[From Mises Blog, July 11, 2006, archived version with comments (below).

For more on this issue see also The Great Mises-Hayek Dehomogenization/Economic Calculation Debate]

Knowledge vs. Calculation

July 11, 2006 9:24 PM

On occasion I’ll see someone try to smooth over the Mises-Hayek “dehomogenization” debate which argued whether and to what extent Mises’s and Hayek’s approaches to the impossibility of socialism differed. One side–what I’ll call the Rothbardian or praxeological-Misesian view–sees Mises’s insight as having to do with the use of money prices to serve as a cardinal unit for purposes of economic calculation. This approach is championed by Rothbard, Hoppe, Herbener, Salerno, Huelsmann, and others, and arguably Mises. This view also sees Hayek’s contribution as different, and as possibly confused or flawed: that prices help to spread otherwise localized information through the economy, thus enabling efficient use of resources. The Hayekians tend to emphasize the knowledge or informational aspects of money, but also maintain that this is just “the other side of the coin” of Mises’s insights.

See, e.g., Yeager, in Mises and Hayek on Calculation and Knowledge, “question[ing] the supposed distinction between calculation and knowledge problems.” See also: Pete Boettke, Hayek and Market Socialism: Science, Ideology, and Public Policy (Don Lavoie [in Rivalry and Central Planning, 1985] argued that one must read Mises and Hayek’s arguments as two sides of the same coin, and I follow him in this regard and will not dehomogenize their different contributions to the analysis of socialism”); also his Economic Calculation: The Austrian Contribution to Political Economy (“the essential argument that Mises and Hayek rose against socialist proposals–the problem of economic calculation–and their understanding of how the private property system affords monetary calculation are complementary contributions to economic theory”).

Also see Steve Horwitz, Monetary Calculation and the Unintended Extended Order: The Misesian Microfoundations of the Hayekian Great Society (“An Austrian economics for the 21st century is going to have to rediscover those Misesian insights and more fully integrate them with Hayek’s work on knowledge and coordination. … a “praxeological” social scientist has both a Hayekian and a Misesian task: The Hayekian task is to recognize and describe the nature of the unplanned order that is to be explained, while the Misesian task is to describe the process by which intentional human action is guided such that it can produce that Hayekian order. … The “de-homogenizers” have … correctly identified microfoundations [including] the importance of monetary calculation and Mises’s concept of “appraisement,” but … they ignore what seems to be the obvious relationship between those microfoundations and Hayek’s vision of the social order. That is, they ignore that the outcome of the use of economic calculation by individual entrepreneurial actors and by firms and households is precisely the “use of knowledge in society” that characterizes the Hayekian spontaneous market order.”).

Also: Bob Murphy in a recent post wondered: “I don’t understand why Salerno (and Kinsella and perhaps others too on their side of this) think it so crucial to hammer home the point that market prices don’t convey knowledge.” Murphy and I had some back-and-forth on this in the comments to this post, as well.

(Some more information is available on the Wikipedia entry on the economic calculation debate.)

So the Rothbardians/praxeologists view the Mises and Hayek approaches as different (and the latter as a weaker point, at best, or confused and distracting, at worst); while the Hayekians claim the approaches are complementary and intertwined.

On occasion I have corralled and summarized some of the resources but do this often enough that I thought it might be useful to put some of the links and references in one place. It is my view that the (primarily Rothbardian/praxeological) sources below, at the very least, make it difficult to argue that the two approaches are “two sides of the same coin”. Below is a brief discussion and summary of and some links to some of these arguments.

***

What I take to be the Rothbardian or praxeological-Misesian approach to the socialism-knowledge-calculation debate is found in the writings of: Salerno (Economic Calculation in the Socialist Commonwealth: Postscript: Why a Socialist Economy is “Impossible”, Reply to Leland B. Yeager on “Mises and Hayek on Calculation and Knowledge”, Ludwig von Mises as Social Rationalist), Hoppe (Socialism: A Property or Knowledge Problem?), Hülsmann (Knowledge, Judgment, and the Use of Property), Herbener (Calculation and the Question of Arithmetic; Ludwig von Mises and the Austrian School of Economics), Rothbard (The End of Socialism and the Calculation Debate Revisited), and, of course, Mises (The Equations Of Mathematical Economics And The Problem Of Economic Calculation In A Socialist State; Economic Calculation in the Socialist Commonwealth; Human Action, esp. Ch. 16, Secs. 1, 2, and 3).

