We evaluate how traditional parties may respond to populist parties on issues that are particularly fitting for populist messages. The testing ground is the 2020 Italian referendum on the reduction of members of Parliament. We implemented a large-scale field experiment with almost one million impressions of programmatic advertising, a concomitant survey experiment, and a follow-up survey in 2023. Our treatments consisted of two video ads: an informative video on the likely costs of cutting MPs aimed at deconstructing the populist narrative, and a trust-reducing video aimed at attacking the credibility of populist politicians. Our field experiment shows that the latter video is more effective at capturing the viewers' attention, and it also decreases the "yes" votes in favour of cutting MPs. This effect is driven by a reduction in turnout rather than by an increase in the 'no" votes, that is, the randomised campaign is effective mainly because of demobilisation. Consistently with this mechanism, treatment effects are larger in municipalities with less college graduates, more people out of the labour force, larger populist votes in the previous elections, and smaller in size. The survey experiment corroborates the field findings. In the subsequent National election in 2022, municipalities treated with the anti-populist campaign vote more for a "new" populist party, Brothers of Italy, and less for both traditional parties and the "old" populists targeted in the 2020 campaign. The follow-up survey shows that, in treated municipalities, interest in politics increases but trust in political institutions decreases, while the new populists fare better than the others, disclosing the existence of a peculiar populist dynamics.
Co-authors: Vincenzo Galasso, Massimo Morelli and Piero Stanig (Università Luigi Bocconi)