Modeling of COVID-19 Panel Time-series Data Based on Generalized Method of Moments

Abstract

The main objective of the chapter is to investigate the dynamic relationships between the number of COVID-19 infected cases and deaths in all the districts of Karnataka state, India, during July 2020 to December 2021 based on the Arellano-Bond estimator using the generalized method of moments (GMM). The panel GMM model with the first difference transformation was found suitable for studying the dynamics of the number of deaths due to COVID-19 infections. For analyzing the dynamics of the number of deaths caused by COVID-19 infections over time, the panel GMM model with the first difference transformation was shown to be useful. The one-period lag (DEATHS(-1)) has a positive and significant effect on the number of deaths (DEATH). The panel GMM method with the first difference transformation has been calculated and presented in Table 15. R² is not used as a statistical standard for determining the model's goodness of fit, but the J-statistics assess the validity of the instrument variable used in the model. The Wald test strengthens the model's explanatory ability and validates the relevance of the coefficients. The number of fatalities at time t is positively associated with the number of fatalities during the preceding time. Additionally, the number of infected cases has a positive and considerable long-term impact on the death rate. Granger pairwise causality test reveals the existence of bi-directional causality relationships between the COVID-19 infected cases and deaths.