Work in Progress

Research Project

The Politics of Shaming and Shamelessness — Semper Ardens: Accelerate grant, Carlsberg Foundation (2026–2030), Principal Investigator.

Why do some politicians break norms openly, and why does it work for them? This project studies shaming and shamelessness as a dynamic of elite political rhetoric: when political actors shame their opponents or opponents’ supporters, when they shrug off shame themselves, and how both shape political hostility and citizens’ tolerance for political violence. The project combines large-scale computational text analysis of elite communication with experimental and quasi-experimental designs, and funds two three-year postdoctoral positions at Aarhus University. Grant page

Selected Working Papers

Tobias Widmann. Far-Right Telegram Communication and Attacks on Politicians: Evidence from Germany.

Violence against politicians is rising, yet its causes remain poorly understood. I test whether radical-right actors in parliament act as “intellectual arsonists,” fueling this violence through their online communication, and whether they matter more than the extra-parliamentary far right. I argue that portraying opponents as moral enemies, combined with calls to action, licenses hostility. Studying the Alternative for Germany (AfD), I link nearly 2.5 million posts from 300 Telegram channels run by AfD politicians, party organizations, far-right activists, and alternative media to a new daily dataset of more than 11,000 politically motivated offenses—from insults to assaults—against other parties’ politicians and offices (2019–2025). Using transformer-based text classification and vector autoregressions, I test whether surges in such rhetoric precede attacks and whether attacks, in turn, reshape the discourse. Knowing whether and how radical-right parties and their networks contain or incite this violence is central to how democracies protect their representatives.

Francesco Colombo, Tobias Widmann. When Language Signals Change: The Emotional and Threat Response to Gender-Inclusive Language in German Political Communication.

Cultural issues have become central to contemporary political debates, with symbolic questions increasingly shaping electoral competition despite their limited material consequences. We argue that reactions to such symbolic cues stem from heightened emotional responses and perceptions of rapid social change, particularly among historically privileged groups. We examine these dynamics using gender-inclusive language (GIL) as a paradigmatic case. Combining an observational study of social media reactions to German politicians’ tweets with two vignette experiments, we assess affective and threat responses to GIL. On social media, GIL elicits emotional polarization, generating both positive and negative reactions. In contrast, experimental evidence reveals a clear backlash: GIL increases negative emotions and perceptions that social change is occurring too quickly. These effects are strongest when GIL is used by women politicians and among lower- and middle-income men. Our findings advance understanding of symbolic political communication and highlight the affective foundations of contemporary cultural conflict.

Tobias Widmann. How Political Discourse Erodes: Radical-Right Success and the Changing Composition of Online Political Debates.

Radical-right electoral breakthroughs are widely thought to erode rhetorical norms by shifting what citizens perceive as publicly acceptable. Yet we still do not know whether aggregate deterioration in online discourse reflects behavioral change within individuals or changes in who participates. Leveraging the parliamentary entry of the Alternative für Deutschland (AfD) across Germany’s 16 states, I analyze 3.8 million citizen replies to state-level party tweets between 2014 and 2020 using staggered difference-in-differences designs. Transformer-based models measure sentiment, negative moral-emotional appeals, and toxic language at scale. I find that discourse becomes significantly more negative, more morally-emotionally charged, and more toxic following AfD entry. Yet within-user analyses show that continuing participants do not become more hostile; if anything, they become more positive. Aggregate deterioration is driven primarily by compositional change: the entry of new, more hostile participants into political discussion. Radical-right success thus reshapes democratic discourse mainly through selection rather than behavioral change.