August 2020

Sun Mon Tue Wed Thu Fri Sat
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 31          
Blog powered by Typepad

« Hayek at His Most Misesian | Main | Austrian Cycle Theory is Not a Morality Play »


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

The piece being criticized is here:

The entire symposium, including Kirzner's reply and Steve's comment, is here:

Dr. Horwitz or Dr. Klein, do you know of a recorded opinion of Hayek on Kirzner's entrepreneurship theory? Has Hayek ever commented on Kirzner?

Thank you

I thought that the point about the Communist Manifesto was to show that "coordination" in not much of a social welfare criterion.

We can think of the desire on the part of agents to effect certain transactions that they find in their interests as enhancing their subjective well-being. So in this sense mutual plan compatibility is something agents want. (Of course, they do not intentionally seek system-wide plan coordination. That is the result of their individual behavior but no part of their design.)

The issue is the distinction between welfare economics in its normative sense and in its positive or analytical sense.

Kirzner's coordination criterion explains the actions of individuals. People will not stop adjusting their plans and actions until a state of coordination is reached.

However, if this state is reached it does not mean that the individuals are in a Pareto-optimal state. (Hayek says as much.) They may be in a Nash equilibrium that is not Pareto optimal. So they may each be better off, by their own lights, if they were in some other equilibrium. (Consider, for example, the free-rider equilibrium with respect to a public good.)

Another distinct point is that even if the agents are in a Pareto optimal equilibrium, each agent could look at the system as a whole and say "What a revoltin' development this is." (If you are old enough or familiar with TV history you will understand the reference.) The individual as a citizen or social philosopher may believe that the outcome is not a "good" one.

I believe that Kirzner has admitted (or at least would admit) all this.

So what I think Klein and Briggeman are saying, on this point, is that there is no Kirznerian normative economics.

And that's fine with me. So I guess I am agreeing with Steve Horwitz.

The good news is that if the economy is in fact in a far-from-equilibrum state, it is always being "jiggled" (or the sides "melted", to use a term closer to that used by Stuart Kauffman), and thus if people find themselves trapped in suboptimal minima, they will eventually pop out of them. But this is only true if the economy is in a far-from-equilibrium state rather than having multiple stable equilibria. Ensuring the economy is in a far-from-equilibrum state will help prevent such daffy situations.

On the positive and normative, what Mario said.

Is he entrepreneur a person or a function? Are the mere optimizers and entrepreneurs, distinct classes, or are we are all entrpreneurs?

How does Kirznerian entrepreneurial discovery differ from Alchian searching?

On Jerry's remark above. The issue of function vs. person is important. One gets the idea from Mises and Kirzner that it is a function. But are all individuals equally capable of that function? Based on casual observation it would seem that some people have a greater propensity to discover opportunities than others. Is that because the former are in a more conducive situation or context? Or is it because they have greater talent? We have only just begun to explore these issues.

Jack Hirshleifer used to quip that some people don't have a labor endowment. Entrepreneurial talent does not appear to be normally distributed. Sowll's work on migration patterns is relevant.

Mario may remember a talk by Jim Buchanan in which he postulated that entrepreneurs must be inherently optimistic people. Since it was a psychological explanation of entrepreneurship, I doubt it appealed to Kirzner.

Anyone see this at the Adam Smith Institute blog:

"I used to go to conferences where incredibly intelligent and well-read liberal political scientists would agitatedly debate issues. The discussions would be full of ifs, buts, maybes and on the other hands, ending up in a sort of confusing fog.

"Then FA Hayek would creak to the microphone, and the fog would be gone. He'd penetrate straight through all the ifs, buts, maybes and on the other hands, and show how simple the core problem was. Then he would show how obvious the solution was. Everyone would look stunned. The old man was right as usual. Next business?"

Rizzo: "The issue of function vs. person is important. One gets the idea from Mises and Kirzner that it is a function."

Huerta de Soto has an interesting take on this in Socialism, Economic Calculation and Entrepreneurship.

He quotes Kirzner, Mises and Hayek and concludes that without the function, the man doesn't matter (without the function no man can do well), but with the function, the man is very important in practical terms.

