Causality and Dependence of COVID-19 Variables from Our World in Data

Abstract

The existing literature on COVID-19 has been dominated by medical sciences as well as epidemiological studies related to different public health reasons or disease related issues. Using the data set of 8 variables of Roser et al. (2020) from Our World in Data we have tested several propositions from conventional wisdom by using Pearson’s Product Moment Correlation, Panel Granger Causality test and Multivariate Bayesian VAR on 209 countries for 143 daily data during COVID19 outbreak in 2020 (31 December 2019-21 May 2020) in an unbalanced panel framework. It has been found that only a small fraction of new confirmed cases are being explained by new test conducted. Only a small fraction of new deaths due to COVID-19 is being explained by new confirmed cases. Stringency index neither contains contagion nor pacifies death rate or mortality. Even though bivariate analysis reveals some support to ‘more test more death hypothesis’ it has become weaker when we switch over to multivariate analysis. Urgent invention of a suitable vaccine is the only viable solution for fighting against this pandemic and save the humanity from this calamity.

Note: Funding: The paper is fully self-financed. We have not received any funding from any agency.

Declaration of Interest: We also declare that there is no conflict of interest from the author or co-author with any person or institution regarding this paper.

Keywords: COVID-19, Corona Virus; Our World in Data; Granger Causality; Bayesian VAR; Dependence; Impulse Response Function; Variance Decomposition

JEL Classification: I1