Lockdowns and more generally restrictions introduced by governments during the Covid-19 pandemic contributed to the disruption of international trade in goods. We show that bilateral trade declined more when lockdowns were initially introduced in Spring 2020 with a higher degree of stringency. Moreover, the quantitative impact of lockdowns on trade weakened in the second half of 2020. This is especially the case of lockdowns implemented in the exporting country, which have little or no significant impact on trade in the second half of 2020.
Figure 2 shows the decline in trade value attributable to decrease in value in goods still being traded (red), and termination or beginning of traded value (blue) – see notes.
Note: Authors calculations based on data from the Trade Data Monitor (TDM) for 31 reporting countries accounting for about 75% of global trade. Exports are measured in current USD. A trade relation is represented by an exporting country, an importing country and a product detailed at the 6-digit level of the Harmonized System of the product nomenclature of the United Nations. A “Continuing relation” corresponds to a trade relation that remains “active” (i.e. is represented by a positive value of trade) in a given month in two consecutive years (e.g. April 2019 and April 2020). The value of entry minus exit corresponds to the change in total bilateral exports, which is explained by new trade relations being created over twelve months, and trade relations being terminated. A positive value implies net trade creation, while a negative value implies net trade destruction. Based on Berthou and Stumpner (2021).
Figure 1 shows the estimated effect of lockdown stringency on imports and exports.
Note: Months of the year 2020. Coefficients (and 95% confidence intervals) estimated from export and import equations, where the main variable of interest is the degree of lockdown stringency in the partner country. Equations are estimated in a panel, and rely on trade data from the Trade Data Monitor and Lockdown stringency from the University of Oxford. For details, see Berthou and Stumpner (2020, forthcoming).