I think that Steve cites too few Austrian economists in the title of this. You need to get Menger, Wieser, Bohm-Bawerk, Rothbard, and Lachmann in there too, if you want anybody to take this really seriously, :-)!

I would direct the honorable members of this board to Thomsen's excellent book --

-- and the accompanying book review by J. Mueller. How many re-interpretations must these eager students suffer before the scions of the Austrian tradition open their doors to new ideas?

Dear austrian:

What on earth are you talking about? There are more than 25 books in the series in which the Thomsen book appeared. Lots of new ideas!

Perhaps it is the solution to the problem

Steve Horwitz writes that "Hayek goes on to say that traditional equilibrium theory assumes away the problem by assuming that everyone knows everything." This isn't true. Neither is this general equilibrium theory, nor is this Hayek's interpretation of it. The clue in GE analysis is that information is local, aggregated by price systems. Hayek's point is that such information is not "given data", but discovered by the process of competition.

... Hayek is explicit that the assumption of perfect foresight is an equilibrium condition, that is, it presumes a solution to the coordination problem and insofar avoids an explanation of how a solution is achieved.

I am not sure what Arash means by "general equilibrium" theory in this context -- that is, when Hayek was writing in the 1930s and 1940s on these issues.

Certainly, in Walras' own conception the "cryer" (the "auctioneer") acquires the full and perfect knowledge of all market supply and demand conditions, and informs the traders at what prices they may now exchange, so to assure a simultaneous "balance" of all markets.

In Pareto's formuation of this, in his "Manuel of Political Economy," he distinguishes between "logical" and "non-logical" action. Logical action, with which economics is primariy concerned, assumes that through sufficient repitition of actions and trades in the market, all market participants have obtained "objectively" correct knowledge of their means and ends, the production transformations, and the market trade-off options. It was in this setting that one derived the conditions and solutions for "general equilibrium."

Non-Logical action concerns those human activities, which were the research domain of sociologists, Pareto said, that included situations of less than perfect knowledge, psycholgoical and ideological biases and prejudices, and institutional uncertainties that resulted in less than "objectively" correct outcomes and indeterminate "solutions."

And certainly the way, by the 1930s and 1940s, the perfect competition model was usually coming to be formulated, it assumed perfect knowledge on the part of the agents (See, Hayek's critique in 'The Meaning of Competition.')

Or have I misunderstood your statement and seeming distinctions?

Richard Ebeling


Do you know of early Austrian responses to Pareto's distinction between logical and non-logical action?

Dear Mr. Ebeling,

Walras's auctioneer relies on the aggregate excess demand function that represents the economic system. He does doesn't know the local information of each agent, but can rely on the aggregate excess demand function. Neither the auctioneer, nor the agents have full information (and I'm talking about Walras's original exposition and not only about modern formulations). The omniscient dictator is more at home in Austrian analysis, e.g. Wieser, and is called social planner today. The social planner problem, also used by Hayek in Utility Analysis and Interest (1936), is not the competitive equilibrium approach usually associated with Walras. Best wishes.

Um, Walras himself had no auctioneer or "cryer." It ain't there. Jaffe somewhere has an extended footnote on this point. Jaffe correctly notes that for most purposes no harm is done by interpreting Walras to have an auctioneer. But strictly speaking, it is simply not there. Prices are "cried," and it is not clear by whom. When QS exceeds QD, the cried-out prices fall; when QS is less than QD, they rise. No one is assumed to have "full and perfect knowledge of all market supply and demand conditions" or "the aggregate excess demand function." Perhaps memory fails, but I do not recall any explicit knowledge assumptions in the Elements.

By the time Hayek is writing his knowledge papers, however, he can complain that we were vague on knowledge assumptions and would often say something to the effect that all relevant knowledge is "given," without investigating how and to whom. The overall result, he says, is that we were missing the role of prices in conveying information and allowing the system to make use of dispersed information.

As far as I know, it was only after Hayek that we could say things like, "The clue in GE analysis is that information is local, aggregated by price systems." Hayek is credited with the insight about utilizing dispersed knowledge for a reason, I think. And Hayek considered the insight to be one of his discoveries rather than something he merely brought out in earlier writers.


