The Impact of Agent Activation on Population Behavior in an Agent-based Model of Civil Revolt

In creating agent-based models of complex adaptive systems, the model designer must specify, as part of the design, the conditions and sequence with which agents activate – sensing their environment and updating their state. Activation can be synchronous (all agents execute simultaneously) or asynchronous, and the latter can be uniform (turn-based and shuffled), random, or Poisson (heterogeneous activation rate dependent on state). In a replication of a well-documented model of civil unrest, a statistically significant difference in emergent population behavior is demonstrated as a consequence of different activation patterns.