Feeling all the (partisan) feels: Exploring the drivers of affective polarization at the individual-level

Thesis

Affective polarization—hostility and aversion between rival political groups—is an increasingly salient feature of Western democratic politics. While scholars agree that this trend raises serious concerns for public life, its underlying drivers are still debated. This thesis enriches this line of research by clarifying some of the mechanisms shaping affective polarization at the individual level. My overarching research question is: What drives affective polarization at the individual level? I tackle this question through three distinct empirical studies that engage with separate aspect of this phenomenon. Study I leverages survey data to show that, among UK partisans, perceived conflict with their in-party on policy issues curbs affective polarization. Moving on to the digital sphere, Study II deploys computational methods to reveal the existence of affective biases in online political discussions between ideologically opposed Reddit users, and shows how the tone of their exchanges can both mitigate and worsen these patterns. Study III, finally, provides experimental evidence that only exposure to certain forms of online partisan messages have downstream consequences for affective polarization: those that explicitly threaten and devalue an individual’s own political party. Taken together, these findings reveal the drivers of affective polarization to be contingent on both situational and dispositional factors, including how ideologically aligned partisans are with their own political party, how their perceived political rivals engage with them in online conversations, and whether they are made to feel threatened by them.

Authors: Marchal, N

• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: social identity theory; partisanship; political communication; affective polarization
• Subjects: Political psychology; Communication--Political aspects; Social media