Tuesday, March 1, 2016

Jason Smith — Post Keynesian blues


Jason Smith comments on Post Keynesianism in light of the debate begun by Noah Smith on "occultism."

Information Transfer Economics
Post Keynesian blues
Jason Smith

3 comments:

Neil Wilson said...

It's very dangerous having yet another round of physics envy in economics.

People are not boron atoms or ants.

Matt Franko said...

Neil they cant separate the math from the physics theory...

We should be using math but it is not "the same thing" as physics or biology or fluid dynamics, etc...

The mathematical illustrations may look similar but it is not "the same thing"...

Its not easy to come up with entirely new theory which includes proper mathematical illustration...

And it has to go beyond correlation studies at some point and on to predictive theory which includes the time domain... which NOBODY is doing..

Tom Hickey said...

Based on past performance, I am skeptical of heavy use of math is the social sciences, as well as psychology. It's been over a hundred years since Alfred Marshall, and economics hasn't come up with a general theory that passes the smell test. Moreover, Jason Smith points out that economics doesn't even have a scientific framework for a theory. Same with other social sciences and psychology.

Maybe there is a reason for this based on the subject matter. Results cannot be more precise then the subject matter permits. Experimentation is impossible, impractical or unethical. So most of the research is based on simulation and approximation. Furthermore, much data is based on "expectations," and expectations are notoriously difficult to determine, let alone quantify. Examining reports of past Fed meetings, for example, there is a lot of speculation and guessing. This is at the top of the data chain with the best information available.

Moreover, Lars Syll tirelessly points out the misuse of statistics, too. Many social scientists are heavy users of statistics without fully understanding the foundations. This is a point that Keynes also called attention to. Cet. par. almost never holds in the real world, so the model is a model based on conventional assumptions is a model of some other world that only exists in imagination.

A good case in point is a recent study about results that cannot be replicated.

This is not only in the social sciences. Good research in medicine is also problematic.

Good science is actually hard. Economists introduce "restrictive assumptions" for "methodological convenience" and "mathematical tractability." While they achieve mathematical consistency necessary for deductive proof, they fail at both causal explanation and predication, which is the basis for empirical testing. The conclusion is that these endeavors are mathematical excursions rather than scientific ones.

The result is piles of BS passing for science because it is dressed up in the garb of science. Cui bono?