The usual background and distinctions between complexity and neoclassical economics are presented. Neoclassical economics deals with perfectly rational representative agents -this creates states of equilibrium. On the other hand, complexity economics relaxes these assumptions to deal with responsive agents in an uncertain dynamic environment -this creates states of disequilibrium.
More specifically, it involves a ‘repeating loop’ of individual actors adjusting their behavior to aggregate outcomes which then leads to new aggregate outcomes. “It is this recursive loop that makes the economy a complex system.”
The article presents the personal experience of W.B Arthur in how the field of complexity economics was initially established. He states that this article is not a formal description, nor is it a survey of other papers now in the field.
After this, the article becomes much more interesting. The third section shows how the complexity approach offers explanations where neoclassical theory can not. It is able to account for actual market phenomena such as the emergence of market psychology, price bubbles and crashes, the heavy use of technical trading, and random periods of high and low volatility.
The fourth section explains the overlap between complexity economics and agent-based computations used in economics. Sometimes complexity economics uses mathematics (such as nonlinear processes), but the sheer complication of keeping track of the decision processes of multiple agents requires the use of computers. Computers allowed economic theorists to venture beyond the standard neoclassical assumptions. Turning these new possibilities into a theoretical framework produced complexity economics, turning them into a solution method produced agent-based computational economics. One could see agent-based computational economics as a key method in complexity economics or one could regard complexity economics as a conceptual foundation behind agent-based economic modeling.
The fifth section shows how, in a complex system, actions taken by individual elements are channeled via a network of connections among them. This part discusses how the network structure or topology makes a difference, how markets self-organize within these networks, how risk is transmitted, how events propagate, and how they influence power structures.
The sixth section discusses policy suggestions of complexity economics and the seventh discuss the frontiers of complexity economics including studies of the formation of the economy, studies of distribution, construction of more realistic models, industry application, and the autonomous economy.
This article is written by William B. Arthur, one of the economists who developed complexity theory. For students who find section two of this paper a bit dull, I suggest reading his other article, Complexity Economics: A different framework for economic thought, which gives a similar introduction to complexity economics but in a more vibrant way.
Nevertheless, the rest of the article gives exceptional insight into complexity economics, its overlaps with computation-based economics (which is often confusing), and its current frontiers.