Contribution
It recasts the socialist-calculation debate as a question about knowledge, not computation. The “data” of the economic calculus “are never for the whole society ‘given’ to a single mind … and can never be so given.” Which arrangement is more efficient depends on which makes fuller use of knowledge that is initially dispersed. It names a kind of knowledge, of particular circumstances of time and place, that a central authority cannot hold in principle, and it reads the price system as a mechanism for communicating the part of that knowledge each decision needs.
Key insight
A decision maker needs to know how much scarcer something has become, not why. In the tin example: “It does not matter for our purpose—and it is very significant that it does not matter—which of these two causes has made tin more scarce.” A price condenses the significance of a change “in view of the whole means-end structure” into one number. Through chains of people whose “limited individual fields of vision sufficiently overlap”, the whole economy adjusts while almost nobody knows the cause. Decisions that depend on local knowledge can therefore be decentralized and still fit together.
Assumptions
- The economic problem is “mainly one of rapid adaptation to changes in the particular circumstances of time and place”. “economic problems arise always and only in consequence of change.” If plans could be laid down in advance and kept, the case for central planning would be “much less formidable”.
- Two kinds of knowledge: scientific knowledge of general rules, which a body of experts might command, and knowledge of particular circumstances, which “by its nature cannot enter into statistics and therefore cannot be conveyed to any central authority in statistical form”.
- Statistics are made “by lumping together, as resources of one kind, items which differ as regards location, quality, and other particulars”.
- The stability of aggregates is not the law of large numbers: “The number of elements with which we have to deal is not large enough for such accidental forces to produce stability.” It is kept by “constant deliberate adjustments”.
- No model is given. The argument is conceptual, with examples (the shipper of tramp-steamers, the estate agent, the arbitrageur, the tin shortage).
Key results
- §I. The optimum problem, given all information, preferences and means, is “purely one of logic”, solved by equal marginal rates of substitution. “This, however, is emphatically not the economic problem which society faces.”
- §II. All economic activity is planning. The dispute is over who plans: “centrally, by one authority for the whole economic system, or … divided among many individuals”. “The halfway house between the two … is the delegation of planning to organized industries, or, in other words, monopoly.” Efficiency turns on whether it is easier to bring dispersed knowledge to a central authority or to convey to individuals “such additional knowledge as they need in order to enable them to fit their plans with those of others.”
- §§III–IV. Knowledge of particular circumstances gives “practically every individual … some advantage over all others”, usable “only if the decisions depending on it are left to him or are made with his active coöperation”. Central planning on statistical information “by its nature cannot take direct account of these circumstances”.
- §V. “We must solve it by some form of decentralization.” But the man on the spot still needs information from beyond his horizon, and needs only the relative importance of the things he deals with. Prices supply it, as “rates of equivalence” do for a single mind solving its own allocation.
- §VI. The price system is “a mechanism for communicating information”, whose most significant feature is “the economy of knowledge with which it operates”. It “fulfils [this] less perfectly as prices grow more rigid”. It is “not the product of human design”, and the problem is “how to extend the span of our utilization of resources beyond the span of the control of any one mind”. The same problem arises “with language and with most of our cultural inheritance”.
- §VII. Against Schumpeter’s claim that consumers’ valuation of goods “ipso facto” values the factors of production: implication “can be meaningfully asserted only of propositions simultaneously present to one and the same mind.” The task is to show “how a solution is produced by the interactions of people each of whom possesses only partial knowledge.”
