Complexity Explorer Santa Fe Institute

09 May 2015

W. Brian Arthur on "Complexity Economics"


In this post we interview economist W. Brian Arthur, a founder of the field of “Complexity Economics”.  Brian Arthur received a doctorate in Operations Research from UC Berkley and has served as both resident and external faculty at the Santa Fe Institute. He is the recipient of the Schumpeter Prize in Economics and the Lagrange Prize in Complexity Sciences, and is the author of several books, including The Nature of Technology (Free Press, 2009) and Complexity and the Economy (Oxford, 2014).  

Brian was kind enough to answer our many questions! 

Q: How does complexity economics differ from neoclassical economics?

A: The standard way of looking at the economy is that there are players or agents who can be producers, consumers, banks, investors, etc. We tend to assume these players’ behaviors are consistent with the outcomes they create, meaning there is equilibrium between behavior and outcome. Standard economics assumes that this behavior has no incentive to change the outcome it produces.  Like the surface of a lake, all the water molecules are flat and even and there are no upheavals.  By contrast, complexity economics sees the economy as something like an ecosystem.  There are new behaviors coming in all the time; the economy is always changing and is virtually never at equilibrium, which is a more realistic scenario. In general, the economy is always in upheaval, and people are acting under uncertainty, exploring and adapting , and therefore causing further disequilibrium.  So basically, complexity economics is non-equilibrium economics. 

Q: What led to the creation of the field of complexity economics?

A: There were antecedents.  Ideas of non-linear physics and complexity existed in Europe for several decades, and similarly, ideas have motored around in economics about disequilibrium, imperfection, a dynamic economy, and an economy that’s like an ecology. Crucially, in 1987 a conference was held in Santa Fe, organized by Nobel Prize winner in Economics Kenneth Arrow, and Phillip Anderson, a Nobel in Physics.  They brought physicists and economists to Santa Fe together, myself included.  This meeting led to a research program on looking at the economy as a complex system, which I ran in 1988. So, it originated in Santa Fe, but was not widely noticed at that time.  In 1999, I was asked to do a piece in Science magazine on the economy as a complex system and the editor needed me to give the approach a name, so, I gave it the name complexity economics.

Q: What makes the shift in perspective provided by complexity economics particularly relevant in today’s world? What do you think has led to it being such a hot topic?

A: Partly, it is a delayed reaction.  It took a while for people to even become aware of it.  However, a couple of things have helped greatly. One of the big emphases in complexity economics is that there can be positive feedbacks as well as negative feedbacks.  Negative feedbacks in the economy mean that if there’s more of something it becomes less interesting or more costly.  It turns out that high tech markets work according to positive feedback, and their arrival led to the economy itself looking more biological or ecological. Then the great crash in 2008 very much discredited the standard approach.  The economy was not in equilibrium and suddenly everything collapsed like a house of cards, rattling the profession of economics, which as a result became open to many other ideas.

Q: Do you think that the tools from complex systems and complexity economics could help predict or prevent another financial crisis?

A: I’m not sure it could.  Imagine, as an example, that there are a lot of cars densely packed on the freeway. Assuming that cars will flow along in an even flow, you can solve for what the equilibrium is going to be, knowing that if density increases, cars will slow down.  In complexity economics, you might say, “well, people lose attention in traffic and slow down spontaneously, causing the car behind to slow down, leading to traffic jams.”  That’s much more realistic.  When traffic gets dense you see spontaneous traffic jams and the jams clear and things go on as before.  But complexity applied to cars and traffic can’t predict traffic jams. It would predict that they exist and it would explain them, but it would not tell you when the next traffic jam is going to arrive.  Complexity economics looks at and answers a lot of questions that need a non-equilibrium approach and examines how the economy behaves out of equilibrium.  Why are there bubbles and crashes? How do sudden structural changes and shifts occur in the economy? Why do they occur? You can’t make exact predictions because many of these shifts and sudden appearances of new patterns are based on small probabilistic events, just as you can’t predict earthquakes very well. You can understand the phenomenon even if you can’t predict the next one. 

Q: It was my understanding from reading some of your articles that the complexity economics perspective allows for things like innovation and time and history.

