Daily Archives: August 17, 2016


What is Econophysics?

How do we define a subject? Traditionally (as written in Wikipedia for instance) a subject matter is defined according to its objects. Traditionally, physics is defined according to its objects which are “matter and energy” (for example). That also applies to other subject as well. The objects of biology is living beings, and the object of psychology is psyche, and so on and so forth.

However, the objects of Econophysics are economic systems which are not subsets of matter and energy, yet we still call it “physics”. The same is true for biological physics or mathematical sociology, for examples. These are examples where the subjects are not defined according to their object, but rather according to their methods.

Now let me stop you for a minute and let us think about the meaning of physics. Is it true that “physics” can be totally defined by its method, and not by its material objects? Let’s look at a specific examples. Suppose we are going to model an economic system using a lattice-gas model. The model itself works on physical objects (“particles”) which are doing their things on “a lattice” (which is also a physical object). We call it “gas” because the particles are able to move and interact as a gas. We call it “lattice” because the setting is a lattice where each cell an only be occupied by a single particle. Now, this system could represent an economic system because the collective behavior of an economic system can be represented by mathematical models which can be generated by the collective behavior of a lattice-gas model.

Here you may disagree with the necessity of using physics to model an economic systems. If the goal is to obtain a set of mathematical equations that are able to model the behavior of an economic systems, then we don’t need physics for that. This opinion arises, I think, when all that we are looking for is a set of equations that may represent the collective behavior of a system. If that is all there is to it, then ALL problems in economics are basically a combination between economics behavioral science and mathematics. Economics behavior science build a theory that explain the economics human behavior in qualitative manner, while mathematics build a set of mathematical equations that will connect that to empirical data hence completing the whole thing by making it quantitative. Don’t forget also the statistical tools to test whether the mathematical models can explain the data in a significant way. Again, no physics is needed.

In my opinion, physics is not really needed when what you are trying to get is a set of empirical equations. As long as you have the empirical data and a qualitative theory in your hand, all you need is a good mathematical procedure to complete the circle.

Physics is needed when you are having troubles building qualitative theories that can be quantified immediately in response to all possible empirical facts that we are having at that moment. In the past, this need has not been arisen since – apparently – the economics systems are evidently quite simple. The economics behavior during those simple days are quite obvious and can easily be explained by some simple qualitative concepts, which in turn can easily be quantified by some simple mathematical operations to be correlated to the empirical data. Because, obviously, the empirical data itself is quite simple. As the systems themselves are getting more and more complex, we need a new method to explain the reality, qualitatively as well as quantitatively. Here, physics is a method by which we build a qualitative theory which can be quantified and used to explain the empirical data.

Some people say that physics can be used for any economical situation, simple or complex. That may be true. But is it really necessary? If the standard economics is able to do the job, why would we need physics for that? Physics, I think, is needed when all other attempts are failed.

The behavior of complex systems is one example where physics can be quite useful. The simulations using a lattice-gas model directly explained -quantitatively- how the interactions between economic agent generates many different collective behaviors. If the only thing that the simulation is capable of doing is to produce some collective behavior that can be quantified, then it won’t be quite useful. Its ability to show HOW those collective behaviors emerge which can be verified quantitatively is what makes it useful.