Catalysing Events? A comparative perspective on referendums and the politicisation of the European Union

Since the Maastricht treaty of 1992, referendums have been a key feature of European integration. While they ostensibly provide much needed democratic legitimacy for the European project, they have also slowed the process of integration significantly. This has arguably occurred through two kinds of effect. First, that referendums have acted as a direct legislative blocks or reversals to European integration in ballots such as those on Constitution of the EU in France and the Netherlands in 2005, the initial rejection of the Maastricht, Nice, and Lisbon treaties in Denmark and Ireland, and the 2016 Brexit referendum. Second, that referendum processes can have indirect educative effects on citizens, which contribute to an increased public awareness and politicisation around the issue of EU integration. Collectively, it has been argued that these two referendum effects have contributed significantly to the broader trend of an increasingly politicised European Union, and to the ending of the era of ‘permissive consensus’. To what extent referendums have catalysed this new politicisation of the European Union, relative to other potentially politicising factors, is, however, contested. Different empirical approaches have yielded different conclusions. In this paper, I adopt a novel approach to (re)analyse the relationship between EU referendums and politicisation. I use extensive media data from eight EU countries spanning from 1992 to the present day, along with the Eurobarometer ‘Good Thing’ questions to construct to construct interrupted time series models in a comparative design. In doing so, I assess the politicising effects of referendums by measuring the relative impact that they have on the both the salience and polarisation of views on o the European Union over time. I find that referendums are closely associated with short-term spikes in salience. Evidence of medium to long-term associations between the occurrence of referendums and higher salience is, however, mixed. I conclude that while referendums are important media educative events in the process of European integration, their politicising effect on EU issues is not uniquely catalysing when compared to extraneous factors such as the sovereign debt and migrant crises. I conclude by proposing a more nuanced conceptualisation of the relationship between EU referendums and politicisation, which acknowledges key differences in national contexts as well as the role of extraneous politicising shocks.