Overview

We are forging a science of complex economies that captures the emergent behaviors, network properties, and self-organizing and evolutionary characteristics of financial systems.

Like ecosystems, economies are living networks of interacting agents. As they transform energy and materials to meet human needs, economies evolve, often in unforeseen ways. Current mainstream economic theories and models are, by and large, derived from the notion that unless disturbed, markets naturally and inexorably relax to an equilibrium state. But these methods can’t capture the turbulence that arises when hundreds of millions of very different people interact with and react to one another in roiling economy immersed in an uncertain world and dependent on the decision making of imperfect agents possessing only partial information. This is because traditional theories and models filter out the intractable complexities of market processes, confining the study of economics to a static problem set to be solved top-down, through analysis.

At the Santa Fe Institute we take a different approach to economic complexity. Using agent-based models, for example, we simulate financial markets from the bottom up, recreating the interactions of the myriad microscale decisions of people, households, and firms to observe the macro phenomena that result. We run these models hundreds or thousands of times to understand the probabilities of outcomes or to test economic policy interventions.

Beyond these models, our complex economies toolkit contains techniques from network science, game theory, behavioral science, biology, ecology, physics, and mathematics. With these tools, we ask how irrational human responses, at the individual level, impact the global market? What patterns give rise to persistent wealth inequality? How does risk accrue throughout the system and can it be controlled? Can we predict the next trillion-dollar financial crisis by spotting collective patterns in individual behaviors?

Only when we embrace and respect economic complexity will we be able to predict the behaviors of real markets and begin the design of effective policy interventions and regulatory controls.