Hilal Sarı
Making the economy as predictable as the weather… Oxford University professor, complex systems scientist, and former physicist J. Doyne Farmer makes precisely this claim in his book Making Sense of Chaos: A Better Economy for a Better World. Explaining why standard economic models often fail with data, Farmer builds a brand new “complexity economics” by adapting intuitive models from physics to economics. This approach abandons the assumption of “perfectly rational individuals” and proposes modeling human behavior with bounded rationality and intuitive decision rules. According to Farmer, this is not a limitation—on the contrary, it is the greatest advantage of the future. Because thanks to these models, we will be able to foresee financial collapses in the next 1-2 years.
“We will simulate the economy with supercomputers”
The paradigm Farmer advocates has the potential to transform not only the economy but also policy-making processes and even democracy. He speaks of an era where governments can simulate policies with supercomputers before implementing them, managing complex systems from energy transition to inflation in a holistic manner. “If we can simulate the future better, we can also develop fairer policies,” says Farmer. This approach, which has already caught the attention of some central banks, is a revolution that will shape the future of both sustainable development and economic justice.
Standard economic models assume that people are “rational.”
You argue that complex systems science can be applied to economics. Yet economics is a human system shaped by expectations, beliefs, and political choices. How can a deterministic concept like chaos theory truly adapt to such behavioral complexity? Where do you think human irrationality sets the limits of this new predictive framework?
Understanding and modelling human behaviour is indeed what makes economics hard. One of the remarkable results that I discuss in the book is that the interaction of decision-making and economics can result in chaos. To show this, we model human decision-making as boundedly rational, based on simple rules of thumb that in some circumstances are a good approximation for how real people make decisions. The ability to build models that do not depend on the concept of perfect rationality, which is so widely used in mainstream economics, is one of the strengths of complexity economics. So this is not a limit — it is a strength.
Predicting the unpredictable, like forecasting the weather
One of the most striking ideas in your book is the analogy of predicting the economy as accurately as the weather. But even meteorology has its limits. Should the ultimate goal of complexity economics be perfect prediction—or rather, a more probabilistic, risk-management-driven understanding of economic systems?
I do not claim that we can predict the economy as accurately as the weather! What I do claim is that we can do a better job of predicting the economy than we do now by predicting the economy similarly to how we predict the weather, i.e. by feeding comprehensive global data into simulations of the economy. The goal is not perfect prediction — just better prediction.
“Even democracy works better with these models.”
You envision a future where governments can simulate policies on supercomputers before implementing them. How might that reshape the policy-making process? Do you see a tension between democratic deliberation and data-driven simulation?
I believe that if we have a better understanding of what policies will do and where they will lead us, we will implement better policies. Right now governments don’t pay much attention to models because their predictions are not very good. If the predictions were better, that might change. There is no tension with democratic deliberation — quite the opposite! If we had data-driven simulations showing us what the future is likely to be conditional on our policies, we might have an easier time coming to an agreement, and democracy would work better.
“Our models will predict crises in a few years”
After the 2008 global crisis, many argued that economies cannot be modeled like physical systems. Do you believe econophysics has now matured enough to foresee another financial collapse? Or are we still living through chaos within chaos?
I would not say that econophysics has matured enough to foresee another financial collapse — yet. However, we are building tools and models that I think may be able to do this (at least better than we can do it now) within a few years. Until then, we are stuck living through chaos within chaos.
“We will navigate the energy transition better.”
You suggest that complex-systems models could help manage the energy transition and the climate crisis. Could this approach serve as the foundation for a new Bretton Woods in sustainable development?
A new Bretton Woods for sustainable development is a wonderful idea. Unfortunately, I think we will have to suffer through even worse environmental degradation — bad enough to threaten the economy — before that will happen. However, in the meantime, we are building models that can help us better navigate the energy transition, to lead us to more prosperity in a more sustainable manner.
“I’m not pessimistic about artificial intelligence, but I am concerned”
Complex systems, big data, and AI now converge in the same equation. In your view, does AI have the potential to make economics more human-centric—or are we moving toward a more mechanised order?
We are at a point in history where many things can happen, making it particularly difficult to predict the future. I do think that AI used with agent-based models (simulations of the economy) can help make things better for us humans. I am worried, though, about where AI may take us in the long run — even if I am also optimistic that the result will not be the doomsday that many are currently worrying about.
This is revolutionary: Central banks have started using it
You describe this as a revolution. What mental or institutional barriers must be broken—across academia, finance, and policymaking—for that revolution to take hold? And do you believe it will lead to a fairer economic system?
This is a revolution because it changes the fundamental assumptions that have underpinned economics for more than a century. Complexity economics will become more and more important in the commercial world, and then start to be used by central banks and treasury departments (this is already beginning to happen). Then, in academia, it will either invade economics departments or entirely new complexity economics departments will be created. Because complexity economics is much better at modelling a heterogeneous world, with rich and poor, young and old, etc., it will give us much better advice on questions of economic social justice. If this advice is followed, it could lead us to a fairer economic system.
“Turkey can use these models to solve the inflation problem.”
Looking at a dynamic and often unpredictable economy like Turkey’s — which key principles or approaches from your book could policymakers here adopt to build more holistic stability and long-term prosperity? In which areas do you think a complexity-based framework could make the most tangible difference for Turkey’s economic policymaking?
Turkey has had serious problems with inflation — I believe that we can provide a better understanding of what causes inflation and how it can be controlled. It could also help guide Turkey’s industrial policy to help Turkey participate more effectively in the global economy.