How the Use of Polarizing Language on Twitter Benefits Political Elites

• In Event: Incivility, Toxicity and the Violation of Communicative Norms

Abstract

Existing studies on elite social media behavior show that a growing number of politicians are using social media in ways that antagonise and provoke, triggering, at the same time, strong reactions from the public. Being divisive, provocative, and even uncivil can indeed pay off with social media audiences, with some research showing that this behavior can drive engagement and even further reward certain politicians in elections. But how exactly does this happen and why does a type of often-times anti-normative behavior that deprecates and impoverishes public discourse create more engagement? In this paper we first demonstrate from a macro-level perspective that divisive language by political elites leads to more engagement with content, using a new approach built for measuring not just negative/positive emotion and moral-emotional language, but specifically political divisiveness. Using a multi-country Twitter dataset we show that this phenomenon can be observed not only around communication in high-stake elections,elections such as the 2016 and 2020 US presidential elections, but also in low-intensity “second order” elections such as the 2019 European Parliament election. Second, we theorize and put into test a number of mechanisms behind this dynamic. Specifically, we provide evidence at the micro level from a set of randomized experiments which allow us to demonstrate with strong internal validity (a) the theoretical idea that divisive and polarizing content on Twitter instigates incivility which then translates into engagement and (b) to gauge why exactly this type of behaviour motivates individuals to engage with it on Twitter as opposed to non-polarising behaviour.

Authors

• Spyros Kosmidis, University of Oxford

• Yannis Theocharis, Technical University of Munich