Individualism During Crises: Big Data Analytics of Collective Actions and Policy Compliance amid COVID-19
44 Pages Posted: 8 Jun 2020 Last revised: 18 Jun 2020
Date Written: June 5, 2020
Collective actions and government interventions are important measures to alleviate the negative impact of the COVID-19 pandemic. However, engagement in these actions and compliance to government policies vary significantly within the U.S. This study explains this heterogeneity through an understudied cultural dimension—individualism. Using novel big data analytics on a quarter petabyte of data, we present the first evidence on how individualism affects offline social distancing and online charitable crowdfunding around state-issued stay-at-home orders. Following economic history studies, we leverage GIS techniques to construct a U.S. county-level measure of individualism that traces the amounts of time counties spent at the frontier during the 1790-1890 period. We then use high-dimensional fixed-effect models, text mining, geo-analytics, and a novel migration-based identification strategy to analyze social distancing compliance and GoFundMe fundraising activities. Our analysis reveals that an interquartile increase in local individualism offsets 41% of the effect of state lockdown orders on social distancing and reduces COVID-related charitable fundraising by 48%. These effects are stronger in counties where social distancing generates larger externalities, suggesting our results are partly driven by individualism lowering the tendency to internalize the externality of their actions. Finally, we show that government interventions, such as stimulus checks, can mitigate the negative impact of individualism. Our study is the first to identify the downside of individualism during crises. It also demonstrates the importance of big data-driven, culture-aware policymaking.
Keywords: COVID-19, individualism, crises, social distancing, charitable crowdfunding, culture-aware policymaking, collective action
JEL Classification: D62, D64, D70, I10, I30
Suggested Citation: Suggested Citation