Saturday, January 28, 2012

On The Lack Of Persuasiveness Of Austrian-School Economists

Mattheus von Guttenberg exemplifies what I think are defects in many fanboys of Austrian school economics. Among these defects is an uncritical acceptance of Ludwig von Mises' characterization of his own theories. And another defect is uncritical acceptance, likewise, of what Mises, or even worse, Murray Rothbard, had to say about the mainstream economics of their day. And a third defect is to apply these characterizations to mainstream economics of our day, while remaining quite ignorant of relevant trends in contemporary economics. Without more widespread correction of such defects, advocates of the Austrian school should not be able to persuade many economists, both orthodox and heterodox, of the worth of their views.

For this post, I focus on the theory of choice.

Here are examples of arguably a weak understanding of both the Austrian school and of mainstream economic theory:

"...we're not rejecting cardinal utility functions because it's hip and counter-culture. There's a distinct reason utility functions are impossible and unrealistic, and that's because utility cannot be known or measured... The degree to which we draw swooping utility functions overlaying cost curves is a unacceptable practice borrowed from coordinate geometry. Utility, again - is ordinal, it is intrinsically subjective, and it cannot be made known by other people." -- Mattheus von Guttenberg
"The concept of diminishing marginal utility is implicit in the logic of action, the Austrians just draw it to the fore." -- Mattheus von Guttenberg
The claim that utility reaches an interval-level measurement scale is a conclusion formally drawn from the Von Neumann and Morgenstern axioms (which can be considered independently of game theory). Most introductory economic textbooks claim that utility only reaches an ordinal-level measurement scale, anyways. The introductory textbooks have a different set of axioms, where choice among a set of goods with specified probability is not formally modeled. And they assert that the utility obtained is not interpersonally comparable. Mattheus' objections are not addressed to any views prominent in mainstream economic teaching for at least half a century. And to assert that diminishing marginal utility is consistent with utility reaching only an ordinal-level scale requires an argument. (I'm actually intrigued by J. Huston McCulloch's 1977 attempt to make such an argument, the one example of which I know in the last quarter of the last century.)

Mises incorrectly asserted that much of his theory could be deduced from a single postulate.

"The only axiom is 'man acts' and we draw the entire body of economic science spanning a thousand pages." -- Mattheus von Guttenberg
"...I have always been interested in rewriting [Human Action] 'as a set of numbered axioms, postulates, and syllogistic inferences using, say, Russell's Principia.' I believe it can be done." -- Mattheus von Guttenberg
I think such a rewriting, as it starts from the above informally stated premise, would be unconvincing.

Furthermore, the current state of decision theory suggests that analyses other than Mises' approach, are consistent with this axiom. The Austrian school approach is roughly akin to Samuelson's revealed preference theory. (One important difference is that Austrian advocates have some silly things to say about the impossibility of indifference.) Anyways, the idea is that an acting human, when presented with two lists of goods, decides between them. But social choice theory, as developed by, say, Amartya Sen in the late 1960s and early 1970s, has shown how to dispense with the formalization of choice as a binary relation as a primitive notion. Instead, one can start with a choice function, that is, a mapping from each menu that an agent might be presented with to a set of best choices for that menu. The derivation of a complete and transitive binary preference relation from a choice function requires additional structure on how menu choices relate across menus. And why the imposition of those additional requirements follows from human action needs to be argued. For example, why are not increasingly prevalent models, at least in research literature, of divided selves consistent with human action?

Update (3 July 2014): The blog free radical has a blog post pointing out Austrian confusions about mainstream teaching on ordinal utility.

7 comments:

Grant McDermott said...

I left two comments in the actual thread of Unlearning Econ's post that take issue with exactly the same thing.

This is not to pick on Mattheus; I like to think that our previous interactions have been pleasant and productive affairs. However, this whole ordinal vs cardinal utility debate is a bizarre non-controversy that has been allowed to persist for far too long.

To state the entirely obvious: Austrians will struggle to gain any kind of traction with other economists if they insist on clinging to demonstrably false distinctions like this.

Magpie said...

You touch an interesting subject in this post.

