Why do politicians share conspiracy theories on social media? The role of ideology, incumbency, and electoral cycle

Over the past decade, the increased visibility of conspiracy theories has raised significant concerns about their impact on democratic processes. Historically perceived as fringe phenomena, conspiracy theories have now entered the mainstream, influencing public discourse and challenging the epistemic foundations of democracy. While substantial research has explored the determinants of individual-level conspiracy beliefs, the role of political elites in propagating such narratives remains underexamined. We argue that examining the supply side of conspiracism is crucial since elite cues remain instrumental in shaping public opinion. Therefore, this paper addresses this gap by investigating the patterns of conspiratorial communication among political elites on social media in Belgium, Finland, Greece, and Poland. Using the concept of conspiracism as a social identity frame, we assume that political elites utilize conspiratorial communication to strengthen the cohesion of the in-group via epistemic appeals while vilifying political opponents. Hence, we categorize a post as conspiratorial if it contains any of the conspiratorial frames: exclusive access to hidden Truth, signaling the existence of a plot against the in-group, identifying powerful conspirators and their alleged interests, and secrecy. Substantively, the analysis has entailed meticulous quantitative manual coding of posts coming from politicians from all political parties (N = 34) on Facebook and X. Firstly, we have delineated our inquiry to the 7 months before the parliamentary election in each country, with the 3 months before the election serving as an electoral period and 4 months prior as non-electoral. Our data includes 49,940 annotated posts on Facebook from 889 politicians and 36,761 posts on X produced by 713 politicians. The analysis finds that actors on the radical left and right are most prone to share conspiratorial content. We find the effect to prevail even after controlling for incumbency [government vs opposition] and populism. When splitting data into non-electoral and electoral periods, we find that the results hold, albeit conspiratorial discourse appears more frequently during a non-electoral period, a phenomenon we aim to explain further. This paper contributes to the literature on conspiracy theories by offering three key contributions: (1) an operational framework for studying conspiratorial communication at the elite level.; (2) a first cross-country analysis of top-down conspiracism; and (3) outlining a novel research agenda that would combine insights of political science, communication science and social psychology.