“Unmotivated bias” and partisan hostility: Empirical evidence

Abstract

Extreme partisan animosity has been on the rise in the US and is prevalent around the world. This hostility is typically attributed to social group identity, motivated reasoning, or a combination thereof. In this paper, I empirically examine a novel contributing factor: the “unmotivated” cognitive bias of overprecision (overconfidence in precision of beliefs). Overprecision could cause partisan hostility indirectly via inflated confidence in one’s own ideology, partisan identity, or perceptions of social distance between the parties. Overprecision could also cause this hostility directly by causing excessively strong inferences from observed information that is either skewed against the out-party or simply misunderstood. Using a nationally representative sample, I find consistent support for direct effects of overprecision and mixed support for indirect effects. The point estimates imply a one standard deviation increase in a respondent’s overprecision predicts as much as a 0.71 standard deviation decline in relative out-party favorability.