« Return to Contents of this issue
Complexity Economics: Proceedings of the Santa Fe Institute's 2019 Fall Symposium (Dialogues of the Applied Complexity Network)Arthur, W. Brian, Beinhocker, Eric D. & Stanger Allison (Eds.)
The Santa Fe Institute Press: Santa Fe, New Mexico, United States, 2020
ISBN 1947864351 (pb)
Order this book
University of Turin
The volume, edited by W. Brian Arthur, Eric D. Beinhocker, and Allison Stanger, includes panel and talk transcripts from SFI’s 2019 Applied Complexity Network Symposium, with newly written introductions and reflections.
The Table of contents is at https://www.sfipress.org/books/complexity-economics. The US price is $ 3.26 (Kindle) or $ 9.99 (Amazon) for a wonderfully printed, high-quality book. Now, the content.
It was another era, in September 1987, when the then-nascent Santa Fe Institute (SFI) held a conference to look at the economy as a complex evolving system. The organizers were the physicist Philip Anderson and David Pines, and the economist Kenneth Arrow.
As the introduction of Arthur, Beinhocker, and Stangers remembers, the idea of complexity in economics has a very ancient antecedent in economics: Adam Smith (p.10) «had a deep, intuitive understanding of emergence and was arguably the first complexity economist.» It is important to understand that Smith and the other early economists were aware that aggregate patterns emerge from individual actions and interactions and that the results influence individual behavior. Again p.10:
Smith’s famous metaphor of the “invisible hand” of markets is popularly misinterpreted as a message that “greed is good” (something Smith did not believe), but in reality was a statement about emergence–how individual actions “without intending it, without knowing it” lead to collective outcomes which feedback to influence further actions.
To understand the complexity revolution in economics, we have to accept the quoted sentence as critical. Being aware of this concept, we naturally refuse the logical consequences of the so-called Lucas critique. Shortly, the idea that models need "microfoundations" to incorporate not only the supposed capability of the economic agents to predict the effects of their actions but they have to include the agents' ability to take into account the changes in politics in their evaluations; only from there, models can aggregate the individual decisions to calculate the macroeconomic effects of the policy changes. As above, people act "without intending it, without knowing it." Suppose oligopolistic or monopolistic agents act within the system: for sure, they intend and know. In that case, another perspective is running (see as an example Mazzoli et al. 2019), where the market's endogenous structure is profoundly influencing the agents' behavior.)
The book is very rich in arguments and perspectives, with the philosophical analysis of Eric D. Beinhocker on “Can complexity economics save the world?” and a computational section on agent-based models (ABMs). Essential contributions on calculation are from Robert Axtell, Joshua Epstein, Blake LeBaron, Melanie Mitchell, and many others. Also, the pandemic situation is considered, with post-meeting insertions.
Globally, this is a significant step ahead in diffusing complexity economics, which is still not mainstream, but more and more considered also thanks to the diffusion of tools as agent-based models or models coming from physics as behavioral swarms and active particle theory.
Following complexity studies, we use ABMs for their flexibility in describing any detailed textures of social structures in a fine-grained way. The risk is the paradox of the map that Borges pictures in On Rigor in Science, where “the Colleges of Cartographers set up a Map of the Empire which had the size of the Empire itself and coincided with it point by point.” Sure, we can avoid that risk, and we do, but we cannot forget the lack of an ex-ante mathematical foundation. The behavioral swarm models’ opposite situation is grounded on theory but ignores essential details and information about actual agents. We have to work toward a grand unification of agent models, where also the construction via a formal computer language is verifiable as consistent with a mathematical framework.
So, a vast research task for Santa Fe's internal and external community and the mathematics of life and society. This book is a mainstay to proceed in that direction.