I understand your description of Walras' process. However the "cryer" cries out alternative prices, and the auctioneer derives the individual supply and demand curves for each market participant, from which, he, then, constructs the respective market supply and demand curves. And the presumption, surely, in Walras' mind, is that this is the necessary and sufficient knowledge for determining a state of general equilibrium.

So for Walras, from my reading, this is all the relevant knowledge that is needed and meaningful for coordination. With an auctioneer of this type, the issue or the meaningfulness of Hayekian special knowledge of time and place, etc., doesn't even arise in the argument.

In Pareto's analysis, the pattern of repeated actions gives the individual market participants the required "objective" knowledge of all relevant market conditions for them to make no errors. That is, to assure that they supply and demand and price goods -- continuously -- at the correct amounts and relationships. They have in the relevant sense, "perfect knowledge." And which results in Pareto defining it as "Logical Action."

(Jevons also speaks of a "perfect market" in which the agents know the market conditions with sufficient completeness that they know all that is needed to know, to know the correct equilibrium prices at which they should trade.)

Frank Knight, in the first part of "Risk, Uncertainty, and Profit" (1921) is usually credited with the first formal, detailed specification of the conditions for perfect competition, and he does assume that the market participants have "perfect knowledge."

But Knight understood, from the start, the unreality and, indeed, absurdity, of this assumption. Several years before his book appeared, in an article entitled, 'Neglected Factors in the Problem of Normal Interest' "Quarterly Journal of Economics" (Feb. 1916) p. 283, Knight referred to "the impossible conditions of ideally perfect competition, where time and space were annihilated and universal omniscience prevailed."

Indeed, this explains the attention that he gives to the nature and relevancy of "risk" and "uncertainty" in the remainder of his book.

Yes, I am aware of the use that Wieser and in some places, such as you mention, Hayek make of the idea of the omniscient "planner."

However, in the case of, say, Walras there is no intention to use this as a conceptual "first approximation" of a perfectly coordinated market to then ask and pursue the question of how a world of decentralized, real "groping" by the market participants, themselves, would move the market to a possible state of general equilibrium that the "crier" is artificially allowed to solve.

(Walras' only reference to this problem, in his "Elements of Pure Economics," is his passing comment [pp. 184-185] to the fact that the market "empirically" solves the problem that his system of equations are supposed to only represent a logic explanation of.

(And Hayek has pointed out that Pareto, too, said that the actual problem of market equilibrium is worked out in the market itself, the complexity of which could not really be solved, empirically, with his system of equations. ["Manual of Political Economy, p. 171].)

While I admit that I may be trying to give too much of a generous interpretation, at least Wieser, (if we take his at other works into consideration as well, other than only "Natural Value") understood some aspects of these "knowledge problems." In his "Social Economics" (1914) he very clearly emphasizes, as much as Menger to whom he refers, the evolutionary processes that generate socially useful institutions, which the individual social participants whose actions are creating them could neither know or appreciate, even though their actions were the elements in the emergence and formation of those institutions. Only, long after, may the analyst, with historical reflection, explain how and what has been generated. (p. 162)

And in his "The Law of Power," (1926) Wieser explains the unique and special role and function of the entrepreneur who sees possibilities and undertakes actions to bring about innovative change in the market that others in the market neither see nor have the same knowledge to appreciate. And which brings about new (coordinated?) patterns in market relationships over time. (pp. 347-348).

And, certainly, for Hayek, any such conception of the omniscient "planner" is really a foil for various analytical purposes, the real purpose of which is to explain the "indogenous" process of knowledge use and tendencies toward coordination.

Richard Ebeling

@ Roger,

you are talking about footnote 27 in Jaffé 1980: 538. As Donald Walker argues, Jaffé's account is wrong. There are cryers. Not a super-auctioneer, of course, but the sellers and buyers themselves. Walker writes against Jaffé's interpretation (and translation):

"In fact, Walras specified precisely many times by whom prices are cried, and he made clear they could not possibly cried by an auctioneer. He stated that suggested prices and the associated individual supply and demand quantities are cried out by the buyers and sellers - the owners of commodities and money - in some markets and by their agents in others." (Walker 1996: 85) Walras is in fact about real-world markets.

