A panel path analysis approach to the determinants of coronavirus disease 2019 transmission: does testing matter for confirmed cases?

Purpose The main purpose of this study is to examine the role of the coronavirus disease 2019 (COVID-19) test on transmission data globally to reveal the fact that the actual picture of transmission history cannot be exposed if the countries do not perform the test adequately. Design/methodology/approach Using Our World in Data for 212 countries and areas and 162 time periods daily from December 31, 2019, to June 09, 2020, on an unbalanced panel framework, we have developed a panel-based path analysis model to explore the interdependence of various actors of COVID-19 cases of transmission across the globe. After controlling for per capita gross domestic product (GDP), age structure and government stringency, we explore the proposition that COVID-19 tests affect transmission positively. As an anecdote, we also explore the direct, indirect and total effects of different potential determinants of transmission cases worldwide and gather an idea about each factor's relative role in a structural equation framework. Findings Using the panel path model, we find that a 1 standard deviation change in the number of tests results in a 0.70 standard deviation change in total cases per million after controlling for several variables like per capita GDP, government stringency and age population (above 65). Research limitations/implications It is not possible to get balanced data of COVID-19 for all the countries for all the periods. Similarly, the socioeconomic, political and demographic variables used in the model are not observed daily, and they are only available on an annual basis. Practical implications Countries which cannot afford to carry out more tests are also the countries where transmission rates are suppressed downward and negatively manipulated. Social implications Cross country collaboration in terms of COVID-19 test instruments, vaccination and technology transfer are urgently required. This collaboration may be sought as an alternative to foreign development assistance. Originality/value This article provides an alternative approach to modeling COVID-19 transmission through the panel path model where the test is considered as an endogenous determinant of transmission, and the endogeneity has been channeled through per capita GDP, government stringency and age structure without using any regression-based modeling like pooled ordinary least squares (OLS), fixed-effects, two-stage least squares or generalized method of moments (GMM). Endogeneity has been handled without using any instruments.