Partisan Endorsement Experiments Do Not Affect Mass Opinion on COVID-19

Abstract

The partisan politics and polarized messaging surrounding COVID-19 have attracted wide interest. We present the findings of a novel survey experiment, fielded March 21-23, 2020, on a nationally-representative sample of Americans. We found no statistically significant effects of partisan endorsements or messaging from President Trump on a wide range of health behaviors and policy attitudes. We speculate on potential explanations for these null results, such as America’s saturated media environment or heterogeneous effects by party. Our results suggest that priming experiments face serious obstacles when implemented at the same time as a national crisis is unfolding.

Keywords: COVID19, coronavirus, endorsement experiments, survey experiments, partisanship, polarization, Trump