Policy Effectiveness on the Global COVID-19 Pandemic and Unemployment Outcomes : A Large Mixed Frequency Spatial Approach

We measure the effectiveness of policies in a mixed frequency spatial VAR (MF-SVAR) modeling framework, conditional on the spillover and diffusion effects of the global pandemic and unemployment. Specifically, we examine two aspects of policy effectiveness from a spatio-temporal perspective, namely the impact of policy start dates and policy timeliness on Covid-19 weekly new case growth and monthly changes in unemployment rates for 68 countries across six continents from January 2020 to August 2021. We find that government policies have a significant impact on the growth of new cases, but only a marginal effect on the change in unemployment rates. A policy's start date is critical for its effectiveness. In terms of both immediate impact on the near term and total impact over the following four weeks, starting a policy in the 4th week of a month is most effective at reducing the growth of new cases. At the same time, starting in the 2nd or 3rd week can be counterproductive. In addition, our estimates suggest that the spillover and diffusion effects are much stronger than a country's temporal effect during a global pandemic, both for new case growth and changes in unemployment. We also find that new case growth influences changes in unemployment, but not vice versa. Counterfactual experiments provide further evidence of policy effectiveness in various scenarios and also reveal the main risk-vulnerable and risk-spillover countries during pandemic