September 2022

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30  
Blog powered by Typepad

« Coverage of my ASI Talk in The Spectator | Main | Markets in Everything: Delta Airlines More Efficiently Bumps Passengers »


Feed You can follow this conversation by subscribing to the comment feed for this post.


Let me add that one of the reasons that many economists like equilibrium theory is that it makes testing hypotheses easier. In the traditional theory, for example, a cet. par. supply shock (eg., political turmoil preventing the export of oil to the US) will have an unambiguous effect on equilibrium price -- it goes up.

But now look at the process -- depending on expectations the path of prices could be rather messy. What to predict outside of equilibrium?

This is one reason why the process approach is considered imprecise. I am sure that this whole issue is related to Hayek's idea of pattern prediction or explanation of the principle. However, I have not worked it out.

As Mario would surely agree, the precision and predictability of equilibrium theory is spurious even on its own grounds. Our own "arash" reminds us constantly of the Sonnenschein-Mantel-Debreu result that most properties of individual demand curves do not aggregate. Before SMD H. Hotelling wrote a great paper on Edgeworth taxation paradoxes showing that you cannot say whether a tax will raise or lower price in a general equilibrium context. ("Edgeworth's Taxation Paradox and the Nature of Supply and Demand Functions", 1932, JPE) I wish we would pay more attention to that sort of thing.

Roger reminds me that during my first year in graduate school we studied an application of Harry Johnson's two-sector model of general equilibrium to tax incidence (and unionization as well). The upshot is that if there are different capital intensities in the two sectors the incidence of the corporate income tax is much more complicated than usually assumed. Now when we have a many-sector GE model knowing the incidence requires a great deal of factual knowledge. However, in principle, the answer is determinate.

The difficulty with process analysis is that while we could build a mechanical model specifying, say, expectations at every turn, it is unlikely to have wide applicability. There are problems of path-dependency, and so forth.

Right on, Pete, right on.

I think the point you make here buttresses your claims about Kirzner, by the way. After all I think it was Kirzner more than anyone who made the point that competition is _rivalry_ among potential competitors, not the _number_ of competitors.

And if _assessing_ claims about efficiency of the market process requires a formal model (re: your Stiglitz-Hayek point), then screw it. I was really inspired by Prof. Carilli at the SDAE to not talk the "efficiency" and "equilibrium" talk. We can't make efficiency claims in an objective way (like miles per gallon), so maybe we should be content with "markets work".

Finally, I thought I'd chime in that the copper pennies story really captures all this stuff. I can show people a crystal clear profit opportunity via current market prices to buy a commodity at 1 and sell it at 2. Why isn't everyone jumping on this and arbitraging it away in short order? All together now: "Transactions costs!" But nobody who has not explored this market and dreamed and thought and imagined and scribbled, can really grasp the true nature of the these costs. Indeed, these "transactions costs" are to some extent what you make them, as you poke around, learn from others, gain knowledge by experience, trial and error, and eureka moments. Eventually even some mentally slow entrepreneur like me can gain the knowledge and "conquer" the transactions costs, with machinery that runs 8 coins per second, knowledge of the best buy and dump banks, nicely developed relationships with the tellers (chocolate candy), etc.

You just can't formally "model" this stuff (how would you handle the chocolate "gifts," let alone the proper answers to various tellers' questions?). Nevertheless, where entrepreneurs are free to do so, they will latch on to this profit opportunity, as revealed by prices, and run with it. Markets work, baby!

Which leads to the question of what should Austrians do? Should we continue to try to reform mainstream? That seems to me like waiting for the drunk to recognize he is a drunk.

It also reminds me of the debates within the Church before the Reformation. People tried reform for centuries and it never worked because the people in power would benefit little from reform.

Maybe instead of trying to reform mainstream, Austrians should strike out boldly on their own, Reformation style, and make a clean break with mainstream. Advertise the differences. Win the battle of ideas in the market place instead of the journals.

Economic degrees are very popular with incoming students. I think most would be attracted to Austrian econ. Austrian challenged schools would envy the enrollment in Austrian schools and be forced to offer something in order to compete.


