Static and dynamic inefficiencies in an optimizing model of epidemics

Several externalities arise when agents shield optimally to avoid infection during an epidemic. We classify externalities into static and dynamic and compare the decentralized and optimal solutions when agents derive utility from social interaction. For low infection costs agents shield too little; for high costs they shield too much because of a "rat race to shield": they delay social action until other agents contract the disease and society reaches herd immunity. Other externalities drive more wedges between the private and social outcomes. The expectation of a fully effective vaccine that ends the disease faster changes results, reversing excessive shielding.