A multi-level analysis of the spread of COVID-19.

This paper uses extensions of the traditional methods for evaluating panel data to evaluate the effect of Non-Pharmaceutical Interventions (NPI) on the spread of COVID-19. I utilize data from weather conditions, policy interventions, past outcomes, and political landscapes at the county level. These components allow me to navigate confounding issues with traditional models such as heterogeneity, endogeneity, and measurement error. The results of this model support the efficacy of policy interventions. I also find that poor weather conditions contribute to the spread of the disease, which indicates that the disease spreads less effectively outdoors. Finally, I find that the share of GOP voters in the previous election is positively associated with the spread of the disease. The ability to combine time variant and invariant components with minimal assumption, makes this model a helpful foundation for further research.