I like the theme you're developing...

I can see it now, Pete posting his "Ten Propositions of Austrian Economics" on the door of Princeton/ Harvard/ MIT econ. department!

Tyler, Thanks! I'm thinking we should just ignore mainstream and start our own "churches." My guess is that businesses would hire Austrians and governments would hire mainstreamers.

As Skousen points out in his text many in business already follow a version of the ABCT even though they know nothing of the theory. For example investment experts divide the stock market into cyclical and non-cyclical stocks, which is nothing but the Austrian division between capital goods and consumer goods production. Also, Austrian econ is very business friendly and practical whereas mainstream fixates on perfect competition.

Austrians should establish schools of economics that break sharply with mainstream and advertise it. Let the public decide.

PS, we have tried Hayek's plan for changing the intellectuals and made very little progress. Let's try something new instead of expecting different results from the same strategy.

In the 1970s and 1980s, there were a variety of attempts to deal with disequilibrium processes in both microeconomics and macroeocnomics, and by some leading up-and-coming mainstream economists (for example, Robert Barro).

But they all seem to have turned away from this "research agenda" after a period of time. And economics (more or less) fell back into a focus on equilibrium models and applications.

I think the reason is (as both Pete and Mario hav suggested) that the problems (expectations, changing knowledge, path-dependency sequences, etc.) make it extremely difficult to come up with precise and "predictable" conclusions.

What if they are not reducible to the methods used by mainstream economics? In other words, what if mathematics is not the appropriate "language" and logical tool to deal with this aspect of human actions (and the mental processes from which they arise)?

Perhaps it is best to try to do market process economics outside of the Neo-Classical conceptual "box."

In other words, to try to follow along and further develop the Austrian methods (and any complementary insights that may be useful from, say, new institutional economics, etc.), and see where this "research program" may lead.

If the mainstream refuses to recognize it, or assign it legitamacy, well, so be it.

After all, think of the uphill fight the early Austrians had in the German-speaking academic world for decades due to the domination of the profession by the Historical School at German universities. And there, too, the dividing line was, to a great extent, based upon differing conceptions of "methods" and "tools" of analysis.

Menger, Boehm-Bawerk, and Wieser did not conform (or to my knowledge did not recommend to their students to conform) to the Historicist methods. But rather to show the theoretical "fruit" that the new subjectivist-marginalist "tree" could bear.

Richard Ebeling

Preach it Dr. Ebeling!

Look at how much Ron Paul has done for Austrian econ. I meet people quite often whom he introduced to Austrian econ for the first time. Glenn Beck has done a lot, too. The people are much more open to the truth than are the scholars, who have too much invested in the old, failed paradigm.

We could promote Austrian econ as anti-establishment, which appeals to young people.

@ Richard Ebeling,

let's make a bet. Once there will be proper mathematical tools to display proper disequilibrium dynamics. We are not there yet. But this does not imply that the language the profession will detect is non-mathematical. I think "Roger" agrees with me on this. The only problem: we won't be alive to see it. ;-)

@ "Roger",

I love you too. :-)

@ the other Roger,

Austrians economics detached from mainline economics (to use Peter's nice distinction between mainline and mainstream) is in danger of becoming a 'sect' instead of a respectable church (whatever that may be). See the Mises Institute.


I have no desire to exclude or 'challenge' anyone's use of the method that they find useful.

And, indeed, it is of course possible that in the future some mathematical methods may be found to solve or better 'handle' such analytical problems.

But . . . as of now, that does not seem to be the case. So I think that those Austrians, who so desire, should "go their own way," and see what they can succeed in doing in dealing with market process analysis.

Richard Ebeling

Math isn't the problem. The insistence that equilibrium accurately reflects the real world is the problem. I'm certain that good econometric models using Austrian theory could be developed and I have seen some excellent steps toward them by several authors. And I have promoted the use of structural equation modeling before. And Hayek uses a lot of graphs, which are nothing but visual representations of math models.

As for the Mises Institute being a sect, that may be true, but they haven't limited themselves to economics; they engage in politics much of the time. They have chosen that role. But a university school of economics doesn't need to do that. It can stick with economics and ignore the politics.