A: The only way that all of those things can be brought into standard economics is to say, “well, we have one equilibrium and then innovation suddenly happens and we have a new equilibrium.”  Complexity economics allows you to look closely at how structural changes happen and how things like creation, innovation and sudden changes and patterns take place.  It’s a widening of the field.  It’s not a replacement of the standard approach. It opens the door to a new set of ideas and problems.

Q: Is that the meso-layer that you discussed in some of your articles?

A: Very broadly speaking, standard economics assumes that there are always some elements or agents---decision makers like banks or investors or firms---and then they produce some outcome that’s in equilibrium with their behavior.  And there are only two layers: the individual agents and the outcome.  So it’s like saying there are individual cars and there’s an overall flow of traffic.  But a traffic jam doesn’t consist of all the cars on the highway and it doesn’t consist of one car, it might be a few hundred cars.   So that’s in the middle: that’s not micro; it’s not macro; it’s meso. Complexity economics opens up a layer of meso-phenomena between individual actors and the overall outcome and patterns start to appear. It’s like saying that individual molecules of air create the atmosphere and the atmosphere could be in perfect equilibrium in the imagination, but it’s not.  There are cyclones and there is rain, etc. There are patterns that form in between the individual molecules and the overall atmosphere of the planet and that’s the meso-layer. Basically, I’m saying that the economy has weather. 

Q: Do you think that that’s becoming more relevant as emerging economies and all of our interactions become more globalized and more complex?

A: The economy is becoming more complicated but that’s not the point.  The point is that when you start to look at the economy from a complexity point of view, you can start to see patterns forming just like you can see traffic jams forming. When one bank collapses, another bank that leans heavily on it may also collapse. You can see propagations of change ripple through the economy in characteristic patterns and all kinds of phenomena that you don’t see in equilibrium. For me what’s exciting about this is it’s like you’ve been in familiar territory all your life in some mansion or other and then you open this door and you say, “there’s stuff in here where nobody’s ever been.” Also, what is striking is that nearly everything that we have seen in complexity economics has been thought of and commented on by political economists.  Looking at the economy over the past century or two they have observed bubbles and crashes, but they worked in words and in specific cases. Now, we are able to study all of this rigorously using mathematically based or probabilistically based models and very often solving these things by computation.  We can make this field just as rigorous as standard equilibrium economics. We can start to understand exactly what is going on and this is really exciting.

Q: What areas of research in Complexity Economics do you personally find to be most exciting?

A: One area of interest to me is that we are starting to find out how technology evolves, what its mechanisms of evolution are through time, how the economy is built up from these technologies, and how structural change in the economy really works. The other thing we’re finding is that we’re getting a much better understanding of economic policy.  When small events can make large differences, the government doesn’t have to intervene suddenly, but it can instead give a small nudge to the system, often nudging it into something practical and more beneficial than if the system went its own way.  We’re also discovering that if the economy is not in equilibrium, it leads you to think about how breakdowns and crises and meltdowns happen and how to prevent those.  So we are starting a kind of forensics of the economy. In equilibrium there’s no need for such forensics because if everything is in equilibrium you don’t get these crashes and meltdowns.  But once you take those seriously and you start to think in non-equilibrium terms, you begin to see that there are incentives for some players to find new ways of exploiting the system and that can be to the detriment of others, and if pushed to an extreme can bring the whole house of cards tumbling down.

Q: Are there many programs in the US that are embracing this? I know you mentioned there was a smattering.

A: It’s kind of on the upswing. If you can get the right dissertation advisor there are small courses here and there. You could still learn a lot on your own if you read all the right papers and if you had the right dissertation advisor you’d be ok.  But I’m very concerned with young people. Last year at the Institute for New Economic Thinking (INET) they had a young scholars program and I talked to them one morning.  What struck me was that they were from all over the world and very interested in complexity economics. I’m particularly concerned that people in graduate school have access to this way of thinking. I do think the field of economics is changing.  It’s not just embracing this complexity approach, but also many new approaches.  I think that the best thing you can do as a graduate student is to absorb all that you can of the standard stuff, but also reach out and into these new and more exciting pathways. 


Comment on this article:

You must be logged in to comment


← Back to news stories