Do the determinants of COVID-19 transmission differ by epidemic wave? Evidence from U.S. counties.

This paper uses data from the United States to examine determinants of the spread of COVID-19 during three different epidemic waves. We address how sociodemographic and economic attributes, industry composition, density, crowding in housing, and COVID-19-related variables are associated with the transmission of COVID-19. After controlling for spatial autocorrelation, our findings indicate that the percentage of people in poverty, number of restaurants, and percentage of workers teleworking were associated with the COVID-19 incidence rate during all three waves. Our results also show that dense areas were more vulnerable to the transmission of COVID-19 after the first epidemic wave. Regarding the density of supermarkets, our study elaborates the negative aspects of wholesale retail stores, which likely provide a vulnerable place for virus transmission. Our results suggest that sociodemographic and economic attributes were the determinants of the early phase of the pandemic, while density showed positive association with the transmission during subsequent waves. We provide implications for regions serving as gateway cities with high density and number of population. To add, we further provide evidence that non-pharmaceutical interventions in the early stage may mitigate the virus transmission.