The standard account of a price is that it transmits information. Nobody in a large economy possesses more than a fragment of what would be required to allocate resources deliberately, and yet allocation occurs, and occurs well enough. The explanation offered is that a price aggregates the fragments. A rise in the price of a metal informs every purchaser of that metal that it has become scarcer relative to demand, without informing any of them why, and each adjusts his own use of it. The information required for the adjustment is contained entirely in the number.
The account is powerful and the illustration usually given is a genuine one. A seam runs out and a mine on the other side of the world shuts. The man who buys metal for a small workshop in another country hears nothing of this. He has never heard of the mine, and news of the closure would tell him nothing he could act on. What he sees is that the bar stock he has bought for years costs more than it did. So he uses thinner stock where thinner will do, he keeps the offcuts he used to throw away, and instead of the job that would eat most metal he takes on the job that eats least. A deliberate allocation would have required a central authority to discover the closure, estimate its magnitude, identify every affected use, and issue instructions. The price accomplishes the equivalent redistribution without the discovery, and the man in the workshop has done what a planner with full knowledge would have told him to do without ever learning why.
What the account does not establish, and is frequently taken to establish, is that the resulting distribution is desirable. A price aggregates willingness to pay, and willingness to pay is a compound of desire and of means. Two purchasers who want a commodity equally will transmit different signals if their resources differ, and the aggregation records the difference faithfully. A man of means who would rather like a thing and a man of none who needs it badly do not send signals of the same size, although the wanting may be. The number reaching the market carries no mark of which man it came from, because that is not the question the market answers. The mechanism is therefore an accurate instrument for the question it answers -- who will give up most for this -- and answers no other question at all.
The confusion is not merely verbal: a distributive conclusion drawn from an informational premise is an argument with a step missing. The distinction is easily lost because the two are habitually run together. An economist demonstrating informational efficiency is demonstrating something about coordination; a reader who concludes that the outcome is therefore just has moved to a different claim without noticing the transition. The demonstration supports nothing about justice, and it is not weakened by that limitation. It was never addressed to the question.
