Three Essays in the Economics of Vaccination

Under-vaccination is a usual concern of disease control studies. However, the equilibrium vaccination pattern may not only deviates from social optimum by the number of vaccinations but also the type of individuals that gets vaccinated. This paper employs a vaccination game in a three-agent contagion network to show that it is only one of the three inefficient patterns. Chapter 2 shows that when the network structure is incomplete or individual characteristics are heterogeneous, there exist new types of Nash equilibrium outcomes with either the right number but wrong set of people getting vaccinated or too many vaccinations, and these equilibria are robust to standard refinements. However, if the pure-strategy Nash equilibrium assigns the vaccine to some agents but not to everyone, the vaccination game becomes anti-coordinative because everyone prefers the other agents to take the costly vaccine. Thus, agents achieve a mixed-strategy Nash equilibrium. Chapter 3 shows that in an incomplete network, the central player has a greater centrality but chooses a lower vaccination probability than the peripherals. These theoretical predictions differ from the traditional epidemiology models, and this paper tests the various models' predictions with U.S. workers' flu vaccine uptake. Chapter 4 finds that among all workers, high-centrality workers are more likely to get vaccinated than low-centrality workers. However, the difference diminishes as we exclude health care personnel. Both the theoretical and empirical results suggest that targeted vaccination policies on high-centrality people can improve social welfare.