Agent-Based Modeling

A computational approach to understanding complex systems by simulating the behavior of individual agents following simple rules—and observing the emergent system-level patterns that arise from their interactions. Rather than modeling the system top-down (macro equations), agent-based models build it bottom-up: define agents, give them rules, let them interact, and see what emerges. Schelling's segregation model showed that agents with only mild same-group preference produce extreme neighborhood segregation—an emergent outcome no agent intended. Traffic jams, market crashes, and organizational culture all emerge from agent-level behaviors in ways that aren't predictable from individual rules alone.

When to use it

When system-level outcomes seem disconnected from individual intentions; when top-down policy interventions aren't producing expected results; when understanding how individual incentives aggregate into collective patterns; when designing systems where emergent behavior matters more than designed behavior.

How it can help

Think about complex problems agent-by-agent rather than system-by-system. Instead of asking 'why does the market behave this way?', ask 'what are individual actors doing, and what happens when they all do it simultaneously?' For organizational design: instead of mandating a culture top-down, understand what individual incentives produce when everyone follows them. The agent-based mindset reveals unintended emergent consequences: 'if every manager optimizes for their own team, what happens to the whole organization?' This bottom-up thinking often reveals why well-intended policies produce perverse system-level outcomes.