Impact of COVID-19 Containment Measures on Unemployment: A Multi-country Analysis Using a Difference-in-Differences Framework

Background: At the start of the coronavirus disease 2019 (COVID-19) pandemic, in the absence of pharmaceutical interventions, countries resorted to containment measures to stem the spread of the disease. In this paper, we have conducted a global study using a sample of 46 countries to evaluate whether these containment measures resulted in unemployment. Methods: We use a difference-in-differences (DID) specification with a heterogenous intervention to show the varying intensity effect of containment measures on unemployment, on a sample of 46 countries. We explain variations in unemployment from January-June 2020 using stringency of containment measures, controlling for gross domestic product (GDP) growth, inflation rate, exports, cases of COVID-19 per million, COVID-19-specific fiscal spending, time fixed effects, region fixed effects, and region trends. We conduct further subset analyses by COVID-cases quintiles and gross national income (GNI) per capita quintiles. Results: The median level of containment stringency in our sample was 43.7. Our model found that increasing stringency to this level would result in unemployment increasing by 1.87 percentage points (or 1.67 pp, after controlling for confounding). For countries with below median COVID-19 cases and below median GNI per capita, this effect is larger. Conclusion: Containment measures have a strong impact on unemployment. This effect is larger in poorer countries and countries with low COVID-19 cases. Given that unemployment has profound effects on mortality and morbidity, this consequence of containment measures may compound the adverse health effects of the pandemic for the most vulnerable groups. It is necessary for governments to consider this in future pandemic management, and to attempt to alleviate the impact of containment measures via effective fiscal spending.