A summary of some of these views is found in my essay Knowledge, Calculation, Conflict, and Law. See. e.g. p. 53 and n. 8, discussing Hülsmann‘s discussion of Hayek’s tin example:

In this example, what information, exactly, is supposed to be conveyed by prices? Let us explore the possibilities. Can the original cause of the price increase (i.e., the change in demand or supply) itself be conveyed via prices? Well, no. Prices are the result of action. Thus, action that changes the prices must already be informed by knowledge.8

8 In other words, the prices generated on the market are past prices, which are always the outcome of action, not its cause. Hülsmann (p. 26) explains that “all information that this action was based upon had to be acquired beforehand. The price itself could not have communicated the knowledge that brought it [the price] about.” With regard to the tin example, “tin does not become scarcer and then this fact can come to be known to someone and lead to adaptations. Rather it is the other way around. The very fact that demand increases means that someone already knows of a more value-productive employment of tin” (p. 28).

Note that even Hayek says that mere users of tin do not know “anything at all about the original cause of these changes.” So prices might rise for a number of reasons: 1. because some people correctly assess that supply is reduced and therefore bid prices up; 2. because some people mistakenly believe supply is reduced and therefore bid prices up; 3. because some people correctly assess that demand will increase; 4. because some people mistakenly forecast that demand will increase. Etc. So if price goes up does it give you any information? All you know is it went up for some reason. You don’t know why. The people who bid it up know why they bid it up, based on their own assessment and knowledge–which is of necessity information they have that they did not get from prices; it is their knowledge and opinions that they use to form the price, not the other way around.

In fact it’s important to realize, in my view, that it is not a bad thing that information is “dispersed.” In fact, as Salerno points out (pp. 114-15), “dispersed knowledge is not a bane but a boon to the human race; without it, there would be no scope for the intellectual division of labor, and social cooperation under division of labor would consequently, prove impossible.”

Prices are important because they serve as an “accessory of appraisement.” “Current” (immediate past) prices tell only what the current price structure is, and thus serve as a basis for forecasting what the future array of prices will be, given the current starting point. For this reason, Hülsmann argues (p. 47) that present prices “can have no communicative function because they are only the, if indispensable, starting point for our understanding of the future.”

Some of Mises’s writing is extremely useful here, on the formation of prices and the distinction between future and past prices. See, e.g., Human Action, pp. 336-37:

In drafting their plans the entrepreneurs look first at the prices of the immediate past which are mistakenly called present prices. Of course, the entrepreneurs never make these prices enter into their calculations without paying regard to anticipated changes. The prices of the immediate past are for them only the starting point of deliberations leading to forecasts of future prices. The prices of the past do not influence the determination of future prices. It is, on the contrary, the anticipation of future prices of the products that determines the state of prices of the complementary factors of production. The determination of prices has, as far as the mutual exchange ratios between various commodities are concerned, no direct causal relation whatever with the prices of the past. The allocation of the nonconvertible factors of production among the various branches of production and the amount of capital goods available for future production are historical magnitudes; in this regard the past is instrumental in shaping the course of future production and in affecting the prices of the future. But directly the prices of the factors of production are determined exclusively by the anticipation of future prices of the products. The fact that yesterday people valued and appraised commodities in a different way is irrelevant. The consumers do not care about the investments made with regard to past market conditions and do not bother about the vested interests of entrepreneurs, capitalists, landowners, and workers, who may be hurt by changes in the structure of prices. Such sentiments play no role in the formation of prices. (It is precisely the fact that the market does not respect vested interests that makes the people concerned ask for government interference.) The prices of the past are for the entrepreneur, the shaper of future production, merely a mental tool. The entrepreneurs do not construct afresh every day a radically new structure of prices or allocate anew the factors of production to the various branches of industry. They merely transform what the past has transmitted in better adapting it to the altered conditions. How much of the previous conditions they preserve and how much they change depends on the extent to which the data have changed.

The economic process is a continuous interplay of production and consumption. Today’s activities are linked with those of the past through the technological knowledge at hand, the amount and the quality of the capital goods among various individuals. They are linked with the future through the very essence of human action; action is always directed toward the improvement of future conditions. In order to see his way in the unknown and uncertain future man has within his reach only two aids: experience of past events and his faculty of understanding. Knowledge about past prices is a part of this experience and at the same time the starting point of understanding the future.

