The effects of non-pharmaceutical interventions on COVID-19-related mortality: A generalized synthetic control approach across 169 countries

Background: Most governments have introduced various non-pharmaceutical interventions (NPIs) in response to the pandemic outbreak of Coronavirus disease (COVID-19) since early 2020. While NPIs aim at avoiding fatalities related to COVID-19, the previous literature on their efficacy has focused on infections and on data of the first half of 2020. Still, findings of early NPI studies may be subject to underreporting and missing timeliness of reporting of cases. Moreover, the low variation in treatment timing during the first wave makes identification of robust treatment effects difficult. Methods: To circumvent problems of reporting and treatment variation, we analyse data on daily confirmed COVID-19-related deaths per capita from Our World in Data, and on 10 different NPIs from the Oxford COVID-19 Government Response Tracker for 169 countries from 1st July 2020 to 31st May 2021. To identify the causal effects of introducing NPIs on COVID-19-related confirmed fatalities per capita, we apply the generalized synthetic control (GSC) method to each NPI, while controlling for the remaining NPIs, weather conditions, vaccinations, and NPI-residualized COVID-19 cases. Findings: We do not find substantial and consistent mitigating effects of any NPI under investigation on COVID-19-related deaths per capita. We see a tentative change in the trend of COVID-19-related deaths around 30 days after workplace closing, public transport closing, and stay at home rules have been implemented, but none of them exerts a statistically significant effect. Interpretation: The study enhances the literature on the effectivity of NPIs with respect to the time frame, the number of countries, and the analytical approach. The results provide further guidance to judge the proportionality of NPIs.