Let's compare the axiomatic base of neoclassical ordinal utility, which, as you pointed out, is what's required for decision problems that do not involve random outcomes (i.e. those forming the bulk of neoclassical microeconomics):

1. Axiom of Preference
each comparison of any two bundles A and B of goods and services results in one of:
(1) bundle A preferred to bundle B (A P B)
(2) bundle B preferred to bundle A (B P A)
(3) indifference between bundles A & B (A I B).
( P : "is preferred to")
( I : "is indifferent to")

2. Axiom of Transitivity
if A P B and B P C then A P C.

Whether one agrees with those two axioms or not, they have two desirable properties: (i). One can deduce all results in household demand theory from them; (ii). They are clearly, unambiguously stated (and, because of that, they can be empirically tested).

Compare this state of affairs with Mises' action axiom:

"HUMAN ACTION IS PURPOSEFUL BEHAVIOR. Or we may say: Action is will put into operation and transformed into an agency, is aiming at ends and goals, is the ego's meaningful response to stimuli and to the conditions of its environment, is a person's conscious adjustment to the state of the universe that determines his life. Such paraphrases may clarify the definition given and prevent possible misinterpretations. But THE DEFINITION ITSELF IS ADEQUATE AND DOES NOT NEED COMPLEMENT OF COMMENTARY". (Human Action, Chapter 1. Acting Man; Section 1. Purposeful Action and Animal Reaction. My emphasis).

Clearly, Austrian economists can easily deduce all their own household demand theory from Mises' action axiom. That is, it has property (i), too. The link below [*] shows how some use the action axiom to deduce their results (a warning: the examples could be unfair, but it was impossible for me to assess the credentials/knowledge of those providing the "proofs").

But it is undeniable that the axiom itself is neither clear, nor unambiguous: it does not satisfy property (ii). Incidentally, that's why the first chapter of Mises' book is entirely spent explaining it.

One such axiom cannot be put to empirical test.

In defense of Austrian economists, I will say that Austrian economics is broad, and I understand, in some senses quite heterogeneous, and perhaps (I am guessing here) some of them might use a, to put it bluntly, more serious axiomatization.

[*]http://mises.org/Community/forums/t/24572.aspx

Robert Vienneau said...

Thanks for the comment, stickman.

Magpie, I did think of quoting Edsger Dijkstra in the original post. He said something like, "You cannot go from the informal to the formal by formal means". I've previously touched on why I think neoclassical utility theory may be more complicated than textbook simplifications. But generally I agree with your contrast.

Magpie said...

Thanks Robert,

I did not know of the quote you provided ("You cannot go from the informal to the formal by formal means"), but that is the gist of my comment.

I fully agree with your opening remarks on uncritical acceptance and misleading characterization.

Where I have some doubt is here: "Without more widespread correction of such defects, advocates of the Austrian school should not be able to persuade many economists, both orthodox and heterodox, of the worth of their views."

Correcting those defects potentially could leave little of Austrian school economics (at least in its Misean variant, that I am more familiar with).

Besides, I don't think Austrians really care much about convincing scholars about the worth of their views. I suspect they are more interested in the wider public.

Philip Pilkington said...

For I think the first time in my life I think I agree with an Austrian -- or at least the sentiment expressed. Any sort of 'theory' of choice is probably nonsense when you drill down.

Even when sophisticated psychological experiments are used it is far too difficult to separate out the 'choice' made by the subject and the influence of environment in which the choice was made.

There is no basis on which to begin a realistic appraisal of Man except by assuming that he acts freely. And if we accept that he acts freely we must acknowledge that we cannot construct an apparatus which will explain his future free acts. These are very old philosophical arguments that go back to the 19th century at least -- if not further.

Does accepting them mean that microeconomics is doomed? Maybe.

Robert Vienneau said...

Philip,

I'm not sure I agree with you. But, I, too, find myself agreeing today with an Austrian school economist, namely Mario Rizzo.

Magpie said...

Robert,

Hopefully my two cents on the subject of rationality will be of some interest:

"Rationality = Methodological Individualism" http://aussiemagpie.blogspot.com.au/2012/02/rationality-methodological.html