If the memory of all prices of the past were to fade away, the pricing process would become more troublesome, but not impossible as far as the mutual exchange ratios between various commodities are concerned. It would be harder for the entrepreneurs to adjust production to the demand of the public, but it could be done nonetheless. It would be necessary for them to assemble anew all the data they need as the basis of their operations. They would not avoid mistakes which they now evade on account of experience at their disposal. Price fluctuations would be more violent at the beginning, factors of production would be wasted, want-satisfaction would be impaired. But finally, having paid dearly, people would again have acquired the experience needed for a smooth working of the market process.

[Update: See also, on this, Salerno, Ludwig von Mises as Social Rationalist, RAE, Vol. 4 (1990), p. 44.

But see Rothbard, in MES, ch. 4, §5.B, who seems to have a different take than Mises here, though he is not discussed or cited:

“In such a [barter] society, if all previous markets and knowledge of previous prices were somehow wiped out, there would, of course, be an initial period of confusion while each individual consulted his value scales and tried to estimate those of others, but there would be no great difficulty in speedily re-establishing the exchange markets. The case is different in a monetary economy. Since the marginal utility of the money commodity depends on previously existing money prices, a wiping out of existing markets and knowledge of money prices would render impossible the direct re-establishment of a money economy. The economy would be wrecked and thrown back into a highly primitive state of barter, after which a money economy could only slowly be re-established as it had been before.”]

For some other interesting views on this:

• Rothbard (p. 66): “the entire Hayekian emphasis on ‘knowledge’ is misplaced and misconceived”
• Hülsmann (p. 39): discussing “the irrelevance of knowledge problems”
• Salerno (p. 44): “[t]he price system is not–and praxeologically cannot be–a mechanism for economizing and communicating the knowledge relevant to production plans. The realized prices of history are an accessory of appraisement”
• Hoppe (p. 146): “Hayek’s contribution to the socialism debate must be thrown out as false, confusing, and irrelevant.”
• Kinsella: “The encoding metaphor seems to be a pseudoscientific and scientistic attempt to give this kind of economic theorizing a patina of scientific respectability by borrowing engineering terminology. It is scientistic because, in vainly trying to borrow natural science terminology, there is an assumption that only the “hard” or natural sciences have true validity. It is akin to using such inapt phrases as the “momentum” of the leading team in a basketball game, the “energy” of crystals and astral forms, or, even worse, “revving the engine” of the economy. Both economics and ethics can be sciences, but not in the same way as the causal, natural sciences.”

Update: See also these comments from Matt Machaj:

The above analysis brings us to the conclusion that socialism is (contrary to the thesis of many modern Austrians) neither a calculation, nor a knowledge problem. It has nothing to do with numbers (calculation is based on them), and it has nothing to do with the states of mind and gathered data of any kind (knowledge). It has to do with actions of private property owners and the results of the entrepreneurial process of imputation.

…

Prices certainly are “signals.” But they are not the signals transmitting knowledge—they are the signals of exclusion by other property owners. When some factor is priced, it means that one is excluded from using it, unless he is able to meet the condition of paying the asking price. In socialism, the central planner setting the price sets the exclusion or price for himself by himself. From this stems the absurdity of market socialism.

Prices also are not information signals about the immediate past. When an entrepreneur buys the factors today, looking at current prices, it is not the case that prices manifest only the past conditions. Today’s prices manifest current conditions—present expectations of the private property owners—and they include current anticipations of future conditions. Therefore, when an entrepreneur looks at today’s prices, he sees not only the immediate past, but conflicting visions about the future between different private property owners. So the prices do inform us about important conditions, and their function is not just to register the past.

Archived version with comments:

• Person

  Stupid question: what about the supplement of the prices in futures markets? Combining those with spot prices does seem to convey the information about whether the cause of the price shifts are transient or more fundamental and long-term. This is useful information for the entrepreneur. And if he buys a futures contract, he can’t “guess wrong” because he gets paid damages if those goods aren’t delivered. This fact coincides with the emphasis in Hayek’s work on, not action coordination, but pattern coordination. While these prices can’t tell one the optimum action given the values of other agents, they can specify, in general terms, what must be done to better coordinate with the plans of other agents.

  Bonus point to the first person call this question stupid.

  Published: July 12, 2006 12:15 AM

• Paul Edwards

  Person,

  You’re the first person to call your question stupid. 🙂

  Published: July 12, 2006 12:33 AM

• cynical

  Person,

  I would say that, too, is an interesting tangent, ala the Hayekian tradition. I don’t, however, think that it takes away from the correctness of the Rothbard/Salerno/Hoppe position on what Mises meant by “the calculation problem”.

