In the 1980s and 1990s, the Santa Fe Institute (SFI) was a center for the development of agent-based modeling (ABM), a method that simulates economies by focusing on individual differences and interactions rather than averages. Over time, ABMs have been used in various fields, including economics, where they now model complex phenomena such as housing bubbles, supply chains, and pandemic impacts.
Despite these advances, mainstream economics has often sidelined ABMs because early models were highly stylized and did not address specific real-world economies. However, with the rise of detailed economic microdata and increased computing power, ABMs are becoming more effective at making predictions relevant to policy decisions.
To promote broader adoption of ABMs in economics, SFI organized its longest-ever working group this August. The event was held at the new Gurley Forum and brought together economists, central bankers, computer scientists, social scientists, physicists, and other researchers from August 4–29. The Zegar Family Foundation and Omidyar Network supported the initiative.
“Mainstream economic models have not been able to solve some big problems: what causes inflation? What determines interest rates? What are the main factors behind recessions? What forces drive inequality?” said SFI External Professor J. Doyne Farmer of the University of Oxford. “At the SFI working group, we developed an approach to building an agent-based macromodel that could finally resolve these paradoxes.” Farmer co-organized the event to address gaps in current models and expand their practical impact.
Participants discussed how ABMs can improve understanding of issues such as supply-chain dynamics, job vacancies, unemployment rates, and transitions to green energy. According to SFI External Faculty Fellow Rob Axtell from George Mason University: “As a result of the SFI working group, we can see our way to creating a model of a whole country’s economy in software.”
The meeting included economists from Columbia University, Johns Hopkins University, Cambridge University, L’Observatoire Français des Conjonctures Economiques (OFCE), as well as representatives from central banks in Italy, Canada, Hungary, and Chile who described their experiences using ABMs for policymaking purposes.
“We saw some of the first examples of agent-based modeling competing directly with the conventional approach,” said Axtell. “In the last year or two, several countries have built high-quality agent-based models for central banks and policymakers to use because they’re so good at ingesting microdata instead of statistical averages.”
The extended format allowed established experts to collaborate closely with newer researchers across disciplines. Jordan Kemp—who recently completed his Ph.D. under SFI External Professor Luís Bettencourt at the University of Chicago—said he valued discussions about integrating complexity science with economics: “It’s not so often that economists talk to people who study ABMs. We are all asking the same questions in very different ways; at the working group we actually made progress towards finding compatibilities between complexity and classical economics.”
Participants also explored how machine-learning tools like large language models (LLMs) could enhance ABM predictions. Valentina Semenova from the International Monetary Fund commented: “We can now encode agents in ABMs with the rich priors of LLMs which allows agent behavior to adapt fluidly to context... We are only scratching the surface of what these cognitively enhanced agents can reveal about the real world.” Examples presented included LLM systems simulating responses to environmental disasters or modeling political polarization.
SFI Vice President for Applied Complexity Will Tracy reflected on changes over time: “In the 1990s and early 2000s ABMs played an important role in helping understand underlying generative mechanisms that animate economic systems but struggled to produce tactical insights on specific interventions... Recent advances in AI data and compute have fueled new ABMs capable of guiding policymakers. This extraordinary working group is doing intellectual work needed to make modern agent-based modeling available to a larger cohort of decision-makers.”
