Econophysics & Economics

“Blue Aurora,” who frequently leaves comments both here and at Facts and Other Stubborn Things, has recently had published a guest post on econophysics over at the latter.  He asked me to offer my opinion on the viability of econophysics, and although I feel overwhelmingly underqualified to give it, I will at least provide an elementary Austrian criticism of what econophysics seeks to do.  Unfortunately, this comment will not extend into evaluating the econophysicists’ criticism of the mainstream.

From the information provided in the post, I can make two criticism,

  1. The general criticism: economics is not a predictive science;
  2. The more specific criticism: human beings are not comparable to physical particles.

I will explain these criticisms in reverse order, since the first one is largely dependent on the second one.

What makes human action unique?  The human mind is what makes human action unique.  Ludwig Lachmann, in The Market as an Economic Process, emphasizes the idea that all action must be precipitated by plans, which are mental constructs that organize means and ends into viable physical actions.  This gives human action a broad scope of possibilities in terms of choice; possibilities that are constrained by the creativity of the mind (constrained, too, by other factors, including income and the actions of others, but one could conceivably think of an essentially infinite number of actions, even with these constraints, if the human mind was infinitely creative, so to speak).

The econophysicists’ claim that their ability to predict comes from the use of mathematical tools which break down the subject studied into the smallest divisible unit (i.e. [I am not scientist, so excuse any errors] subatomic particles, and, in economics, the individual).  They can then empirically test patterns, and from these predictable patterns they can lay the explanatory groundwork behind scientific events, and even extend this logic to predict events which cannot yet be empirically verified.  Again, I am no physical scientist, but this is done in astrophysics, and I assume elsewhere.  For fields where individual particles have a limited range of “action” (or outcomes) this approach is great.  I am not convinced the same is true for economics.

The individual human does not act predictably.  She has a choice of means and ends, and chooses between these, and acts upon plans to utilize them.  Her purpose is to bring about “satisfaction,” or reach a more preferable state of affairs.  These means and ends are always-changing and fluctuating, meaning that the range of choices in action are always changing; no less, the human mind is always devising new plans, which also changes the range of choice.  My point: individual human is incomparable to the subatomic particle. Humans do not act, collectively, in predictable patterns — if they did, there would be no such thing as changes in the state of economy (ostensibly, increases in wealth).  Nobody would think of products that did not exist beforehand, and nobody would think of new actions to remain competitive on the market.  These are all unpredictable events, precisely because you cannot predict the human mind (especially on a collective scale).

I do believe you can make predictions that are very, very general.  An example would be Austrian business cycle theory.  This theory suggests that if the rate of interest in the loanable funds market falls, and this causes price distortions between goods of different orders, mass malinvestment will ensue.  It is general in the sense that it does not predict what exact type of malinvestment will be created, and it does not predict when the bust will occur.  This also means that the longer the period of time you are predicting over, the less accurate your prediction will be (because there is more room for changes in action).  No less, these types of actions also depend, to some degree, on the homogeneity of action.  This homogeneity is caused by central direction, and in the case of the business cycle this central direction comes from a key distortion (the rate of interest on loanable funds).  Similarly, regulations can serve to homogenize action, since it limits the scope of action.  These are not the type of predictions econophysicists want to make: like the mainstream, they want to make increasingly specific predictions of the future.

Because of these constraints on the predictive power of economics, I believe that economics is principally an explanatory science.  In other words, it serves to theoretically elucidate on certain relationships that make up our world.  These relationships deal with the individual and action, and the consequences of action.  Lachmann does not approve of this approach either, because in a sense this still about making predictions (especially because, oftentimes, some of these relationships cannot be empirically verified by marginalizing other, contributory variables).  The difference is that these predictions are really over very short periods of time: the amount of time between one action and the next.  These explanations can then be used to explain historical phenomena.

We can see, though, how Austrians part ways with the mainstream and others who believe economics’ main task is to predict.  Austrians do not seek to pass judgment on how humans act, or in what ways they will act.  Austrians simply look at explaining what will occur if a certain action is taken.  Those who want to predict future events are undertaking a much more ambitious task: predicting how humans action.  I am sure there is some level of consistency, somewhere; consistency in some parts of action, however, will be marred by inconsistency elsewhere.  Time will tell how accurate econophysicists’ models turn out to be, but I feel that they will find the need to constantly update their models to consider underlying changes in preferences, expectations, and creativity.

4 thoughts on “Econophysics & Economics

  1. Blue Aurora

    Thanks for responding to the guest post, even though the second part isn’t up yet! I’m waiting for the second post to be up…

    Reply
  2. Dave Marsay

    If Keynes et al are correct, and the course of an economy cannot always be predicted in a narrow sense, and if econophysics is capable of modelling economies, then econophysics should be able to produce models of toy economies where the ’emergent properties’ are unpredictable. In so far as econophysics is founded on statistical mechanics, which relies on conventional probability theory, this may seem paradoxical, but so be it.

    Reply
    1. Jonathan Finegold Catalán Post author

      The idea behind econophysics is to use models based on patterns of behavior to predict events, such as financial crises. These would fall in what is unpredictable. I’m not denying we can broadly conceive of what will lead to a financial crisis, but all these “predictions” are so broad that they are almost useless (for that purpose).

      Reply
      1. Dave Marsay

        For me, the critical issue is the extent to which the models are ‘based on’ patterns. Conventionally, one would use tools such as those of econophysics to suggest models / theories, which would then be tested more broadly before applying. Some econophysicists seem to suggest that it is enough to test the theories against the same type of data, but I doubt it. For example, I cannot see how one could have spotted the up-coming 2007/8 financial crisis just by looking for patterns in the financial data. On the other hand, Keynes’ General Theory suggested failure modes, such as that the end of the Beijing Olympics would cause a hiatus that the US financial system couldn’t cope with. Econophysics type models illustrated these potential failure modes, and suggested issues to be explored. These came down to ‘who was minding the shop’? I do not see how econophysics on its own could answer this question. Indeed, some econophysicists seem to suppose that financial markets are ‘nothing but’ agents in a market reacting to news, and do not need minding. This is not how I see it.

        Reply

Leave a Reply