Brian Arthur (left) and Kenneth Arrow (right) circa 1988, at SFI's former location on Canyon Road in Santa Fe, NM

About 30 years ago, economist Kenneth Arrow and physicist Philip Anderson brought together a small group of economists and scientists at the then-new Santa Fe Institute, to discuss the economy as an evolving, complex system. Stanford’s W. Brian Arthur* was one of them.

The group, which developed a year later into a research program led by Arthur, suspected something was amiss with conventional, neoclassical economics. The problem wasn't in the math or the models; rather, they worried that the discipline was simplistic in its worldview. For the last 150 years, economic theory has depended on assumptions — made mostly for mathematical convenience — that consumers and investors think hyper-rationally — they respond to well-defined problems using optimal strategies. 

This approach treats the economy like a well-oiled machine. It's elegant but not realistic, Arthur argues in an essay published recently in Nature Reviews Physics. The economy runs more like an ever-evolving ecology of beliefs, principles, and behaviors, Arthur writes, populated with actors whose decisions — often necessarily based on incomplete information — feedback into the system itself. It's neither neat nor deterministic; it's dynamic and complex. It's "always creating itself, alive and full of messy vitality."

In those early days, SFI was a community of researchers exploring complexity in its many guises: In the dynamics of traffic or the behaviors of individual cells in the immune system. In these systems, individual elements interact and produce patterns, which in turn influence systems and cause the elements to change or adapt.

Such characteristics describe economies, as well, Arthur recognized. Over the years, he and his colleagues developed the core tenets of complexity economics, which eschews the idea that people act rationally, or that the economy has an equilibrium state. By using tools including agent-based, computational modeling, and recursive algorithms, the researchers developed new models that explored economic problems as they might play out in the messy real world.

Complexity doesn’t remedy all the limitations of neoclassical economic theory, Arthur admits, but it can allow for a widening of economic ideas and admit the influence of a greater diversity of forces. And this shift isn't limited to economics: All scientific disciplines are evolving, "embracing openness and process, and asking how structures or phenomena come into being."

*Arthur's current affiliation is Santa Fe Institute External Professor.

Read the essay, "Foundations of complexity economics," in Nature Reviews Physics (January 29, 2021)

Check out a related book, Complexity Economics: Proceedings of the Santa Fe Institute's 2019 Fall Symposium, edited by Arthur and SFI External Professors Eric Binhocker and Allison Stanger (SFI Press, 2020)