Responsive withdrawal? The politics of EU agenda-setting

ABSTRACT

This contribution asks whether and why the newly political environment of EU law-making impacts on the European Commission's choice (not) to announce the withdrawal of legislative proposals. We argue that the Commission uses ‘responsive withdrawal’ in response to bottom-up pressure, so as to signal self-restraint or policy-determination to different audiences. Bottom-up pressures are driven by (1) the national contestation of ‘Europe’; (2) visible controversy about optimal (crisis) governance; and (3) the domestic salience of EU legislation. Our hypotheses are tested on a new dataset of all codecision files concluded, withdrawn, rejected or ongoing between 2006 and 2018. We show that the Commission reacts to bottom-up pressure by either politicising or depoliticising the EU's legislative agenda: ‘withdrawal announcements’ are more likely when Euroscepticism is high and when legislation touches core state powers, but less likely when legislation is domestically salient. We also demonstrate the continued importance of cyclical and technical reasons. Our analysis complements extant explanations of withdrawal as the upshot of functional factors or of uncertainty, and contributes to the nascent debate about whether, why and how supranational actors respond when the systems in which they operate – and the policies they produce – come under attack.

KEYWORDS:

• Agenda-setting
• European Commission
• legislation
• politicisation
• depoliticisation
• withdrawals

Acknowledgements

We would like to thank Berthold Rittberger, two anonymous reviewers and participants at the Workshop ‘Politicising and De-Politicising the European Union’ (King's College London, 15 and 16 December 2017) for their helpful feedback. Nicolò Conti, Simona Piattoni as well as audiences at the King's College Public Policy and Regulation Workshop (15 February 2019) and the European Governance Colloquium at the Hertie School (4 March 2019) also provided constructive comments on earlier versions of this article. We are grateful to Niccolò Massei, Francesca Minetto and Michele Scotto di Vettimo for their excellent research assistance.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Christine Reh is Professor of European Politics and Dean of Graduate Programmes at the Hertie School, Berlin, Germany.

Edoardo Bressanelli is ‘Montalcini’ Assistant Professor at the DIRPOLIS Institute, Sant’Anna School, Pisa, Italy.

Christel Koop is Senior Lecturer at the Department of Political Economy, King’s College London, UK.

ORCID

Christel Koop http://orcid.org/0000-0002-7125-6439

Notes

1 We only include ongoing legislation proposed before the end of 2016, because by June 2018, these files would have either been withdrawn or would still be ongoing.

2 This is also reflected in the Juncker Commission's WPs. For instance, the first programme stated that the Commission was ‘voted into office with a commitment to make a difference: to do different things and to do things differently’ (2014, p. 2).

3 Procedures 2005/0127(COD) and 2013/0059(COD).

4 Unlike Boranbay-Akan et al. (2017), we exclude ‘procedural change’ as a variable. Our ‘codecision-only’ dataset starts in 2006 and does not include proposals transferred to the OLP.

5 As robustness checks, we ran a model with an alternative operationalisation of ideological conflict; a model estimated on a subset of observations excluding Barroso I (where withdrawals were rare); and two models changing the sequence of including different categories of variables. The results do not differ substantively. All models are presented in Table C in the Online Appendix.

6 Our data also include the Barroso I period, but the number of announced withdrawals is so low (only eight) that we cannot run a separate analysis for that Commission. This is relevant in itself, suggesting that the power to withdraw has only recently been embraced as an agenda-setting tool.

7 The difference in approaches is confirmed when adding an interaction between Juncker and substantive files to Model 4 in Table 1 (see Table D and Figure A in the Online Appendix).

Additional information

Funding

Research for this article was supported by a British Academy/Leverhulme Small Research Grant (SG160582; ‘Politicising Europe’).