Although COVID-19 emerged as a global shock, governments adopted non-pharmaceutical policy responses that were rather heterogeneous, depending on cultural and institutional characteristics. At the country level, the stringency of ‘lockdown’-type policies should be set to achieve the best possible trade-off between economic and fatality dynamics, obviously accounting for possible cross-border influences. To allow for policy learning, I assume that the first country implementing a policy initiative that is worth emulating must either get the best possible health or the best possible economic outcome. I propose a combination of sign and magnitude restrictions, embedded in a global VAR model, to identify idiosyncratic policy shocks that spill over and influence policy responses abroad. Once policy shocks are identified, I run a comparison exercise between two model specifications, i.e. with and without policy emulation. Within a given a sample, this methodology can be used to find when and where policy lessons can be identified. I find that, among 17 developed and developing countries, few can offer lessons based on their policy initiatives, but several others might get better trade-offs through policy emulation, although in reality this outcome is not guaranteed to have occurred.
- COVID-19, policy stringency, sign and magnitude restrictions, GVAR