Jason Smith — The representative macro-theory agent differs from micro-theory agents

Jason Smith comments on David Glasner's recent post.

Here is my comment, which I tried to post there without success. The problem is not unique to that site and it affects other blogs using Blogger, although not MNE. Seems to be a Blogger issue:

I think that what Glasner is saying in summary is that the representative agent plays a key role in general equilibrium theory in economics to make GE modeling tractable. The basic idea is that "free" economies tend to general equilibrium naturally, even though they may never actually converge on it at a point in time.

Then the question logically arises that if human agency expressed in markets is always tending toward GE, whhy the chronic boom-bust cycles that affect free market economies.

Glasner is saying that the reasons given are either implausible or conveniently accidental, and not the result of human agency, which is assumed to be "rational." The representative agent is an aggregate of rational actors rationally pursuing somewhat homogenous preferences targeted at utility maximization.

The implication is that such models are really just tautologies that can't be tested because of the role the representative agent plays in the model. The definition of representative agent embeds equilibrium in the assumptions.  The model is internally consistent and can't be disproved from within.

In fact, we regularly heard that the model did not fail when it failed to predict the GFC because the shock that resulted in the crisis was "exogenous to" the model. How can a model be expected to foresee "acts of God." Econ is not fortune-telling.

Then, when narrative alternative explanations were offered, such as the financial fraud that the FBI warned was rampant at the end of 2004, the retort was, "Where's your model?"

Information Transfer Economics
The representative macro-theory agent differs from micro-theory agents
Jason Smith

See also

Lars P. Syll’s Blog
Representative agent models — macroeconomic foundations made of sand
Lars P. Syll | Professor, Malmo University

"We believe that the confounding of the aggregate with the individual is as dangerous as it is pervasive...."
—Angus Deaton and John Muellbauer, Economics and Consumer Behavior, page 81.

Richard's Real Estate And Urban Economics Blog
A book that changed my life

Richard Green

Jason Smith — Economics as and versus social science

My blurb on how I think that the whole "economics is too complex to make neat mathematical models" argument really just tends to assume itself (i.e. the original meaning of question begging) was picked up over at Mike Norman Economics, where it was put in an interesting way. Sociologists et al say economists need to prove economics is not too complex to model and economists say sociologists et al need to prove it is too complex to model.

That is to say we have sociologists and economists making a play for the null hypothesis.

My personal view is that economists should get the null hypothesis in this case, but not for reasons that economists think they should.…

Information Transfer Economics
Economics as and versus social science
Jason Smith

Jason Smith — Noah is stealing my material

More on econ as science.

Noah says to prove the assumptions are wrong. Lars says show that they are correct. Who has the better case?

The issue is how well a theoretical model works. Assumptions are always simplifications for economy and tractability of explanation. 

There is nothing inherently wrong about assumptions not being precise. They only need to be precise enough to yield results within an acceptable degree of tolerance.

The proof of the pudding is though hypothesis testing more than verifying assumptions, although if assumptions are not reasonably correct, then the model is questionable as an explanation that is generalizable. 

It is possible that a dodgy model can sometime yield positive results (pace Friedman's instrumentalism), as broken clock is correct twice a day. The test of theory is how well the model performs as an explanation of how things stand over time, that is, taking change into account. Theoretical models that don't reliably predict as not generalizable explanations. They are only generalizable in terms of restrictive assumptions that may or not hold in specific cases. That is to say they are models of special cases. 

I would say that scientific method is applicable in econ as it is in other social sciences and also in philosophy, since even speculation must take established truths into account. The question really is whether econ is a natural science like physics, a hybrid science positioned between natural science and life and social sciences, a narrative explanation of occurrences like history, or speculation based on principles grounded in intuition, like speculative philosophy. 

I would say that econ as practiced is some of each, and all approaches make their own contributions to the field. Defenders of econ as science can cite examples to make their case, and opponents can do likewise.

Jason Smith has written on this previously, to which I have linked here at MNE.

Is human agency Noah's big unchallenged assumption?

Falsifiabilité, simplicité, succès ... ou la mort

On the use of hypotheses: or, what do you get when you assume non-ergodicity?

Prescott and Lucas aren't Romer's problem

What does the AD-AS model mean?

There are more posts there, but these are representative.

The title,  Is human agency Noah's big unchallenged assumption?,  hits the nail on the head. The social sciences are about human agency, and so are behavioral psychology and motivational psychology and some other branches of psychology. So is theory of history. And in philosophy, so are theory of man, ethics, theory of action, and social and political philosophy.

Economics is based on a theory of man and theory of action. The theory of man and of human action are not subjects of study in the field of economics, or at least not exclusively so (pace Ludwig von Mises).

There is no agreed upon general theory of man or of human action in either the sciences or philosophy. Why? Foundational disagreement is often due to lack of criteria that are agreed upon, or failure of agreed upon criteria to determine a definitive result. It is also possible that data or method are insufficient. 

As a result, general agreement is usually limited to quite specific cases under particular conditions that are not generalizable, that is, special cases. As a result, the case method is generally used in business schools, for instance, that than theoretical economics.

Information Transfer Economics
Noah is stealing my material
Jason Smith

Jason Smith — Is human agency Noah's big unchallenged assumption?

More on foundations.

Information Transfer Economics
Is human agency Noah's big unchallenged assumption?
Jason Smith