Agent-based computational economics

Agent-based computational economics (ACE) is a research paradigm that models economies as complex adaptive systems composed of interacting autonomous agents rather than as equilibrium systems. Each agent follows local rules, and aggregate behavior emerges from decentralized interactions without any assumption of market clearing or representative agent behavior. ACE grew out of the recognition that the representative agent assumption of mainstream economics systematically obscures the emergence of phenomena like bubbles, crashes, and persistent inequality. The paradigm is computationally intensive: it trades analytical tractability for empirical realism, and its results are often sensitive to initial conditions. From a systems-theoretic perspective, ACE is the natural extension of cybernetic thinking to economic systems — it treats the economy as an information-processing network rather than as an optimization problem. The field owes much to Thomas Schelling's pioneering segregation model and to researchers at the Santa Fe Institute who showed that simple local rules can produce globally complex dynamics. The central question is whether the computational cost is worth the gain in descriptive accuracy, and whether the field can develop standards that distinguish robust emergent properties from modeling artifacts.