From Fear to Radicalization: Emotional Dynamics of Extremism

This study investigates the relationship between fear and extremist attitudes in periods of heightened socio-political uncertainty, drawing on two cross-sectional surveys conducted in October 2024 and October 2025 in Germany. Existing research in political psychology highlights the central role of emotions in radicalization processes: collective anger and contempt have been shown to mediate the effects of personal and collective deprivation on the legitimization of both normative and non-normative, including violent, political action. Personal emotions can aggregate into collective emotional climates that license radical behavior, particularly when grievances are framed as shared and morally justified. Survey research further demonstrates that dispositional anger and anxiety, as well as emotionally triggering social interactions, are associated with support for extremist groups and a willingness to engage in political violence. Voting studies similarly indicate that anger directed at politics increases support for radical parties on both the left and right, even after controlling for ideological attitudes, whereas fear and positive emotions tend to show weaker and more conditional effects. Despite this rich theoretical landscape, much of the empirical evidence relies on cross-sectional data, limiting our ability to assess temporal dynamics and potential causal pathways between emotions and extremist attitudes. To address this limitation, I employ a Bayesian linking approach to construct a pseudo-panel using repeated cross-sectional data. Both surveys include identical measures of emotional states—most notably fear—and of extremist attitudes. Using Bayesian hierarchical modeling, respondents are probabilistically linked across the two waves based on observable characteristics and latent constructs, allowing the generation of individual-level trajectories of fear and extremism while accounting for measurement uncertainty and sample differences. This approach enables estimation of the effect of fear on subsequent extremist attitudes, adjusting for demographic and socio-political covariates, and situates fear within a broader emotional framework that emphasizes its interaction with anger-based mobilization processes. While anger has been shown to directly mobilize radical political behavior, fear may operate as a precursor emotion that heightens threat perception, increases receptivity to exclusionary narratives, and facilitates the transformation of personal anxiety into collective grievance. By integrating repeated cross-sectional data in this way, the study provides a more robust assessment of the temporal dynamics linking fear and extremist attitudes than traditional cross-sectional analyses. Substantively, the findings contribute to debates on the emotional drivers of radicalization by clarifying the role of fear in extreme times. Methodologically, the study demonstrates the utility of Bayesian linking as a viable strategy for studying emotional and attitudinal change when true longitudinal data are unavailable, with broad applicability to research on political extremism, public opinion, and socio-political behavior.