There doesn't seem to be much interest in a "Luther" moment for Austrian scholars. Am I right? Why not? You have your own scholarly journals.

It's just that I don't plan a Luther moment and rather wait for it to arise spontaneously. ;-)

Take a look at the literature on far-from-equilibrium systems. If there is any math of any use, that's it.

If I may add one point of clarification, so I not be misunderstood.

When I said, above, that "Austrians," so motivated, should "go their own way," following their own methods of doing economics, I do not mean that Austrians should ignore, or fail to debate, discuss, and try to persuade those in the mainstream of the profession.

The early Austrians -- Menger to some extent, and certainly Boehm-Bawerk, and Wieser to a lesser extent -- all entered into debates and discussions (sometimes in heated and colorful language as seems to have been the custom in the German-language, even scholarly, literature of the time!) with their "mainstream," which was predominantly the Historical School.

But they did not let that mainstream set the "terms of the debate," or allow themselves to be deflected from formulating their new theories and ideas, and their applications, in the ways that they considered most useful and insightful by the prevailing mostly Historicist methods and concepts of the time.

I see know fundamental reason why we should not follow in their footsteps.

Richard Ebeling

Given the comments on math, I would like to say once again that computable economics has math that can be very useful from an "Austrian" POV.

I guess because the rules of the game have changed tremendously. The game is more professional so that it is less possible to have any impact from the outside. Most economists only listen, if you speak their language or somehow show that what you say is within their universe.

Instead of taking an honest outside position, I prefer a more strategic (or sneaky) approach: why not be Trojan horses and invade from within?

with respect to Roger's comment: computable economics is a good example for the strategic approach. Given that the profession respects mathematical arguments, why not formulating the limits of contemporary economics in such mathematical terms and offering Austrian narratives as potential beacons that guide the search for more appropriate tools?

Final comment on the Luther analogy: Even though you guys are certainly top NBA level, we don't have the Micheal Jordans, Magic Johnsons, or Larry Birds that could have impact by sheer ingenuity.

Years ago there appeared a book called "Less than Words Can Say" by Richard Mitchell

The book began with the contrast of two imaginary "tribes," one of which had verbs in its grammar and vocabulary, and the other tribe that did not.

He contrasted how and what each of these two groups would be able to say and express, given the structures of their respective languages.

Natural languages, it seems to me, often have the capacity to explain, express, and capture meanings that another (artificial) language may not. And even when attempts are made to translate these things into the artificial language, something may and can be "lost in translation."

It often seems to me that such is the attempt to restate logic and thoughts and meanings explained in a natural language through which people actually think, reason, experience, "feel," and understand their "meanings" of things into an artificial language such as mathematics.

I do not wish to be misunderstood. This is not a criticism of mathematics and its many uses, applications, and clarifications of certain types of relationships.

But that does not mean that all ways of thinking, experiencing, understanding, and conveying meanings can successfully be translated into it without such losses in translation. Losses that can represent crucial elements of the human condition, including aspects of "action" as lived in terms of man's interactions with others that represent aspects of what we call the "market process" and its workings.

What is really being search for, it sometimes seems to me, is a "closed system" of some form of determinism. A search for equilibrium systems that are able to incorporate disequilibrium states and show how and why in a deterministic fashion that disequilibrium position "has to" unfold along a particular path leading to a new or restored and precisely specified equilibrium.

In other words, the search for a holy grail in which everything is predictably predictable. Even stochastic systems are meant to be predictable, with the goal of even making the range and degree of the "unsystematic" predictable.

I ask, why? Why do we have to assume that the world works this way, at least all aspects of the human condition? Maybe all the attempts to reduce mind to matter -- how else in some ultimate sense could we successfully predict people's expectations, for instance? -- are misguided.

There is a certain strange "circle" in this pursuit. Shouldn't the analyst's search for a solution and the mental products of his search been predictable from the prior conditions? Because the search for the answer and any answer found must have been part of the predetermined equilibrium system?

