Complexity Economics – A New Approach


Economics has been dominated by the neoclassical school of thought. Economists early were keen on proving that economics is also a science and were heavily influenced by the scientists and scientific enterprise of their time. They start moving towards ‘equilibrium’ as a natural state of thing or the end objective.

Complexity economics is a new way to approach economics and understand the dynamics in the economy. Complexity is the study of the consequence of interactions. It studies patter, structures, phenomenon that emerge from interactions among elements. It shows us how change plays out. Complexity provides a new way to understand economics in a non-equilibrium environment. Equilibrium is not necessarily always present.

There are many instances where there is a non-equilibrium environment or an environment where there are multiple equilibria. Economic agents constantly change their actions and strategies in response to the outcome they mutually create. This further changes the outcome, which requires them to adjust afresh. Agents live in a world where they constantly test their beliefs to see which ones work and which ones don’t.

The ones that don’t work are discarded. This implies that strategies and actions constantly evolve over time, making time also an important element. This would mean that structures would also constant evolve, making the entire system adaptive. The level of analysis here would be ‘meso’ rather than micro or macro. The resultant would be a world that is organic, evolutionary and historically contingent.

Understanding the economy 
What is an economy under this view? An economy is a “vast and complicated set of arrangements and actions wherein agents – consumers, firms, banks, investors, government agencies – buy and sell, speculate, trade, oversee, bring products into being, offer services, invest in companies, strategize, explore, forecast, compete, learn, innovate and adapt.” In short, it is a massively parallel system of concurrent behavior.

Compare this with the standard definition of an economy which only looks at production, consumption and distribution. The standard only looks at what behaviors would be upheld by and be consistent with the aggregate patter these caused i.e. what patterns would call for no changes in micro-behavior and would therefore be in equilibrium. When we compare the two approaches, we can conclude that complexity has a much more realistic and comprehensive view of the activities in the economy.

Equilibrium is an obsession. But the moment we ask how one might react to something, we are operating in an environment of disequilibrium. The view is that disequilibrium is caused by ‘external’ or ‘exogenous’ shocks. This isn’t the only cause of disequilibrium. There are also internal causes of disequilibrium – uncertainty and technological evolution.

The moment we bring in an element of time, uncertainty follows. When we look at choice, it is fundamentally something that will happen in the future – bringing in the element of time and uncertainty with it. No matter how well informed a person is, s/he cannot know all possible outcomes, their values and their probabilities. The problem of uncertainty is commonly confused with risk.

When we look at uncertainty in a non-equilibrium environment, where everyone is adapting and responding to other, uncertainty engenders further uncertainty. To the degree to outcomes are unknowable, the decision problem itself cannot be well defined. Therefore, there cannot be a logical solution by way of deductive rationality cannot exist. The economy is permanently in the state of disruptive motion as agents explore, learn and adapt.

This is true also of technology, which is often assumed to be constant. Schumpeter recognised this and referred to it as a “source of energy within the economic system which would of itself disrupt any equilibrium that might be attained”. Novel technologies call forth more novel technologies. Computers brought with it further technologies such as data storage, algorithms, chips, etc. Novel technologies make other novel technologies possible. So technological disruption isn’t a one time thing, but happens in ongoing waves of disruption. Technological change endogenously breeds further technological changes. This is a continuous process which also ensures the economy is in a perpetual state of disruption.

Non-equilibrium Filter
So, what happens when we take away the equilibrium filter? We operate in a state where we can see general patterns of formation. The accuracy of these patterns is much higher than the accuracy of particular points. This enables to understand the dynamics and evolution of a system. There are three things that can be notice. The phenomenon is spontaneous, the phenomenon is temporal and that the phenomenon happens in the meso-level. To understand these dynamics better, let us take the example of a stock market. There are ‘self-reinforcing asset price changes’, which we commonly refer to as bubbles and crashes.

These are spontaneous. For a more detailed discussion on this aspect, you could refer the article on Soros’ theory of fallibility, reflexive and human uncertainty principle. The temporal phenomenon is captured by ‘clustered volatility’ which refers to random periods of high and low activity in the market. The third phenomenon is captured by ‘sudden percolation’. Depending on the density of the network, a change will propagate and continue to propagate itself or it will die out.

There is also another phenomenon called ‘phase transition’ where a phenomenon doesn’t appear until there is some underlying parameter of the model that depicts the intensity of the adjustment or the degree of connection passes some point and reaches some critical level. This leads to a change in the overall behaviour. This leads us to three broad phases – simple behaviour, complex behaviour and chaotic behaviour. The means we shift from equilibrium to complexity to multiple equilibria.

When we connect this to the understanding of complexity which a science that understands the outcomes of interactions, we can arrive at certain general properties that will help us understand systems better. Propagation of events causing events can be viewed from mathematical aspects such as power laws, heavy tailed probability distributions, long correlations, etc. These phenomena appear and disappear in distinct historical time or space. They are localised and appear in one part of the network and are diffused from there. They operate at all scales – from individual to the entire economy, but they usually take place between the micro and macro space, the meso space. If we insist on an equilibrium approach, all the dynamics that actually drive events remain hidden and unrecognised.

-Contributed by Bhargav Dhakappa

Picture Credits:

Most Popular

To Top
Please check the Pop-up.