Enhancing Long-term Forecasting: Learning from COVID-19 Models

Abstract

Long-term forecasts are hard, but also indispensable in personal and policy planning. How could long-term predictions of complex phenomena, such as COVID-19 contagion, be enhanced? While much effort has gone into building predictive models of the pandemic, some have argued that early exponential growth combined with the stochastic nature of epidemics make the long-term prediction of contagion trajectories impossible. We leverage the diverse models contributing to CDC repository of COVID-19 death projections to identify factors associated with prediction accuracy across different projection horizons. We find that better long-term predictions correlate with (1) capturing the physics of transmission (instead of using black-box models); (2) projecting human behavioral reactions to an evolving pandemic; and (3) resetting state variables to account for randomness not captured in the model before starting projection. A very simple model, SEIRb, that incorporates these features and includes few other nuances offers predictions comparable with the most accurate models in the CDC set. Key to the long-term predictive power of multi-wave COVID-19 trajectories is endogenously capturing behavioral responses: balancing feedbacks where the perceived risk of death continuously changes transmission rate through the adoption, and relaxation, of various Non-Pharmaceutical Interventions (NPIs).

Note: Funding: The authors received no specific funding for this work.

Declaration of Interests: The authors declare no competing interests.

Keywords: epidemics, forecast, SEIR, behavior change, system dynamics