Effect of Behavioural Strategies on Individual Quality of Life in Agent-Based Social Simulations of COVID-19

Abstract

Since the onset of the COVID-19 pandemic, many models have been made to predict the spread and responses to it. Although moral decision-making during the uncertainty pandemics is suggested to be more motivated by individual incentive than collective incentive, decision-making in COVID-19 agent-based models is often modelled implicitly by cultural, social and economic characteristics, making it difficult to investigate the relation between decision-making and individual pay-off. The main research question of this paper is: ``What behavioural strategies can maximise an individual's quality of life in an agent-based social simulation of COVID-19?''. This is answered for two basic behavioural strategies, Isolate and Ignore, modelled in an existing agent-based model of COVID-19 called \textit{ASSOCC}. Patterns in quality of life per strategy are investigated by what they look like over time and how they are influenced by the degree of isolation and distribution of isolation strategies. The two main conclusions from the experiments in this paper are (1) isolation behaviour can be an acceptable or even a preferable strategy if and only if it is done full-time and collectively and (2) although collective isolation is preferable over collective non-isolation, there is an individual incentive not to join in on this collective isolation. These results can only be seen within the confines of the agent-based model and any proof towards its extension to real life is left for future work.