in the EU member states always take place in a multi-level political system without any involvement of the supranational level. The EU institutions are simply presented with the outcomes and only become involved when referendums have a negative impact for the EU. An EU decision requiring consensus among all member states can be brought down by a referendum: If a referendum makes it difficult for a member to ratify EU treaty amendments or Association Agreements with third countries, for example, then the EU is at least prepared to clarify matters. However, negative referendum outcomes cannot directly prevent specific EU policy actions. A member’s “No” vote in the Council that is backed politically by a referendum has greater weight if a legal act requires unanimity in the Council than in the case of qualified majority voting. The referendum held in Hungary in 2016 that rejected the introduction of a distribution mechanism for refugees at best politically strengthened the Hungarian government. But there was no need for the EU to take notice or react.
Peculiarities of the Exit Referendum
In referendums on accession, amendments to European treaties, or even individual issues, a “Yes” usually means agreement with a change to the status quo, which is laid down in legal texts (Accession Treaty, treaty changes, etc.), open to public scrutiny, and subject to legal review. In the event of a “No” vote, the status quo remains. In the Brexit vote, however, the population had a choice between the status quo (remain) and a leap into the dark (leave). Parts of the ruling Tory party advocated remaining in the EU after Prime Minister David Cameron negotiated slightly altered membership conditions with the EU. However, the special provisions of the so-called February Settlement were hardly mentioned in the referendum discourse.
It remained vague as to what future relations with the EU would look like in concrete terms under the condition of independence. The only thing that was clear was that the withdrawal had to be based on Article 50 of the Treaty on European Union (TEU). The withdrawal article is designed to protect the acquis communautaire and the interests of the remaining member states. Accordingly, the British government had to conclude an agreement with the Union on the details of the withdrawal, taking into account the framework for future relations. The undefined meaning of the “leave” vote gave rise to all sorts of projections. That is why the debate on future relations did not really begin until after the referendum. Ultimately, it depended on how the new government under Theresa May would interpret the tight result (51.9 to 48.1 per cent) and when it would put exit negotiations into motion, which were to last two years.
The extremely narrow interpretation of the referendum following the campaign for restoring national sovereignty and independence and regaining control over borders, jurisdiction, and finances marked red lines that Prime Minister May elevated to dogma: no customs union, no participation in the internal market and the four freedoms, and no jurisdiction for the European Court of Justice. With this rigid mandate, the May government subjected itself to (erroneous) party-political calculations and pitfalls. From the outset, this excluded a number of tried and tested models of partnerships with third countries (European Economic Area, Turkey model) in the hopes of developing a unique new model.
What Governments and the EU Can Do
If a government does not receive the desired approval for its position in an EU-related referendum, it must (re-)position itself politically in both the national and EU arenas. The most important step is the interpretation of the vote’s outcome. Even if it is a binding vote and not a mere advisory vote (as in the case of Brexit), it is up to the government to assess it.
In 2005, the French and Dutch governments considered the rejection of the Constitutional Treaty to be final. The EU had to accept the failure and make a new attempt with the Lisbon Treaty. However, the Danes voted twice on the identical Maastricht Treaty (1992/93), as did the Irish on the treaties of Nice (2001/02) and Lisbon (2008/09). The governments thus created room for manoeuvre and dialogue with the EU institutions, and both agreed on clarifications, protocols, and anticipatory resolutions for future treaty changes. These amendments were put to a vote again without requiring the ratification process to be repeated in the other member states. Another example of flexibility and creativity on both sides is the struggle over the Association Agreement with Ukraine, which Dutch voters rejected in 2016 in a referendum initiated by EU-sceptical associations. The government then acted on individual points of criticism and obtained legally non-binding clarifications from the EU, but it ratified the agreement after it was adopted in parliament.
Whether an exact implementation of the outcome of the vote is unavoidable depends, above all, on the strength and determination of the government and its ability to engage in dialogue with the population and the political parties represented in parliament. The Tories’ severe losses in the early elections, the formation of a minority government tolerated by the Democratic Unionist Party, and the deepening divide on Brexit within and between the Conservative and Labour parties all provided adverse conditions for a pragmatic approach. There was also no real change of opinion in the electorate. As a result, neither the Leave nor the Remain camp clearly gained the upper hand.
The repeated rejection of the Withdrawal Agreement in British Parliament shows that the government has not yet been able to offer an interpretation of the referendum outcome that can be approved by a majority. Much therefore depends on the repositioning of the British government. Three possible paths emerge: Conservatives and Labour agree on the basic lines of future relations with the EU in the critical areas of external tariffs, free trade, and the internal market; new elections to the House of Commons; and/or a second referendum. It is up to the government and the House of Commons whether one or more of these paths are taken.
The EU saw the high turnout (72.2 per cent) for the Brexit referendum as a sign of legitimacy and did not press for a second referendum. In order not to jeopardise internal cohesion, the EU does not want to renegotiate the Withdrawal Agreement from November 2018. However, the 27 member states are prepared to interpret or modify provisions in the Political Declaration accompanying the Withdrawal Agreement in such a way as to support an internal compromise in the UK and to pave the way for ratification of the Withdrawal Agreement.
In addition to the two-level constellation – whereby the government negotiates simultaneously with British actors and European partners – the time limit created incentives for tactical manoeuvring. The EU has extended the deadline twice already in order to give London room to find a compromise in the government and the House of Commons. In principle, the EU is demanding the adoption of the Withdrawal Agreement, but it is flexible in the choice of model for future relations – even if it means remaining in the EU.
Political Lessons Learnt
In the European multi-level system, national referendums are not the end, but part of a political process, even – and especially – when the government’s position is rejected. What is critical for moving forward is how the governments interpret the referendums’ outcomes, because governments are the negotiation partners for EU institutions and EU countries.
Leaving the EU is a huge step. In this respect, legitimisation by means of a referendum is plausible. By analogy with constitutional amendments, it would be advisable to establish a quorum of votes in favour that goes beyond the simple majority as well as a quorum for voter turnout in the respective member state. It should also be apparent to the electorate what a withdrawal means for their future relations with the EU and what mandate the government will be given with the vote. The EU institutions will remain reluctant to intervene in referendum campaigns. However, parties and civil society may do so.
It can be assumed that the provisions of Article 50 TEU will be examined more closely in light of first experiences from the Brexit process. The withdrawal article is deliberately not designed in a user-friendly way, as it is not intended to provide incentives to renegotiate the rights and obligations of membership. At the same time, it must offer the state wishing to withdraw a viable way out of the Union. Whether this leads to a free fall, however, is largely up to the withdrawing state and its government.
Dr Barbara Lippert is Director of Research at SWP.
Erik Brandes, M.Sc., is research assistant to the Director of Research.
© Stiftung Wissenschaft und Politik, 2019
All rights reserved
This Comment reflects the authors’ views.
SWP Comments are subject to internal peer review, fact-checking and copy-editing. For further information on our quality control procedures, please visit the SWP website: https://www.swp-berlin.org/en/about-swp/ quality-management-for-swp-publications/
Stiftung Wissenschaft und Politik
German Institute for International and Security Affairs
Telephone +49 30 880 07-0
Fax +49 30 880 07-100
(English version of SWP‑Aktuell 22/2019)