Just accept the fact. Man has thoughts. He imagines things in ways differently than he did before, and acts upon it. And it generates unpredictable, and unintended, results for both him and for the analyst trying to determine his equilibrium conditions. And it may not all be successfully expressed in mathematics.

Mathematics may be a language that can only express less than words can say.

Richard Ebeling


I'm not sure I fully understand your comment. The parts about determinism and predictability seemed to suggest that math is somehow deterministic and predictable. Or, I suppose, it can only describe deterministic and predictable systems. Was that indeed a part of your message?

You know, I'm into the computability stuff and I have this recent JEBO paper that sort-of-kind-of pulls hermeneutics out of a computability hat, in a sense and as it were. So I resist any notion that math is somehow about determinism and predictability. Anyway, first things first. Please indicate whether you were indeed linking math to determinism and predictability.


No, I would not want to suggest that, necessarily.

But . . . I think there are some (many?) who use math and believe that is what it can do or should be used for, because they are looking for such a "closed system" and think that mathematics is the appropriate language to create such a system.

Richard Ebeling

@ Richard

Cool. It sounds like you think that sort of thing is more common in economics than I do. I think about multiple equilibria, sun spots, complexity, computable economics. But I am straying from your main point, which is methodological dualism if I have understood you. I am totally with you on that one. And you used the crucial word IMHO, meaning.

In the recent JEBO paper I've been hyping ("Some epistemological implications of economic complexity"), I try to get at the same basic point, but from the math side.

Let's consider a flock of birds. Basic geometries can indeed be modeled using some tidy equations, and when code is fed into a computer, we spin up the model and such a model flock of birds can look quite realistic as it moves along smoothly, sometimes lurching a little one way, then another, with no perceptible artificial time steps. Each bird in the flock is quite dumb, but still we know that a real flock is made up of individual birds each reacting as an individual bird. Still, we may learn a great deal about the "macro" state of the birds by watching the patterns evolve on the computer screen, all the while knowing nothing about what each bird is "thinking". Now here is where it gets tricky. Let's suppose a hunter shoots up amongst the flock of birds, or a raindrop falls on the bill of the lead bird, or a bird flying higher drops a torpedo on the lower bird. When we try to model the pattern reaction on the computer, the change in geometry and direction is amazingly non-linear. Tiny variations of the bullet’s position relative to the flock changes the pattern of the flock in very unpredictable ways. Now there are two important details that destroy the ability for even a supercomputer to accurately predict a precise change in flock geometry by integrating differential equations forward in time. First, there is the lack of ability to precisely initialize the precise location of the shot with respect to the flock. Even in the flock is perfectly mechanical, owing to non-linearity, the geometry of the flock in the future is highly sensitive to initial conditions. Second (and even more devastating to our model), even in the case of stupid birds, if we were theoretically able to ascertain a precise initial condition, our computer model would still suffer from the very real shortfall of not being able to code in the exact brain function of each bird. This may mean that our model is too simplistic to possibly model reality with precision. At some point into the forward iterations, it will quickly break down. Still, there are some very fixed rules which will limit the possibilities. First, each bird can fly only so fast or high. We know the flock will not ascend into the stratosphere. However, the rules or "boundary values" do not really add much utility to the model in this example. In the case of a human flock, it has been suggested that the brain function of an individual human being is somewhat more complex than that of a bird. The future motion of each human actor in the community will by necessity involve much more in the way of mathematical uncertainty, owing to the complex phenomena occurring within the human brain. Now like in many fields of learning, when we are not able to capture all the mathematically cogent details due to real limitations in knowledge of small processes, we often "parameterize" with spoken language or simplistic graphs, pictures, or more commonly in modern economics, highly truncated equations which do not delve into the fine detail of that which is being described. Biology does it, geology does it, and economics by necessity is forced to do it most of all due to the incredibly complex phenomena of human action. By parameterizing processes in spoken language, we beautifully work out details that cannot be captured even in the most sophisticated computer code, and we thereby communicate to others truth in a manner of speaking. However, when parameterizing with truncated mathematical equations, there is a real danger that the arrogant and unwise develop an ill-placed confidence that such truncated math can be fed into a line of code without sacrificing skill or vital information. Spoken language at its best is governed by wisdom and experience. Equations on the other hand just equate without regard to wisdom or experience. If some scientists insist (in the name of high brow "science") on holding fast to the fairy tale that by continuing to feed numbers into our simplified equations in the hopes of actually forecasting changes in a community of human beings, what do we call this other than scientism? Thus by our example, we see that in physics or even ecology, complex systems may only be governed by initial conditions and boundary conditions. In human social interaction, the situation is much more daunting for our predictive model, since predictability is also destroyed by vast oversimplification and truncated wooden equations that by their very nature can never be expected to accurately approximate reality. In many ways, many of these types of attempts at mathematical sophistication are major steps backwards in gaining of understanding of that which really determines the geometry of the flock, namely the individual actions of the individual birds.

