Background: Understanding public perceptions of government responses to COVID-19 may foster improved public cooperation. Trust in government and population risk of exposure may influence public perception of the response. Other population-level characteristics, such as country socio-economic development, COVID-19 morbidity and mortality, and degree of democratic government, may influence perception.
Methods and findings: We developed a novel ten-item instrument that asks respondents to rate key aspects of their government's response to the pandemic (COVID-SCORE). We examined whether the results varied by gender, age group, education level, and monthly income. We also examined the internal and external validity of the index using appropriate predefined variables. To test for dimensionality of the results, we used a principal component analysis (PCA) for the ten survey items. We found that Cronbach's alpha was 0.92 and that the first component of the PCA explained 60% of variance with the remaining factors having eigenvalues below 1, strongly indicating that the tool is both reliable and unidimensional. Based on responses from 13,426 people randomly selected from the general population in 19 countries, the mean national scores ranged from 35.76 (Ecuador) to 80.48 (China) out of a maximum of 100 points. Heterogeneity in responses was observed across age, gender, education and income with the greatest amount of heterogeneity observed between countries. National scores correlated with respondents' reported levels of trust in government and with country-level COVID-19 mortality rates.
Conclusions: The COVID-SCORE survey instrument demonstrated satisfactory validity. It may help governments more effectively engage constituents in current and future efforts to control COVID-19. Additional country-specific assessment should be undertaken to measure trends over time and the public perceptions of key aspects of government responses in other countries.