  Published: July 12, 2006 2:02 AM

• zuzu

  Although I could say this about myself as much as anyone else, I think Mises requires significantly multiple readings of his voluminous output before lumping him and Rothbard together so quickly.

  (Yes, Rothbard was Mises protege or whatever, but Rothbard took a much more modernish social-science flavor to his descriptions and analysis. I do agree with lumping Rothbard and Hoppe together for this reason.)

  From an interdisciplinary perspective, Mises’ “human action” methodology derives from what many would now, with the gift of hindsight, call a branch of cognitive science.

  Likewise from that same perspective, Hayek’s analysis follows more of a graph and information theory methodology.

  Much of the same ground has been tread in swarms, complexity, combinatorial optimization, network theory, immunological computation, biosemiotics, and other such sub-disciplines studied at the Santa Fe Institute.

  Although both Mises and Hayek concerned themselves with epistemology, the distinction I would make between Mises and Hayek has to do with scope… as in, beware the fallacies of composition and division. Mises addressed how the individual gathers information and acts on that information. Whereas Hayek addressed how the distributed computation between actors provides more useful information faster to each node than a centralized computation network could.

  Published: July 12, 2006 7:17 PM

• Dennis Sperduto

  Below are a few observations regarding this informative posting, which is a good summary of an important issue.

  Despite the views of some others, I believe it is clear from reading pages 695-715 (1966 edition) of Human Action that Mises himself considered the calculation problem, and not the knowledge issue, to be socialism’s fundamental (and insurmountable) obstacle. In this discussion, Mises went to considerable length to differentiate the two issues, even conceding for argument purposes only, that the socialist planners did have full knowledge of all relevant data.

  The differences Mises and Hayek regarding economic calculation in a socialist state may arise from a more fundamental difference: Specifically, the epistemological status of the general equilibrium framework. Mises firmly and consistently rejected the general equilibrium construct except as a mental tool for understanding the incessant change of the real world, while some have argued that Hayek can be viewed as a “proximal equilibrium�? theorist. Interestingly, Hayek’s emphasis on the knowledge issue in the socialist calculation debate was a response to the proposals of the mathematical, general equilibrium socialists.

  Finally, I believe that Professor Herbener’s article “Calculation and the Question of Arithmetic�? [RAE Vol. 9, No. 1 (1996)] should also be considered a significant contribution to the literature. In particular, he discusses an important aspect of Mises’s argument, specifically, that the cardinal numbers that constitute money prices enable entrepreneurs to appraise, using a common denominator, disparate factors of production. Without the common denominator of money prices, appraisement and calculation are not possible. While some may consider this a trivial insight, it is nonetheless powerful and critical to Mises’s argument that denies the possibility of economic calculation in a socialist state.

  Published: July 12, 2006 8:01 PM

• Stephan Kinsella

  Dennis: quite right about Herbener’s seminal piece; I have added it to the main entry.

  Zuzu: I confess to not being quite clear as to what you are trying to say. In any event, you have no basis for concluding that I lumped in Mises and Rothbard “quickly”.

  Published: July 12, 2006 8:30 PM

• Dennis Sperduto

  One other point, related to the following quote of Steve Horowitz: “The ‘de-homogenizers’ have … correctly identified microfoundations [including] the importance of monetary calculation and Mises’s concept of ‘appraisement,’ but … they ignore what seems to be the obvious relationship between those microfoundations and Hayek’s vision of the social order. That is, they ignore that the outcome of the use of economic calculation by individual entrepreneurial actors and by firms and households is precisely the ‘use of knowledge in society’ that characterizes the Hayekian spontaneous market order.”

  The reason I believe that Misesians “ignore” the Hayekian concept of the spontaneous market order is that for Mises the key societal concept is human cooperation under the division of labor. Society, the “social order”, is most accurately characterized, if not defined, as interpersonal cooperation under the division of labor. In Mises view, material economic progress and the advancement of civilization are the result of expanding and deepening the division of labor.

  Published: July 13, 2006 10:58 AM

• Curt Howland

  Person, it may be true that the buyer of a futures contract cannot be making an “error”, but the seller certainly can. The futures contract does communicate information, but it is the information of the educated guess.

  Peoples preferences can still fluctuate in the ensuing time period, which is why “futures” were created in the first place, to dampen such changes’ effects on prices.

   

  Published: July 14, 2006 9:27 AM