Polarized Public Responses to Pandemic: A Homo Economicus’ Perspective; Audrey Xianhua Hu (City University of Hong Kong)

Abstract

In this paper, we seek to understand why, in a liberal society, the public response to pandemic may exhibit extreme divisions. We present a model of dynamic game with incomplete information in which a continuum of agents, each endowed with a private or imperfectly observable (multi-dimensional) type, choose communicable activities non-cooperatively. In the absence of government interventions, agents are free to choose any level of communicable activity at 2 [0; 1] in each period, ranging from "no communicable activity (at = 0)" to "full communicable activity (at = 1)." A higher level of at increases an agent's current period utility but also increases his probability of contracting the disease in the subsequent period.
We model transmission probability by extending the standard SIR model in three aspects that (1) the transmission and death rates can be stochastic, (2) the transmission can be attributed to unknowingly infected asymptomatic agents, and (3) the individuals' communicable activities mutually affect each others' probability of contracting the disease. In this context, we find that each agent's objective function is quasi-convex in his choice of at under the standard matching technologies. As a result, the equilibrium is characterized by polarized public responses: in every period agents choose either at = 0 or at = 1; or randomize between the two polar choices. The free-riding problem prevents all agents from choosing 0 in any period even when it is socially optimal to do so.