The Bass Model: a parsimonious and accurate approach to forecasting mortality caused by COVID-19

Abstract

Purpose The novel coronavirus disease 2019 (COVID-19) pandemic has presented unique challenges in terms of understanding its unique characteristics of transmission and predicting its spread. The purpose of this study is to present a simple, parsimonious and accurate model for forecasting mortality caused by COVID-19. Design/methodology/approach The presented Bass Model is compared it with several alternative existing models for forecasting the spread of COVID-19. This study calibrates the model for deaths for the period, March 21 to April 30 for the USA as a whole and as the US States of New York, California and West Virginia. The daily data from the COVID-19 Tracking Project has been used, which is a volunteer organization launched from The Atlantic. Every day, data is collected on testing and patient outcomes from all the 50 states, 5 territories and the District of Columbia. This data set is widely used by policymakers and scholars. The fit of the model (F-value and its significance, R-squared value) and the statistical significance of the variables (t-values) for each one of the four estimates are examined. This study also examines the forecast of deaths for a three-day period, May 1 to 3 for each one of the four estimates – US, and States of New York, California and West Virginia. Based on these metrics, the viability of the Bass Model is assessed. The dependent variable is the number of deaths, and the two independent variables are cumulative number of deaths and its squared value. Findings The findings of this paper show that compared to other forecasting methods, the Bass Model performs remarkably well. In fact, it may even be argued that the Bass Model does better with its forecast. The calibration of models for deaths in the USA, and States of New York, California and West Virginia are all found to be significant. The F values are large and the significance of the F values is low, that is, the probability that the model is wrong is very miniscule. The fit as measured by R-squared is also robust. Further, each of the two independent variables is highly significant in each of the four model calibrations. These forecasts also approximate the actual numbers reasonably well. Research limitations/implications This study illustrates the applicability of the Bass Model to estimate the diffusion of COVID-19 with some preliminary but important empirical analyses. This study argues that while the more sophisticated models may produce slightly better estimates, the Bass model produces robust and reasonably accurate estimates given the extreme parsimony of the model. Future research may investigate applications of the Bass Model for pandemic management using additional variables and other theoretical lenses. Practical implications The Bass Model offers effective forecasting of mortality resulting from COVID-19 to help understand how the curve can be flattened, how hospital capacity could be overwhelmed and how fatality rates might climb based on time and geography in the upcoming weeks and months. Originality/value This paper demonstrates the efficacy of the Bass Model as a parsimonious, accessible and theory-based approach that can predict the mortality rates of COVID-19 with minimal data requirements, simple calibration and accessible decision calculus. For all these reasons, this paper recommends further and continued examination of the Bass Model as an instrument for forecasting COVID-19 (and other epidemic/pandemic) mortality and health resource requirements. As this paper has demonstrated, there is much promise in this model.