K is dead-on right. And this is the kind of modeling I think is worth pursuing. It is also, btw, the very cutting-edge of science. The Austrians are in a unique position to jump on the complexity sciences and use them, as most of what the traditional Austrians talked about involved exactly what the complexity sciences claim. For example, what Hayek called spontaneous order, the complexity theorists call self-organizing processes. Problems with equilibrium? Complexity theorists talk about far-from-equilibrium states in which creativity is maximized. There are clear parallels between these two strains of thought, and the Austrians are in a unique position, already believing what the complexity theorists have since discovered, to bridge the two and demonstrate that the Austrians have been right all along -- from a scientific point of view.

Complexity theory shows promise and should be pursued, but what do people really want? They want a model to predict the business cycle. It doesn't have to explain everything in the economy; just predict with reasonable, useful accuracy when depressions are likely to occur. Look at how useful people find the simple model of the yield curve.

I would like to point out that at GMU we have the Center for Social Complexity, which is very much involved in this sort of modeling and has strong ties to the Santa Fe Institution -- see the work of Rob Axtel, and on the business cycle and macro see the write up in The Economist on Rob's work.

Roger Koppl's SDAE is right on this and actually should be a sort of rallying call for those of you who are intrigued by complexity science.


Thanks for that, Pete. I don't know how much life is left in this thread, but I did want to chime in with your on his original point. Which is the real thing, the formal theory or the appreciative theory? Now that question does not sustain too much critical scrutiny. Obviously each has its role and so on. But in some loose sense we could probably divide economists into those who think of the formal theory as the real thing and those who think of appreciative theory as the real thing. It seems pretty clear that a lot of the formal guys, though absolutely not all of them, think the formal stuff is the real thing and appreciative theory is just hand waving or otherwise inferior to formal theory. How mistaken they are! It is the appreciative theory that does most of the work in economics, including applied economics. Indeed, appreciative theory is the secret handshake of economists. You know when you're talking to an economist, because that person can do appreciative theory in casual conversation.

Should have been "chime in with you on your original post."

Let me add an "extraneous" formatting point. People should divide their posts into frequent paragraphs, especially on this blog. I, for one, find a monster paragraph on the computer screen to be off-putting and an effort to read.

I think much of this has to do with (my) expectations about computers -- everything needs to be fast and to the point. Books or journals, on the other hand, carry different expectations for me.

I wonder if younger people agree.

Mario, I think you're completely right. As I try to tell students, it's best to assume that your reader is lazy and stupid. If the reader has to work to make sense of what you're saying, the odds increase that she or he won't do so. Paragraphing is one of several complementary ways of holding the reader's hand and providing needed guidance.

"It is the appreciative theory that does most of the work in economics, including applied economics."

I would say this is true outside of economics as well. It seems to me that appreciative theory should come first. To the dismay of purely mechanical rational choice theorists, grand technical theories might best be formulated based first on some sound hermeneutics.

It should bother us a great deal when people assert that "the more grand a theory, the *less* it will be applicable to detailed problems." What has just happened is that their "grand theory" has paramaterized too much of the important detail, and truncated too much reality. It is a foolish assertion to state that models constructed on lines and lines of these types of functions will have much predictive skill.

Even in a physics problem, we often must step back a moment from the technical minutia of the problem, and *first* ask ourselves in verbal language, "what's really going on here?" It's often only after this process that we can delve into the technical details and try to formulate in a way that might be mathematically tractible.

