There is a large body of literature on police abuse and discrimination, but knowledge about public opinion on the topic in limited, particularly as it relates to citizen understandings of what constitutes police misconduct and whether misconduct will be punished through mechanisms of democratic accountability. We address this gap by mimicking the way in which citizens typically acquire information about possible police abuse: through the news media. Building on previous literature which finds that consumption of news media impacts on attitudes towards the police, we present results from a survey experiment fielded in Germany (n=16500) which was designed to capture how public attitudes on police accountability shifts as the informational landscape evolves. We hypothesize that individuals holding extremist political ideologies will be more resistant than respondents with moderate political beliefs to changing their opinions on possible police misconduct in light of new information. We test this argument in an experimental survey containing news media-style vignettes describing cases of possible police misconduct. Each vignette is divided into three stages which are designed to mimic real world news media practices. In the first stage, the vignette describes a police-citizen encounter, relying exclusively on police officials as the primary source for information. In the second stage, we mimic the additional reporting that typically occurs as journalists identify more varied sources, resulting in a contradictory account of the event. Finally, the third stage reflects the follow-up stories which are typically written long after events, and which report on outcomes of the event, in our case, a misconduct investigation. We ask about the same outcomes after each stage and use fixed-effects estimators to explain how individuals respond across the different information environments. This novel research design allows us to study how media reporting of a contentious issue is filtered through ideological beliefs, affecting public support for accountability mechanisms.
