Optimal COVID-19 Quarantine and Testing Policies

We study quantitatively the optimality of quarantine and testing policies; and whether they are complements or substitutes. We extend the epidemiological SEIR model incorporating an information friction. Our main finding is that testing is a cost-efficient substitute for lockdowns, rendering them almost unnecessary. By identifying carriers, testing contains the spread of the virus without reducing output. Although the implementation requires widespread massive testing. As a byproduct, we show that two distinct optimal lockdown policy types arise: suppression, intended to eliminate the virus, and mitigation, concerned about flattening the curve. The choice between them is determined by a "hope for the cure" effect, arising due to either an expected vaccine or the belief that the virus can be eliminated. Conditional on the policy type, the intensity and duration are invariant to the welfare function's shape: they depend mostly on the virus dynamics.