State Media Labels Increase Perceived Accuracy of News Content from Democratic Sources *

Abstract

Social media platforms like Facebook and YouTube apply state-media affiliation labels to flag content from government-backed sources. In a preregistered survey experiment, we examine how such labels affect users’ perceptions of content accuracy across both true and false news headlines. We find that content labeled as originating from democratic governments is perceived as more accurate, regardless of its factual accuracy. In contrast, labels indicating affiliation with non-democratic governments have no significant effect. These findings challenge the assumption that state-media labels promote informed engagement. Instead, they appear to activate motivated reasoning: users trust content more when it aligns with their favorable priors about democratic regimes. As platforms experiment with labeling strategies to combat misinformation, our results highlight the need to account for cognitive biases that shape credibility judgments and may undermine the intended effects of transparency efforts.