Research

How do social ties between administrators and defendants affect the implementation of personnel transitional justice policies? To address issues of inherited personnel, countries undergoing transitions from autocratic rule frequently implement mechanisms like purges or criminal trials. While a considerable number of studies investigates the consequences of these policies, less is known about factors that shape their implementation. Building on insights from administration research, I analyze how differences in social ties shape the outcomes of a mass purge by drawing on data of approx. 50000 denazification trials from British occupied Nazi Germany. I investigate the effects of two different forms of ties: social identity and networks. I leverage variation regarding the level of ties between administrators and defendants resulting from the delegation of denazification to German authorities and defendants' different migratory backgrounds. My results point to an ambivalent effect of ties in the context of transitional justice policies: while identity ties lead to more lenient ruling, network ties seem to have the opposite effect. The study's findings shed new light on transitional justice and the persistence of political personnel in the shadow of dictatorship.

This study investigates how individual beliefs regarding punitive transitional justice measures shape the support for policies of remembrance. While previous research on transitional justice has predominantly focused on its immediate consequences, less is known about long-term consequences and the potential for unintended side-effects. We investigate such potential "hidden long-term costs" of punitive transitional justice in the case of the Denazification of Germany after WWII. We implement an online survey in a sample of approx. 2000 German voters. To assess the effect of different perceptions regarding the extent of punitive transitional justice, we (1) manipulate respondents' beliefs regarding the degree to which Nazi perpetrators have been sanctioned through Denazification in a survey experiment and (2) exploit historical variation in the intensity of Denazification across former occupation zones.

How do voters react when self-serving collective memories are challenged? We study the controversial first “Wehrmacht Exhibition” (1995–1999), which, for the first time, exposed the German public to gruesome details about war crimes committed by the German military during WWII. The traveling exhibition challenged the ‘myth of the clean Wehrmacht’: the widespread but incorrect post-war narrative that Adolf Hitler and the Nazi Party alone designed the war of annihilation and that war crimes were only committed by the SS, not by the German military. To study how different segments of the electorate reacted to the exhibition, we leverage high-frequency survey data of more than 160,000 voters in a staggered difference-in-differences setup. We find that the exhibition triggered fierce political backlash. The social democratic SPD - the main proponent of the exhibition - loses about eight percentage points in electoral support following the opening of the exhibition in a given locality. These results are driven by the sons of WWII veterans, especially those cohorts whose fathers were most likely to have been killed in action or died as prisoners of war. We support these analyses with qualitative evidence drawn from over 1.200 letters to the editor commenting on the exhibition.

How does right-wing terrorism affect electoral support for populist radical right parties (PRRP)? Recent research has produced contrary answers to this question. We argue that only high-profile attacks, whose motives and targets mirror PRRPs’ nativist agenda are likely to generate a media backlash that dampens electoral support for PRRPs. We test this argu- ment combining high-frequency survey and social media data with a natural and survey experimental design. We find that right-wing terror reduced support for the right-wing populist party Alternative für Deutschland after one of the most intense nativist attacks in recent German history. A case comparison with three other high-profile attacks and an analysis of all 98 fatal right-wing attacks in Germany between 1990 and 2020 support our argument. Our findings help to understand how political violence triggers partisan detach- ment and have important implications for media responsibility in the aftermath of terrorist attacks.

How do voters respond to information about corruption? Research investigating electoral responses to corruption has thus far generated mixed results and research in this area generally faces a number of methodological problems. While observational studies linking aggregate corruption levels and political behavior usually raise concerns about endogeneity, experimental methods may suffer from problems like social desirability bias or a lack of external validity. In this paper, I bypass these issues by analyzing a natural experiment generated by a corruption scandal around the procurement of face masks during the COVID-19 pandemic in Germany. I exploit the differential exposure to the corruption scandal of mail and ballot-box voters prior to a major regional election in a difference-in-differences design. Comparing electoral outcomes across mail and ballot-box electorates in 1109 municipalities over time, I find that corruption leads to a reduction of the affected party's vote share of approximately 5% in the exposed electorate. Furthermore, I find similar effects for the party's coalition partner indicating that corruption may produce negative spillovers for other incumbent parties. This study sheds new light on corruption and voting behavior by showing that misconduct by public officials can produce substantive electoral responses that have the potential to move well beyond an affected party.