Time to complicate things

If we stop trying to simplify our economic models, we can improve policy.

According to the increasingly influential school of complexity economics, decisions that at the "micro" level might seem rational, when they become manifest at the "macro" level produce outcomes that are detrimental to all. Several "rights" often combine to produce a "wrong". This can help to explain problems like why there is such an unequal distribution of wealth in many developed economies and why some regions remain depressed for long periods of time.

Neo-classical economics finds it difficult to account for such emergent problems because it is based on a framework of simple, bilateral exchanges between individuals (people and firms). It is forced to regard economy-wide problems as the result of some external disruption to the normal running of free bilateral exchange. As a result, it has failed to develop an adequate theory that makes a connection between individual decisions and developments in the aggregate economy.  

The new fields of complexity theory and network theory help us to understand the economy as a dynamic network, rather than as the static model of bilateral exchange, which underlies orthodox economics. With this new understanding, we can start to develop policy options that might respond more effectively to problems like inequality of wealth.  

Although the school of complexity economics and the idea of emergent phenomena are relatively new, the recognition of system-wide economic problems is clearly not. But, for a long time, most economic thinking has held that many of our fundamental challenges, such as inequality and climate issues, are the result of market failures and can only be solved by using the power of the state to correct any such failures.  

A complexity perspective suggests the state’s top-down bureaucracy cannot, as it currently operates, offer effective and sustainable solutions to emergent economic problems. The state's policy-making incorporates the idea that society acts like a machine that responds automatically to a stimulus (such as a tax cut) in the same way each time, in large part because mainstream economics has taken a very narrow view of human nature and interaction. This has been explored most recently by Paul Ormerod in his book, Positive Linking.

Complexity theory is based on the core observation that social systems are dynamic, evolving networks in which individual and collective behaviour can shift and change rapidly and unexpectedly. The fluidity of this system means attempts to control an economy by gathering data, making forecasts and developing policy will always be subject to a high risk of failure. Mechanistic approaches to policy can be extremely problematic, for at least four broad reasons:  

  1. Idiosyncrasies matter and it is a near impossibility for a centrally determined policy to remain sensitive to local circumstances.
  2. Remotely set targets can be inaccurate proxies for real aims.
  3. Network effects can drown out the very incentives that form the core of most policy responses.
  4. Incentives are often set as if people were selfish maximisers of their own utility - but this is very often untrue.  

The neoclassical approach is comforting in the sense it implies that following simple, easy to devise, mechanical policy rules can solve some problems. But it is misleading because the economy does not work in the way it suggests, which often leads to inappropriate policy ideas. With the complexity approach things are, roughly speaking, the other way around. It suggests policy responses to certain problems will be hard to draw up, and the right answer might be found only after experimentation, simulation, and pilot studies. But the policy formed as a result is more likely to be suited to the policy challenge.  

The new fields of complexity and network theory advocate building up an understanding of the real world from the ground up. In doing so, they paint a picture of the real world that is much more recognisable than the abstractions of neoclassical economics. As such, they have the potential to offer new approaches to seemingly intractable policy problems, and, because these approaches are inherently apolitical, they ought to be of interest to all political parties.

This is an edited extract of a chapter from IPPR’s forthcoming book, Complex New World: translating new economic thinking into public policy. For more see here.

A bifurcation diagram showing a common representation of chaos theory. Photograph: Wikimedia Commons

Adam Lent is the Director of Programmes at the RSA. Greg Fisher is the Managing Director of Synthesis.

Show Hide image

The New Statesman Cover | Wanted: An opposition

A first look at this week's magazine.

March 31 - April 6 
Wanted: An opposition