In economics, I suspect it is often the case that we will find that there might be no good way to numerate *many* types of problems formally. I vaguely remember Hayek discussing a similar concept in Individualism and Economic Order, where he is laboring his points about the division between subjective and objective realities.

To come back just once more to the direction that Austrian econ should take, this article at on how the Fed controls economics is important:

Academic economics strives to supply the types of economists that its customers demand. Its main customers are government and the Fed, both of whom demand math economists who praise state and Fed efforts.

Few private enterprises employ economists because for the most part mainstream economics is worthless to private enterprise. I believe some Austrian schools should totally scrap their econ programs and re-write them from the ground up to orient the program toward private enterprise. Undergrad education should be primarily applied econ. "Managerial economics" programs are a step in that direction.

Those schools should make it clear to prospective students that credits won't transfer to traditional, state-oriented econ programs. And as some schools have a class in Austrian econ, Austrian schools could offer a class in mainstream econ so that students will at least be familiar with the differences.

Eventually, students will understand that if they want to work for the Fed or in government they should go to a mainstream econ program. If they want to work in private enterprise they should go to an Austrian school.

Dear Roger,

You make some very good points, but also there are some major misconceptions involved in the article that you link to, and more importantly to the article that links to.

Consider the following from the article:

"Paul Krugman, in Sunday's New York Times magazine, did his own autopsy of economics, asking "How Did Economists Get It So Wrong?" Krugman concludes that "[e]conomics, as a field, got in trouble because economists were seduced by the vision of a perfect, frictionless market system."

So who seduced them?

The Fed did it."

Note what is being claimed, the Fed perpetrated that belief in the perfect market theory. Really?!

The White article linked at the sight is worth looking at, but not necessarily the Huffington Post piece.

Now as to your depiction of the economics profession --- it is again both partially right, and misleadingly wrong. First, economists trained in formal technical economics are in very high demand in the private sector and in academia. A full professor at a top rated economics department in a private university will typically make well over $200k, some more than $300k. In the private sector, they will make significantly more than that. Many economists work in consulting and make significant fees --- consider the investment fund that Shleifer and Vishny ran that is discussed in The Myth of the Rational Market. Consider the employment options of an economists such as Fisher Black during his time on Wall Street.

Do I wish the market only valued those trained in Austrian economics, of course I would. But is that true, no. The market very much values advanced mathematical modeling skills and especially advanced statistical analysis skills. To suggest that the only jobs available for these people are to be at the Fed, etc. is a misunderstanding of the situation in the market economy.

The most valuable skill in the private sector is a demonstration that the individual knows how to learn --- learning how to learn so to speak. And there is a signaling issue associated with studying economics as well as a training issue. If you look at any data on jobs and graduate and professional school admission for students with social science degrees, then you will find that economics students are in a better position than their peers.

Just so you don't misunderstand --- I am very sympathetic to the general sentiment of your post, I just think you have some wishful thinking about the lack of market value for mainstream economics, and the high market valuation for Austrian economics. We who believe in the Austrian approach would do well to keep working harder and harder to make what you believe to be true, to actually be true. It will help tremendously spread the Austrian approach if we could in fact establish that.


I didn't mean to suggest that Austrian econ is now valued by private industry. It's hardly known. I was suggesting that Austrians could stake out that territory and strive to be known as valuable to industry.

Yes, there are jobs in private industry for mainstream economists. I didn't mean to suggest that there were no jobs. But in comparison to the number of jobs in government and the Fed, how many jobs exist in private industry (not colleges) for economists? And yes, investment firms hire economists, but do they hire more with finance degrees? And what about non-financial businesses, do they hire economists? Not very many.

I'm suggesting that Austrians could make a name for themselves as economists who are valuable to private enterprise in general. And yes, that will require training it at least advanced statistics, but Austrian theory behind that math will make the math much better.

PS, anyone know what the job market is for those with a BA/BS in economics? My guess is it's not very good. I think Austrian econ could change that if it takes an approach similar to managerial econ degrees.

The comments to this entry are closed.

Our Books