Co-authored with Marianne Bertrand, University of Chicago Booth Business School, Maddalena Grignani, University of Chicago, Rustandy Center, and Salma Nassar, University of Chicago, Rustandy Center.
Crises can bring citizens together through shared experiences or can increase divisions through greater ideological polarization. In a multi-wave longitudinal panel survey, we track a representative sample of Americans throughout the COVID-19 pandemic to study what factors might increase convergence or polarization. We find that a direct experience with the crisis, such as losing a size- able portion of income or knowing someone hospitalized with the virus, has a marginal converging effect in increasing support for greater public spending on both welfare and temporary relief policies, and in decreasing confidence in institutions. Conversely, consuming politically biased media further polarized views on the same topics. We show that this effect is partly driven by a misunderstanding of the gravity of the crisis in the absence of a direct experience. In a survey experiment on the same sample of respondents, we show that a simple non-incentivized suggestion to correct for misinformation, namely checking the virus death rate on the C.D.C. website, increases accuracy of beliefs and understanding of the gravity of the crisis, with the effect persisting for several months.