Getting around no: how governments react to negative EU referendums

ABSTRACT

Referendums, and negative referendums in particular, are an important constraint on European integration and the most direct way in which politicization affects EU-level decision-making. Yet governments retain considerable room of maneuver in responding to them. Whereas there is a rich literature on referendum campaigns, voting behavior, and intergovernmental bargaining under referendum threats, the (inter)governmental responses to referendum outcomes have not been studied systematically. This article examines how governments respond to negative referendums on European integration. It suggests that the type of integration issue put to the referendum (accession, withdrawal, or integration), and the concomitant size and intergovernmental distribution of the costs of failure, structures the post-referendum choices. This conjecture is explored in a comparative analysis of all 15 negative EU referendum cases since 1972.

KEYWORDS:

• European Union
• referendum
• treaty
• politicization
• negotiation

Acknowledgments

For comments on previous versions, I thank Christine Reh, Stefanie Walter, Natasha Wunsch and participants in the EUP colloquium at ETH Zürich, the Paris conference of the ECPR Standing Group on the EU, and the research seminar of the Department of Political Science, Stockholm University. The paper has further strongly benefited from comments by the editors and the reviewers.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes on contributor

Frank Schimmelfennig is Professor of European Politics at ETH Zurich, Switzerland, Center for Comparative and International Studies.

ORCID

Frank Schimmelfennig http://orcid.org/0000-0002-1638-1819

Notes

1 Count based on Beach (2018) plus four Swiss initiatives.

2 A partial exception is Rose’s list of governmental responses to EU referendums (Rose 2019). Walter (2017) offers thoughts on the strategies available to governments in response to democratic choices that create negative externalities abroad.

3 This reasoning is in line with the idea of loss aversion in prospect theory (Kahneman and Tversky 1979). For its applicability to political science, see Levy (2003).

4 For instance, the Hungarian referendum on refugees in 2016 did not reach the necessary quorum for validity. The Swiss 2001 initiative on initiating accession negotiations with the EU was not about an existing or negotiated agreement. It is not clear to what extent subnational EU referendums follow the same logic as national EU referendums and the subsequent intergovernmental negotiations – and there has only been a single case so far. I therefore exclude Greenland’s withdrawal referendum of 1982 because Greenland was not a (member) state and did not negotiate the subsequent Greenland Treaty with the EU.

5 National Declaration by Ireland at the Seville European Council (21 June 2002), available at https://www.cvce.eu/content/publication/2005/6/17/a442cb15-0528-4560-9d12-0b46fb5c75d4/publishable_en.pdf.

6 Brussels European Council, 18/19 June 2009, Presidency Conclusions, available at http://europa.eu/rapid/press-release_DOC-09-2_en.htm.

7 In areas of high unemployment, the new legal provisions oblige employers to first advertise vacant jobs to job centres and interview unemployed Swiss job seekers; it does not require any justification for not hiring them or for preferring non-Swiss applicants.

8 Whether or not voters believed that a negative vote would lead to Grexit, had a strong effect on their vote choice (Walter et al. 2018).