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | The economic problem of a society is the use of knowledge that is not given to anyone in its totality, not the allocation of given resources | moderate | conceptual argument (§§I–II); the Schumpeter critique (§VII) |
| C2 | Knowledge of particular circumstances of time and place cannot be conveyed to a central authority in statistical form, because statistics lump together items that differ in ways that matter to the decision | moderate | argument from how statistics are made (§IV); no evidence |
| C3 | So decisions that depend on such knowledge must be decentralized to those who hold it | moderate | follows from C2 and the premise that adaptation to change is the main problem |
| C4 | A price system conveys to each decision maker the relative scarcities relevant to it, without the causes, and with great economy of knowledge | moderate | the tin example (§V); an analogy to rates of substitution in a single mind’s calculus |
| C5 | The price system works less perfectly as prices grow more rigid, though the same forces then act through other terms of contracts | weak | asserted in passing (§VI) |
| C6 | Delegating planning to organized industries is a third arrangement, monopoly, which “few like when they see it” | weak | asserted (§II) |
| C7 | The coordination problem is the central theoretical problem of social science, shared by language and cultural inheritance | weak | asserted, with Whitehead quoted (§VI) |
Concepts
- Planning: “the complex of interrelated decisions about the allocation of our available resources”. Everyone plans; the dispute is who.
- Knowledge of the particular circumstances of time and place: “unorganized knowledge which cannot possibly be called scientific in the sense of knowledge of general rules”, fleeting and local.
- The man on the spot: the person familiar with the circumstances, who “know[s] directly of the relevant changes and of the resources immediately available to meet them”.
- Economy of knowledge: “how little the individual participants need to know in order to be able to take the right action”.
- Datum: Hayek’s charge is that the word’s ambiguity, given to whom, sets “a trap” for economists who treat facts known to the observing economist as if they were known to the agents.
Connections
It continues the socialist-calculation debate begun by von Mises “twenty-five years ago”, answering Lange and Lerner, and it rejects the reading of Pareto and Barone as having solved the problem: they “merely … state[d] the conditions which a rational allocation of resources would have to satisfy.” Its foil is Schumpeter’s Capitalism, Socialism, and Democracy. Its mechanism, prices as low-dimensional messages, was later formalized as the informational efficiency of the competitive mechanism (Hurwicz; Mount and Reiter). The record holds none of those works.
Bearing on the record
- THEORY-841. Filed Proposed: C2–C4 as an account, without a model or evidence behind it.
- CLAIM-410. A18 cited Hayek for “the informational advantages of certain governance arrangements”. The essay supports exactly that and no more. It is an argument for where decisions should sit, given where knowledge sits, and §II already frames the question as who plans, not whether. That is the claim’s “profile, not a scale”. Its “halfway house”, planning delegated to organized industries, is a case of the claim’s transfer: authority moves from a central board to an industry, not to the members. Hayek does not argue that removing a public constraint decentralizes. He argues for competition, which he calls “decentralized planning by many separate persons”.
- Legibility. §IV’s account of why statistics cannot carry circumstances of time and place, “lumping together, as resources of one kind, items which differ as regards location, quality, and other particulars”, is the cost of legibility in TERM-064’s sense, stated in 1945. NOTE-649 already reads LIT-848 as compressing Hayek and Scott together.
- NOTE-649 cites the essay through LIT-848. Its gloss, “Hayek’s price as a summary of dispersed knowledge”, is accurate.
- THEORY-429 (North). Hayek’s prices and North’s institutions are both answers to dispersed knowledge, one by messages and one by constraints. Neither cites the other here, and this is the record’s comparison.
No instruction for machine-learning practice; nothing for the anthology.
Limitations
- No model and no evidence. Every result is argued from examples and from the logic of choice. The claim that prices convey “the relevant information” is never stated precisely enough to test: what counts as relevant, and how close to right the resulting adjustment is, are left open. Hayek grants that the adjustments are “probably never ‘perfect’”.
- Prices only. The essay says nothing about what prices fail to carry: externalities, information held strategically, or knowledge that matters for decisions with no market. Its claim that the coordination problem is shared by language and culture is asserted, not shown.
- The comparison is two-sided. It argues that dispersed knowledge cannot be centralized. It does not show that decentralized decision makers get the knowledge they need in every case, only in the tin example.
- Reprint, not page images. The Econlib text was not compared with the original pages.
Open questions
- For which decisions is local knowledge decisive and for which is it not? Hayek grants that experts may command scientific knowledge. The essay gives no criterion for when a decision depends mainly on one kind.
- What does the argument imply for allocations made with no price, inside an organization or a state? The halfway house of §II is named and not analysed.