The knowledge problem
Editorial · Hayek 1945 / dispersed & tacit knowledge / prices as signals
Hayek's 1945 refinement: even with perfect intentions and unlimited computing power, planners cannot obtain the knowledge they would need, because it is dispersed, local, tacit and constantly changing.
The claim
Friedrich Hayek's 1945 argument sharpens the calculation problem into something computing power cannot touch. Grant the planners perfect motives and unlimited processing speed: they still cannot plan well, because they cannot get the knowledge they would need. It does not exist in any single place to be collected.
- The economic problem is not allocating given resources. The textbook problem — distribute known means among known ends — is the easy part. The decisive knowledge is never "given" to one mind; it exists only as dispersed fragments held by millions.
- Much decisive knowledge is local and tacit. It concerns "the particular circumstances of time and place" — the half-idle machine, the supplier about to fail, the customer's shifting taste — and much of it, as Michael Polanyi put it, is knowledge we "know more than we can tell." It never reaches a central office because it cannot be written down.
- A price is a communication device. A market price compresses the knowledge of millions into a single number. When tin grows scarce its price rises, and everyone who uses tin economizes — without needing to know why it grew scarce.
- The system works because people can act on prices they don't understand. Coordination comes not from anyone grasping the whole, but from each responding to local signals — a feat no board can reproduce, because it would first have to possess what the price system instead economizes on knowing.
- This shifts the critique from computation to epistemology. Mises said planners lack prices; Hayek says they lack the knowledge prices are made of. More computing power does not help, because the inputs were never available to compute on.
Hayek made the argument in "The Use of Knowledge in Society" (1945) and developed it in "Competition as a Discovery Procedure" (1968): competition matters precisely because it generates knowledge no one had beforehand.
The scholarly debate
Two honest questions remain. First, how much practical knowledge genuinely resists codification — platforms and sensors now capture "circumstances of time and place" at a scale Hayek could not have imagined, which may narrow the gap or merely relocate it. Second, whether the argument proves too much: if dispersed knowledge defeats planning, how do enormous firms coordinate vast internal operations without internal markets? (See the counterargument.)
Strongest counterargument
◆ Against this entry’s own thesis
The sharpest reply is Ronald Coase's: if the price system were always the superior coordinator, firms would never exist — yet huge firms plan enormous internal operations by command, not internal markets, precisely because *using* the market has its own costs. That suggests the line between planning and markets is drawn by transaction costs, not an epistemological wall, and that the real question is empirical — how large a domain internal planning can efficiently coordinate — not whether planning is impossible in principle. The Hayekian rejoinder: a firm still swims in an ocean of market prices it did not generate; the knowledge problem bites for the economy as a whole, which has no outside prices to borrow.