Climate influences on COVID-19 prevalence rates: An application of a panel data spatial model

The present study aims to measure the impact of climate characteristics on the prevalence rate of coronavirus disease 2019 (COVID-19) in Brazilian states given the exogenous nature of these variables. We used a daily panel for the period from March 10 to April 10, 2020, the first phase of the pandemic, as there were few intervention policies to contain the spread of COVID-19 during that period, and it was estimated through generalized least squares (GLS) spatial models to control the presence of spatial spillover, first-order autoregressive errors, and correlation between cross-sections. Considering the COVID-19 incubation period and the time it takes for COVID-19 symptoms to manifest, the econometric models were estimated using the 14-, 11-, and 7-day moving averages of the climate variables. The results showed that increases of 1% in the solar incidence, average temperature, and relative humidity of the air reduced COVID-19 prevalence rates by 0.16%, 0.049%, and 0.22%, respectively, considering the 11-day moving average.