The scholarly debate about the influence of population density on COVID-19 spread points to a question: whether it is a larger population density, or a larger size of the population, that actually accelerate the spread of the virus . To figure out an answer in the US context, proper considerations should be taken to deal with three highly-influential determinants of the shape of a COVID-19 curve: the timeline of policy interventions, the metro and non-metro division, and the phase of pandemic. To safely unmask the effect of population size and density at county level, I introduce a group of “seasonal surges” and “COVID-19 policy reaction” variables, which measure to what extent a pandemic surge happened in a season, and whether the surge was followed by effective policy intervention within the season. Besides, a group of interaction variables based on the division of metro and non-metro counties are added to address some socio-cultural differences. To generally interpret the results, population density positively correlates with COVID-19 spread, while population negatively correlates with COVID-19 spread.

However, in the early phase of pandemic, density had negative impact only in metro counties, although in later phases the effect of density no longer differed between metro and non-metro counties. The negative impact of population on COVID-19 cases are most observable in non-metro counties, while its coefficient for metro counties was evidently smaller.