- Thursday 24 Mar 2022, 12:00 - 13:00
- Spoken Language
We propose a mixed-frequency spatial VAR (MF-SVAR) modelling framework to measure the policy effectiveness conditional on the spillover and diffusion effects of the global pandemic and unemployment interactions.
We study two aspects of policy effectiveness, the impact of policy effective date and the policy timeliness, from a spatial temporal aspect. The data are mixed frequencies of weekly new case growths and monthly unemployment changes of 68 countries in 6 continents from January 2020 to August 2021. We found that government policy shows significant effect on new case growths, but marginal on unemployment changes.
The effective date of policy plays a key role. When launching in the last week of a month, policies are the most effective in reducing the new case growth rates over the next weeks. Moving to the first week of a month, the effectiveness decreases. It turns to be counterproductive when the effective dates of policies move to the second and third weeks. The lifespan of policy is about 2 weeks, after which effectiveness diminishes. Moreover, our estimates imply that the contagious or spillover effect and the diffusion effect are much stronger than the temporal effect of a country, for both the new case growths and unemployment changes, during the global pandemic.
We also found that the new case growths have influence on the unemployment rate changes, but not vice versa. The counterfactual experiments further demonstrate the policy effectiveness under various scenarios, which also show that the United States and India are the major risk exposurers with the largest direct effect on new case growths, if policy were mildly loosened in the countries. The United Kingdom, on the other hand, is the major risk spillover country with the largest indirect effect of inducing more new cases into other countries, among which the American countries are the most vulnerable.
This is a joint work with Xiaoyi Han, Yanli Zhu and Yijiong Zhang.