A missing behavioural feedback in COVID-19 models is the key to several puzzles

Summary: ⇒ Human actions have played a key role in shaping the COVID-19 pandemic patterns. While theoretically recognised, existing models of epidemics often do not endogenously capture many of the feedback loops connecting people’s choices and epidemic dynamics, for example, adoption of non-pharmaceutical interventions (NPIs) by individuals and governments shapes disease transmission, which in turn alters perceived risks and future NPI adoption.

⇒ Such ‘risk-driven response’ feedback is central to explaining important empirical puzzles of the COVID-19 pandemic, including the convergence of reproduction number to 1 across nations, multiple waves of pandemic, mortality variance and limited trade-off between economic and health outcomes in adoption of NPIs. Capturing that feedback also enhances pandemic forecasting and offers distinct and more effective vaccination strategies.

⇒ Much remains to be explored in modelling diverse behavioural feedbacks, from endogenous testing and vaccination choices to the building of infrastructure for various responses. Integrating those with epidemiological models offers promising new discoveries and enhanced policy design.