What is true, however, is that I myself made use of the auctioneer to make a more important point about Walras (the local-knowledge point). With respect to the knowledge problem, we can in fact forget the real-world groping of Walras and resume to the stylized tâtonnement process dominating Neo-Walrasian theory (since this is a blog, I didn't specify this short-cut; my bad). If you don't believe this, stay with Walras's real-world groping and resume to Walker:

"In fact, Walras described the markets in his models as 'freely' competitive, and stated they are perfectly 'organized', not that they are perfectly competitive in the sense of traders having perfect 'knowledge'. That they are not perfectly competitive in that sense is evident from the workings of his oral pledges model, for such knowledge would mean that the traders or an omniscient price-setter know the market supply and demand FUNCTIONS and hence the solution price before trade occurs" (p. 79, my emphasis).

Thus, my argument stays intact even though I used a misleading short-cut by introducing the language of modern Walrasian theory. Your argument that Hayek somehow innovated the local knowledge stuff is insofar incorrect.

@ Mr. Ebeling,

I like your detailed response a lot. And I agree with you on Wieser. I cannot agree with you on this: "However, in the case of, say, Walras there is no intention to use this as a conceptual "first approximation" of a perfectly coordinated market to then ask and pursue the question of how a world of decentralized, real "groping" by the market participants, themselves, would move the market to a possible state of general equilibrium that the "crier" is artificially allowed to solve." Please go to my response to Roger to see why.

Also evident in my Walker-quotes above, I further disagree on this:

"However the "cryer" cries out alternative prices, and the auctioneer derives the individual supply and demand curves for each market participant, from which, he, then, constructs the respective market supply and demand curves. And the presumption, surely, in Walras' mind, is that this is the necessary and sufficient knowledge for determining a state of general equilibrium."

In contrast to my response to Roger, I return to the language of modern Walrasian theory. The auctioneer doesn't construct excess demand CURVES. This kind of knowledge would indeed be prohibitively restrictive. In fact, however, the auctioneer is a differential equation that drives prices up, if excess demand is positive, and drive them down, if excess demand is negative. That's it.

Further, I agree that "for Hayek, any such conception of the omniscient "planner" is really a foil for various analytical purposes", but for other analytical purposes he makes use of it. I mentioned his growth theory in Utility Analsis and interest. I in fact formulated Hayek as optimal growth theorist, showing that he anticipated some formal general equilibrium theory, employing the social planner. I'm on page 40! So it should be finished anytime soon.

The thing that I find confusing in these discussions is that "knowledge" is used to represent two things, the neo-classical perfect knowledge that must be assumed to calculate an equilibrium in a model, and the knowledge that is needed in reality for buyers and sellers to coordinate their plans. Vernon Smith's (and subsequent others) experiments suggest that perfect knowledge in the NC sense is neither necessary nor sufficient for market equilibrium in the lab. So the market as a platform for buyers and sellers to interact generates a different sort of knowledge more akin to stimulus and response knowledge that is totally unrelated to NC perfect knowledge. Over time this knowledge builds to the point where plans are aligned. To me Kirzner's entrepreneur alters the equilibrium to which this knowledge guides participants over time. So I like the description that Arash Molavi uses above except to say that the auctioneer is unnecessary. The market seems to allow participants in Labs to discover the direction in which they should shift prices.


You description of the auctioneer as an "a differential equation that drives prices up, if excess demand is positive, and drive them down, if excess demand is negative."

This, of course, makes the entire idea even more ludicrous than any hypothetical human auctioneer. After all, we can conceive of a human "crier" crying out prices, recording them, and "drawing" out supply and demand curves.

But a differential equation does not think, "cry" or record. It truly becomes "the God in the machine."

It makes Boehm-Bawerk's simply horse market, where the buyers and sellers form expectations and initiate buyer bids and seller offers moving toward an equilibrium between the "marginal pairs" far more common sensically convincing -- as least as a first approximation conception of the type of thing that actually might go on!

