Buying Time to Save Lives: Evaluating COVID-19 Lockdown Policies Using a Behavioral SEIR Model

This thesis provides a behavioral SEIR model for the ongoing COVID-19 pandemic, and evaluates various lockdown policies based on the saved value of statistical lives. The behavioral SEIR model introduces new and more transmissible variants, vaccines with stochastic arrival time, seasonality feature of virus transmission, and the behavioral change due to fear of contagion. Individuals update their beliefs on the COVID-19 mortality rate over time based on up-to-date information, creating a time lag between belief and the actual time path of the death rate. Based on the simulation from the behavioral SEIR model, we evaluate various lockdown policies by the saved value of statistical lives they offer, since lockdown policies grant more people the opportunity, or a binary option in effect, to get vaccinated in the future and avoid more deaths which is the ultimate form of irreversibility. The option value makes up a significant share of the total value generated by lockdowns and it depends on the uncertainty regarding the vaccine arrival rate and rollout speed. To save more lives and life years for its own population, the government is willing to pay more upfront foran earlier vaccine arrival and a faster vaccine rollout.