A Spatiotemporal Model Based Governance Performance Evaluation in the US During the Pandemic: Nonpharmaceutical Interventions or Else?

Keywords: Balanced governance, Performance evaluation, Spatiotemporal data analysis, Bayesian hierarchical spatiotemporal modeling, COVID-19, Nonpharmaceutical Interventions
Abstract Type: Virtual Paper Abstract

Abstract

Scientific evidence suggests that nonpharmaceutical interventions effectively curb the spread of the COVID-19 prior to a pharmaceutical solution. Implementing these interventions also significantly impacts regular socioeconomic activities and practices of social, racial, and political justice. Local governments often face conflicting goals during policymaking. Striking a balance among competing goals during a global pandemic is a fine science of governance. How well state governments consume the scientific evidence and maintain such a balance remains less understood. This study employs advanced spatial (spatial panel regression) and spatiotemporal (Bayesian hierarchical spatiotemporally varying coefficient process) models to evaluate how state governments in the US use the scientific evidence to balance the fighting against the spread of COVID-19 disease and socioeconomic, racial, and other demands. We modeled the relationships between five nonpharmaceutical intervention strategies and eight COVID-19 caseload information and used the modeled result to perform balanced governance evaluation. The spatial panel regression model found that restaurant closure, nonessential business closure, and mask mandates curb the spread effectively. However, the Bayesian hierarchical spatiotemporally varying coefficient process model produces 612,000 spatiotemporally varying coefficients that suggest all measures work somewhere at some times. Summarized results suggest that states emphasizing NPIs fare well in curbing the spread of COVID 19. With over a million deaths due to COVID in the past two and half years in the US, we feel the scale of the balance likely needs to tip towards preserving human lives. This study aims to provide decision support for policymaking during a national emergency.