Richard Ebeling

Oh, I totally agree. I never claimed that the auctioneer is about reality. I entered this thread because of Steve's claim about general equilibrium analysis. I never argued that GE analysis is about the real world, though I differ with most Austrians in that I find that it ISOLATES some interesting points, like the decentralization properties of the market systems. When it comes to normative statements, I usually side with you guys. I differ in that I think GE analysis is a great step towards understanding how reality does not work. As you see, I like negative results. Further, as a Hayek scholar I believe that the Austrian bias against GE analysis is accountable for some blind spots with regard to Hayek's earlier work. I mentioned Utility Analysis and Interest, which put him at the forefront of modern economics. I try to show that his ideas were taken up by general equilibrium theorists like Uzawa, Epstein, Boyd, R. Becker, Bewley, ... i.e., the who is who of pure theory. I believe that if mainstream audience comes to appreciate Hayek's contribution to their own field of interest, Austrians gain some advantage to convince them of the significance of Hayek's more mature and more sceptical work.

@ Doug,

I like V. Smith's work, but the problem is how to generalize experimental results to the market system as a whole. The convergence on single markets or in labs is granted. The famous Gode and Sunders paper showed that zero intelligent but constraint behavior may also lead to convergence on single market. What is still unknown is how different markets are coordinated, some being substitutes, some being complementary. Further, even if knowledge is generated by the market process, the coordination problem remains unresolved. This is a strength of GE analysis: even if all knowledge is "given data", that is, even if we take for granted that there is no need for learning, we still have no clue how an equilibrium can possible come into existence. Isn't that cool?


This is one of the most fascinating, and at the same time, frustrating aspects of economics, including Austrian Economics, it seems to me.

We see everyday a wide and continuous degree of coordination. To use the phrase attributed to Bastiat, Paris DOES get fed.

And not just fed, but with a huge variety of diverse goods and services, in a world that everyday is changing in various small and not so small ways.

And prices, clearly, do assist that process to "guide" people into the directions that reflect where consumer and market demands are greater rather than less.

And, in the division of labor there are entrepreneurs who are, also, constantly doing a wide variety of "Schumpeterian" and "Kirznerian" things.

Plus, taking the relatively longer view, the "profit and loss system" does seem to tend to weed out the "bad" entrepreneurs and "reward" the "good" entrepreneurs.

And all of this is occurring in that Hayekian world of imperfect and decentralized knowledge of various types.

In other words, the market works.

But if one thinks, in general how "mainstream" economists have attempted to analyze markets in the general setting of general equilibrium theory, and the Austrians, say, with their general "market process" approach that they have attempted to develop using Menger, Boehm-Bawerk, Mises, Hayek, and Kirzner, over the last one hundred years, it is amazing much still is not fully explained or understood about how market actually work in terms of the "equilibration" or "coordinating" processes.

It seems we do not have to worry about any "end points" being reached in the research agendas for any foreseeable future.

Richard Ebeling

I completely agree. But since I like riddles, I find it more fascinating than frustrating ;-). I would give up one of my arms to see what answers the profession will provide in 500 years or so.

What if it's not equlibration but far-from-equilibirum processes that are maximally coordinating?


I think Jaffe is closer to the truth on Walras than Walker. Basically, Don makes Walras into an Englishman, but he was French. In the later editions of the Elements, W. started assuming no trading at false prices (as we put it today), which should suggest that his "groping" was not very real world. Walker's interpretations notwithstanding, I don't think you could find any hot quotes from Walras, or at least from the Elements, that reveal explicit assumptions about who knows what.

Anyway, I do agree that Austrians should not pooh-pooh GE, which some tend to do. It's a great model and has its place. It's clearly at work in much of Hayek's writing and in Mises's "evenly rotating economy." And I agree that Mises & Hayek had more influence on post-war orthodoxy than they generally get credit for, making them more "neoclassical" that we generally seem to think. These are important points worth pursuing IMHO.

Arash, I regularly read about the dichotomy between mainstream GE and Austrian dynamic markets. However, I don't see the dichotomy when it comes to theory. Hayek usually started his analyses from a position of equilibrium. And Mises had the evenly rotating economy which is very similar.

It seems that both Mises and Hayek saw equilibrium analysis as vital to advancing theory. It is the equivalent of the controlled experiment in the natural sciences.

The only time I see Austrians complaining about GE is when mainstream insists that it describes reality, which it doesn't. That's the reason for the attacks on GE, not because Austrians think there is no benefit in GE theorizing.

The comments to this entry are closed.

Our Books