The Role of Demographic and Environmental Factors in the Outbreak of COVID-19 Across Italian Provinces

This research proposes an analysis of the spread of COVID-19 during the three waves at the provincial level in Italy. Specifically, this study examines the distribution of COVID-19 cases across Italian provinces to determine the potential existence of territorial clusters, which are measured by spatial autocorrelation among the provinces. Furthermore, this study examines whether a convergence process occurred in the contagion of the virus among the provinces, thereby estimating its relative speed. To this end, the β-convergence model—commonly applied in economic growth models—is employed in this study to consider the demographic and environmental variables present in each province. The initial level and growth of infections observed in a certain province are then related to the level of infections and their growth rate in all other Italian provinces. This econometric estimation is conducted for the first, second, and third waves of the pandemic between March 2020 and June 2021, using covariates such as population density, aging index, average temperature, pollution levels, and the duration of the waves. The hypothesis of COVID-19 convergence, verified in all three waves, recorded a rather different average speed, with an influence exerted by demographic and environmental variables. Additionally, the econometric analysis is integrated, considering spatial effects using the Spatial Autoregressive (SAR) and Spatial Error (SEM) models, after verifying the existence of autocorrelation of the residuals for the first and third waves. The results obtained in this second estimation only partially confirm the influence of demographic and environmental variables for the first and third waves. In the first wave—in addition to identifying any spillover effects among provinces—the effects are decomposed into direct, indirect, and total effects.