Sunday, June 19, 2016

Mark Thoma — If the Modifications Needed to Accommodate New Observations Become Too Baroque ...


This is important if only because Stephen Hawking gets it wrong.
Hawking: So which is real, the Ptolemaic or Copernican system? Although it is not uncommon for people to say that Copernicus proved Ptolemy wrong, that is not true..., our observations of the heavens can be explained by assuming either the earth or the sun to be at rest. Despite its role in philosophical debates over the nature of our universe, the real advantage of the Copernican system is simply that the equations of motion are much simpler in the frame of reference in which the sun is at rest.
Sorry. Not so. The logic does not hold.

The models are equal as models but not as causal models of reality, as Newton's theory of gravitation proves beyond doubt. Scientific explanation is assumed to provide causal accounts, not just models that fit. Many models can fit a set of conditions. The challenge for science is distinguishing which are correct and this is accomplished by causal accounts.

The planets orbit the sun rather than the earth and while Copernicus did not know why, we do now. So, strictly speaking, Hawking is correct within the time frame. The causal account was only provided later. But it completely resolves the issue causally without appeal to formal elegance.

Newton's account of planetary motion shows not only how (descriptively), but also why (causally) in terms of the gravitational force — Newton's law of universal gravitation and his laws of planetary motion. Case closed. Ptolemy is a historical footnote. Copernicus, Bruno, Galieo, etc. are vindicated.

The Ptolemaic model is not just more inelegant and cumbersome, it is actually wrong about reality in that there is nothing causing the sun to orbit the earth and there is a well-explained cause of the planets orbiting the sun and the moon orbiting the earth.

Even though Ptolemaic models can be adjusted to fit, it's not just a matter of economy of explanation. There is no comparison between the force of the heliocentric causal account of modern science based on math and empirics, and the ancient account based on the assumed importance of the earth establishing it as the center of the universe and jiggling with the data to make it seem so.

This is no small matter. Copernicus was careful in sufficing it, Giordano Bruno was't and got burned at the stake over for asserting heliocentrism (among other similar matters), while Galileo Galilei had to recant to save himself. This was not about models. It was about causality. The authorities of the time saw heliocentrism as religious heresy that undermined the centrality of the Incarnation, for example.

Similarly, economic models are either adequate causal models as representations of reality or not. Potentially an unlimited number of models can be devised to fit and adjusted as conditions require. As supposedly scientific theories, the degree of elegance they display may be suggestive, but it is inconclusive without causal explanation that is supported by observation (data).

Moreover, simple models may appear elegant but they are actually simplistic when matched against reality. Economy of expression cannot be a deciding criterion in science, although it may be in math or philosophy.

Samuelson's insistence on ergodicity in order to make economics a science more like natural science rather than social science was a necessary condition and not a sufficient one. That must be borne out by facts supporting causal effects through hypothesis generation and testing.

Ergodicity is needed for generality in change. Ergodicity ensures that the future will resemble the past in that there is no difference between an ensemble and a time series. Timeless laws become possible, whereas they are not in a non-ergodic environment in which the future is uncertain in the sense of the well-known disclaimer, "past performance is no guarantee of future performance."

General case theories don't allow for exceptions at the margin. If there are exceptions, then the explanations are of special cases.

For example, Say's "law" does not describe a general case but only a special case of full employment. Assuming full employment "in the long run" is meaningless scientifically without specifying parameters. As Keynes joked, "In the long run we are all dead."

"In the long run" gives no useful information even though it as the gravity of a philosophical principle. It actually conceals a petitio principii or vicious circle.

Nor is there a law of supply and demand that governs markets as a market force, or "invisible hand" that left to itself spontaneously generates natural order in which all available resources are employed optimally (efficiently). Erroneously attributed to Adam Smith, the invisible hand as a metaphor for market forces is largely a Paul Samuelson invention promulgated widely in his 1948 textbook and picked up by the echo chamber.

Nor do rational agent optimize economically; they take all of life into consideration motivationally and not just economic interests, nor do all necessarily prioritize economic interests. That is to say, preferences are both economic and non-economic, and some non-economic preferences (personal, social, political) are prioritized over economic preferences rationally, that is, taking all aspects of purpose into account in "optimizing utility" as well-being. The economic "basket" is a limited case in this. economists need to talk to motivational psychologists about this, as well as sociologists and other life and social scientists.

