Thursday, November 03, 2011

Flaws of Economic Models: Differentiating Social Sciences from Natural Sciences

David Freedman of the Scientific American asks “Why Economic Models Are Always Wrong?”

He gives the answer,

The problem, of course, is that while these different versions of the model might all match the historical data, they would in general generate different predictions going forward--and sure enough, his calibrated model produced terrible predictions compared to the "reality" originally generated by the perfect model. Calibration--a standard procedure used by all modelers in all fields, including finance--had rendered a perfect model seriously flawed. Though taken aback, he continued his study, and found that having even tiny flaws in the model or the historical data made the situation far worse. "As far as I can tell, you'd have exactly the same situation with any model that has to be calibrated," says Carter.

That financial models are plagued by calibration problems is no surprise to Wilmott--he notes that it has become routine for modelers in finance to simply keep recalibrating their models over and over again as the models continue to turn out bad predictions. "When you have to keep recalibrating a model, something is wrong with it," he says. "If you had to readjust the constant in Newton's law of gravity every time you got out of bed in the morning in order for it to agree with your scale, it wouldn't be much of a law But in finance they just keep on recalibrating and pretending that the models work."

We can’t simplify, through mathematical models, what truly is a highly complex environment. Repeated “recalibrating their models” or “calibration problems” only exposes on these structural analytical errors.

The ultimate reason why economic models are always wrong is that investigations have been patterned after natural sciences. Yet analyzing natural sciences isn’t the same as social sciences. That’s what modelers and their disciples cannot seem to grasp.

The great Ludwig von Mises draws a clear distinction between the two sciences, (bold emphasis mine, italics original)

Since the elements of social cognition are abstract and not reducible to concrete images one would like to have metaphors. First there were biological metaphors, now mostly mechanistic ones. These are based in positivist view of social science that holds that social science should be built up by experimental method as ideally applied in Newtonian physics. Economics becomes experimental, mathematical and about measurement. This is all wrong:

1. Social sciences cannot be based on experience like the natural sciences. Social experience is of a complexity and cannot be experimented with

2. Therefore the social sciences can never use experience to verify their statements. Every fact and experience is open to multiple interpretations (but see Kuhn. KS)

3. The impossibility of experimenting implies the impossibility of measurement. In human behavior there are no invariable relations like there are between physical properties, which means that it is pointless to mathematize them in order to make predictions. Statistics merely studies history.

4. Mathematics does not deal with actual operations of human actions but with a fictitious concept, static equilibrium (tomorrow is like today, no uncertainty), that economists build up for instrumental purposes. But not only is this unrealistic, it is also inconsistent for lack of uncertainty and change implies lack of actions. The only purpose mathematics can have in economics is the study of the nature of relations between costs and prices and thereby of profits.

5. Mathematics cannot tell us how the market arrives at a static equilibrium.

6. Mathematicians are prone to consider the price either as measurement of value or as equivalent to the commodity. But prices are neither; they are simply the amount of money exchanged for a commodity and there is reversed valuation.

Economics deals with human action, not with objects (as physics does) such as commodities, economic quantities or prices. Therefore economists do not consider their subject matter from without, but from within, through our own understanding of what it is to be human and to act. What makes natural science possible is the power to experiment, what makes social science possible is the power to grasp the meaning of human action….

Social sciences have a distinct method, praxeology and verstehen, due to the special character of their objects, and owe their progress through it and do not have to and cannot use the method of the natural sciences.

Praxeological concepts refer exactly and with certainty to the reality of human action because both the science of human action and human action itself have their toot in human reason. The quantitative approach would not render them more exact.

Nobody denies that economics is not perfect yet, but:

1. the present unsatisfactory state of social and political affairs is not due to deficiencies in economic theory, but in policy. People just don’t use economic theory enough.

2. even if economics needs to be drastically reformed someday it cannot take the direction proposed by those who use the model of the natural sciences. This idea has been thoroughly refuted forever.

Again many people seem to find comfort in models, for many possible reasons such as social signaling, conversation, career, politics and others.

But in terms of the predictive value, as the Scientific American article’s inquiry as indicated by the title, economic models have always been wrong.

No comments: