Exodus from the Pandemic: Measuring Demographic Resilience Of Cities Using Large-Scale Transaction Payment Data

Understanding the spatial and temporal evolution of migration flows as a consequence of disasters is essential to building effective public health and economic policies but challenging. We evaluate urban resilience in a demographic dimension to the COVID-19 pandemic using large-scale high-frequency transaction payment data by tracing the destination cities, volumes and types of spending flows for cardholders whose habitual residence is the city of Wuhan before the pandemic. Our estimates show a large decline in consumer spending due to migration from Wuhan to other domestic cities following the pandemic outbreak, accounting for approximately nine percent of the GDP in 2020. Thereafter, we employ a difference-in-difference strategy to test the gravity theory of migration by exploring the spatial and temporal evolution of demographic resilience. Our results show that both economic mass and distance matter for demographic resilience in Wuhan: migration to economically similar cities in neighboring provinces and eastern coastal areas dominate following the easing of the lockdown and are more likely to be permanent, while migration to higher-ranked and neighboring cities show significant resilience to support the consumption recovery. Our proposed measure of demographic resilience of cities could be utilized to inform public health and economic policy planning.