Assuming equilibrium involves assuming the cause, when the task is to explain the causality. The question is whether general equilibrium prevails as a general or special case and if so, what is the cause. There seems to be no data supporting general equilibrium either as an assumption or as an outcome of market action in a monetary production economy in open economies with government. assuming a closed economy without government is not very instructive. Aiming at this as a goal is insane.

Given chronic unemployment, markets are out of equilibrium more than in it, unless unemployment is defined to fit the assumption of general equilibrium, which is the logical fallacy of begging the question, which is a type of circular reasoning. The arguments based on natural rates (interest and employment) are philosophical rather than scientific. In science, "natural" means observed regularity, whereas neither the natural rate of interest or the natural rate of unemployment are observables. They are theoretical terms. Calling them "natural" is bogus science.

No matter how elegant a model based on such assumptions may be, formal consistency doesn't prove anything about the fitness of the model as a causal representation of reality. The model first must fit, and secondly it has to account for why it fits based on causal transmission in a way that allows for generation of testable hypothesis. Such hypotheses can only be extended generally in ergodic systems, and there are no good reasons to believe that economics is any more ergodic than other social sciences because markets and prices and "economic laws" governing them.

The whole contraption is a Rube Goldberg machine logically because it assumes economic priorities as fundamental and exclusive. Talk about inelegance. The only thing elegant about it is the formal consistency, which alone is meaningless. An elaborate tautology. Impressive tour de force, but empty of content.

We see this sort of faux logic presently in the argument over Brexit, where the (imagined) economic consequence are being used to stoke fear, so as to confuse the actual issues, which are the supremacy of national sovereignty and democracy in a liberal order in opposition to the tyranny of technocracy under unelected and unaccountable bureaucrats. The real issues are not only or even chiefly economic.

The issues are social and political — nationalism and democracy versus internationalism and bureaucracy. Plus, the economics is dodgy. See Remain’s models are built on poor foundations.

Economist’s View
If the Modifications Needed to Accommodate New Observations Become Too Baroque ...
Mark Thoma | Professor of Economics, University of Oregon

5 comments:

Bob said...

The Ptolemaic model is not just more inelegant and cumbersome, it is actually wrong about reality in that there is nothing causing the sun to orbit the earth and there is a well-explained cause of the planets orbiting the sun and the moon orbiting the earth.

Economists have learned from Ptolemy's mistake. The solution is too keep the model elegant and force reality to fit the model. It's still cumbersome, but in a way that economists can live with.

Tom Hickey said...

Right. It is a matter of massaging your data long enough and not making it public.

Matthew Franko said...

https://www.youtube.com/watch?v=0jHsq36_NTU

John said...

"The models are equal as models but not as causal models of reality, as Newton's theory of gravitation proves beyond doubt. Scientific explanation is assumed to provide causal accounts, not just models that fit. Many models can fit a set of conditions. The challenge for science is distinguishing which are correct and this is accomplished by causal accounts."

I referred to this weird turn of events in another thread. Even this doesn't really grapple with what's happened to theoretical physics. The aim is to account for readings on an experimental apparatus. Seeking to explain the real world is a fool's errand. Thus Hawking's claim about Ptolemy's astronomy.

With its pretentious physics envy, economics has seemingly imbibed the same philosophy but only in order to insert ideology under the guise of prediction. This was made famous by Milton Friedman and his ludicrous instrumentalism: all that matters is the ability to make predictions, a sly way to divert attention from the ideological foundations he and likeminded propagandists were successfully embedding into this value-free "science". How unsurprising then that the predictions are as bad as the ideological foundations. Garbage in, garbage out.

John said...

By way of history, the Copernican idea may have been more elegant, but it predictions were much worse than the Ptolemaic ones. It required Brahe's later data and Kepler's elliptic orbits to make the observations fit.

For a different and very interesting take on Galileo and the Church, see Paul Feyerabend. His conclusion is that Galileo was knowingly dishonest in the way he presented the heliocentric evidence. The geocentric model (based on Aristotelian and Ptolemaic theories) was much stronger than the heliocentric model. The Church repeatedly exposed Galileo's deceptions and misdirections. The Church was left with little choice but to punish Galileo. Nevertheless, the Church went easy on him.