Do referendums have spill-over effects on national elections? Despite extensive research on the consequences of direct democracy for civic engagement, this question has received little scholarly attention. Drawing on a spatial model of issue voting, I argue that referendums lead to a re-alignment between voters and parties, with “Yes” and “No” voters moving closer to the parties that share their own issue positions. I test this argument empirically using cross-sectional surveys and panel data related to five referendums that took place between 2015 and 2016 in Denmark, Greece, Italy, the Netherlands, and the UK. Preliminary results show that support for parties changes substantially following referendum results in each country, with some voters “switching” to the parties that are aligned with their own position on the referendum proposals. Left-right ideology, however, plays a crucial role, since the likelihood of vote switching increases as the ideological distance between voters and parties decreases.

In addition to this data analysis, I am planning to conduct a survey experiment with a representative sample of Italian voters in the context of an upcoming wave of referendums in spring 2022. The experiment aims to test the theorized causal mechanism of issue voting underlying observational data. I will present the experimental results at EPSA 2022.

These findings have important implications not only for our understanding of how citizens make voting decisions in referendums, but also for putting into perspective elites’ ability to influence public opinion, and, more broadly, for the connection between direct and representative democracies.