Vaccine Adoption with Outgroup Aversion Using Cleveland Area Data

Innovation adoption can vary by group membership based on identity signaling by adopters or non-adopters in each group. In this study an agent-based model is developed that simulates innovation diffusion of Covid-19 vaccine adoption in the counties surrounding and including Cleveland, Ohio. The derivation of the model emanates from established models of innovation diffusion, outgroup aversion and statistical evidence of association between political affiliation and vaccine behavior. The model enables understanding the impact of several factors related to the outgroup aversion effects on innovation adoption and the resultant polarization. These factors include the proportion of never-adopters in each group, the advertising or innovation effect and the imitation effect with an outgroup factor that amplifies or de-amplifies adoption based on observations of other neighbors and their adoption status and group membership. The model presented here is meant to be a starting point for refinement. While the model directionally matches summary vaccine data in the counties studied, it is limited by the availability of data required for each of its parameters. Despite this limitation, the model can be refined for future analysis of similar innovation adoption scenarios, be a platform for integrating other decision factors, such as opinion dynamics, and can be adapted to other innovation scenarios with geographic dimensions.