Are Affective and Attitudinal Polarization Causally Linked? Experimental Evidence from the U.S.

Abstract: Partisanship is a dominant force in contemporary U.S. politics. Prior research distinguishes between affective polarization—hostility toward opposing partisans—and attitudinal polarization, or the extremity of policy attitudes. While partisan animosity has increased substantially over recent decades, trends in mass policy polarization are more mixed, and the causal relationship between these two dimensions remains unclear. This paper provides new evidence on their causal link using a large collection of experiments. We analyze 26 experimental interventions targeting policy attitudes (N = 4,892) and estimate their effects on partisan animosity, as well as 25 interventions targeting partisan animosity (N = 32,059) and estimate their effects on attitudinal extremity. Instrumental variable analyses indicate that changes in policy attitudes have negligible effects on partisan animosity, and reductions in animosity do not meaningfully moderate policy attitudes. These findings suggest that affective and attitudinal polarization are relatively independent phenomena that require distinct interventions.

Keywords: Persuasion; intergroup conflict; experiment