The Complexity of Turnout: An Agent-Based Simulation of Turnout Cascades

Using an agent-based simulation of electoral participation this paper examines the cascade or second order effects of an exogenous increase in turnout such as a GOTV campaign. Unlike most previous applications of agent-based modelling in political science, we adopt a descriptively complex rather than a simple model of behaviour. Thus factors affecting the turnout of agents include individual-level characteristics (e.g., socio-demographic characteristics), socialisation processes, habit and inertia, ‘rational behaviour’, political interest, social norms, political discussion, and interpersonal and party mobilisation. The experimental focus of the simulations is to gauge the direct and indirect effects of campaign mobilisation on individuals and others in their social networks. Campaign effects are separated into those stemming from the long campaign (socialising and persuading voters) and the short campaign (voter mobilisation). Experimenting with different levels of mobilisation allows estimation of the extent to which mobilisation spreads beyond the primary mobilisation target both across the population and whether or not it survives over time. The article compares results based on a highly complex simulation with those derived from one of the few earlier studies using a simpler simulation model of a voting cascade (Fowler, 2005). We argue that in a more complex world, rather than producing simple geometric increases in turnout, cascades can become truncated or overwhelmed by other factors, often affecting those who might have voted anyway, and even crowding out the effects of peer-to-peer mobilization.