Prices as Signals: The Telegram from Everywhere

Recall the tin story. Somewhere on Earth, tin gets scarcer — a mine floods, or a new gadget starts devouring the stuff; it genuinely doesn’t matter which. The price of tin rises. And within days, millions of people across the planet — a solder manufacturer in Hamburg, a can factory in Ohio, a hobbyist in Osaka — quietly begin to economize on tin, substitute away from tin, and hunt for new sources of tin. Almost none of them ever learns why. They just saw the price and adjusted. The right global response, assembled from people who never got the memo, because the price was the memo.

Hayek looked at this and did not shrug. He called the price system a “system of telecommunications” and — in a working paper by a famously unexcitable economist — a “marvel.” The claim of this lesson is that he was being precise, not poetic. A price is a compressed telegram from the whole world, and once you see what’s packed inside that one number, you’ll never look at a price tag the same way again.

A price is a sufficient statistic

Here’s the fancy name for the tin trick. In statistics, a sufficient statistic is a summary number that contains everything relevant in the data for the decision you face — once you have it, seeing the raw data adds nothing. The claim: a market price is a sufficient statistic for global scarcity and desire. It compresses every flooded mine, every new gadget, every shift in taste, every discovered substitute, into the one number your decision actually depends on.

Think of your car’s fuel gauge. It shows one thing: how much fuel remains. It does not show you a chromatography report on the tank’s contents, the refinery’s maintenance schedule, or a history of your recent driving. And you don’t miss any of that, because for the decision at hand — do I stop for fuel? — the gauge is sufficient. Demanding the full report wouldn’t make you a better driver; it would make you a slower one. The price of tin is the world’s fuel gauge for tin. “It costs more now” isn’t a degraded version of the full story. For deciding what to do, it’s the whole story.

The market as a distributed computer

So where does this magical number come from? Nobody computes it. That’s the part that should feel familiar from your emergence course: a macro pattern (the posted price), produced by local rules (each person deciding buy/don’t-buy at the price they see), with no controller anywhere in the loop. Each participant performs one tiny, private computation — is this worth it to me? — and the network of trades aggregates millions of those private verdicts into a single public number. The market is a distributed computer whose processors are people and whose output is the price.

Let’s run the computer by hand, with an economy small enough to fit on a napkin. Three producers of a widget, three buyers, one unit each. Each knows only their own number — the producers their private cost to make one, the buyers their private value for owning one:

Try a price of $5: only Producer A is willing to sell (B and C would lose money), but X and Y both want to buy. Two buyers chasing one unit — the disappointed one offers a bit more, and the price gets bid up. Try $10: A and B will happily sell, but only X still finds it worth buying. Sellers with unsold stock shade their asks down. The haggling only stops in the corridor between $7 and $9 — say $8 — where exactly two units are supplied (A and B) and exactly two are demanded (X and Y). That’s the clearing price: the price at which quantity offered equals quantity wanted, so nobody is left waving money or holding stock.

Look at what the $8 verdict quietly accomplished. The two cheapest producers make the units; the two hungriest buyers get them. Producer C (cost $11) and Buyer Z (value $6) are correctly pushed out — making C’s unit for Z would burn $5 of value, and the price told them both “not you, not today” without a committee ever reviewing their case. Total surplus: X gains 13 − 8 = 5, Y gains 9 − 8 = 1, A gains 8 − 4 = 4, B gains 8 − 7 = 1. Eleven dollars of value created — and you can check that no other assignment of who-makes and who-gets beats it.

Now perturb the computer with one private fact and watch it re-run.

What prices do that reports can’t

“Fine,” says the skeptic, “information gets around. Why not collect it properly — surveys, reports, a really good dashboard?” Because a price has four properties that no filed report has ever had:

Fast. The price moves the moment the first trades do — no committee meeting, no quarterly data call, no waiting for the report to be typed, approved, and ignored. In the napkin economy, A’s flood re-priced the market in one round of haggling. A survey would still be in the mail.

Incentive-compatible. Here’s the tie to your incentives course: you respond to a price because it’s your money on the line. A report can instruct a factory to “economize on tin”; the price pays the factory to do it, in proportion to exactly how much it economizes. Nobody audits your compliance — your own profit-and-loss does.

Honest at the margin. People lie on forms. Ask factories how much tin they “need” and every one of them needs a lot, urgently (ask any budget office). But buying tin at $14 is a costly, credible act — you only do it if the tin really is worth more than $14 to you. Prices are testimony under oath, where the oath is your own bank account.

Carriers of tacit knowledge. This is the deepest one. Lesson 1’s baker cannot articulate her dough-feel — no form has a field for it. But her bids and asks embody it: her sense of tomorrow’s demand, of this flour’s quality, of which supplier slips in the rain, all get expressed in what she’s willing to pay and charge. The price system transmits knowledge that its owners could never write down. A reporting system can only collect what can be stated; a price collects what can only be acted on.

The socialist calculation debate

This is not an academic parlor game; it was the twentieth century’s biggest economic argument. In 1920, Ludwig von Mises threw down the gauntlet: socialism — meaning genuinely centralized production, with no market prices for capital goods — isn’t just hard, it’s impossible, because the planner cannot calculate. Should the new railway line use steel rails or the sturdier alloy? Should this factory make tractors or trucks? Without prices, you can list the physical inputs of each plan — so many tons, so many hours — but you cannot add them up. “Calculation in kind” can tell you a bridge uses less steel and more labor than a railway; it cannot tell you which one wastes more of what society values. The planner isn’t merely uninformed. He’s blind, doing arithmetic without numbers.

The market socialists fired back, and — intellectual honesty requires saying this clearly — their best shot was genuinely clever. Oskar Lange proposed in the 1930s that a socialist planning board could mimic the market: post trial prices, tell state factory managers to follow simple rules, watch where shortages and surpluses appear, and adjust the prices by trial and error — exactly what our napkin market did by haggling. On paper, Lange’s board computes the same clearing prices. Many economists at the time scored the debate for Lange.

Hayek’s 1945 essay is the counter-punch, and you already hold all of its pieces. The trial-and-error board fails on speed (real economies have millions of prices, each needing continuous adjustment as conditions change hourly — the board is always solving yesterday’s economy), on incentives (a state manager following rules with no skin in the game has every reason to overstate his needs and understate his capacity — no oath, no bank account), and on tacit knowledge (the baker’s dough-feel never reaches the board at all, because it can’t be filed). Lange’s scheme assumed the knowledge problem away; Hayek pointed out it was the entire problem.

History ran the experiment anyway. Gosplan, the Soviet planning agency, ended up administering on the order of millions of prices, set by committee and revised at a pace of years — stale telegrams from a world that had moved on. The results were the queues and the warehouses: chronic shortages of what people wanted standing right next to gluts of what they didn’t, the signature failure pattern of an economy whose signals are set by decree rather than discovered by trade.

Run the experiment yourself

Enough history — run the twentieth century on your screen. The economy below has about a dozen producers and a dozen consumers, each holding one private number (a cost or a value) that nobody else can see. You can hand the economy to a central planner — who, like Gosplan, sees only aggregates and decrees one price for everyone — or to the price mechanism, where the clearing price emerges from local trades. Then hit the local shock and watch a cluster of producers’ costs jump, tin-style. Watch the readouts: total welfare, unmet high-value need, and — the number this whole course is about — how many of the economy’s 24 private facts each regime actually puts to use.

Where the signal garbles

Before you leave this lesson a price mystic, one section of cold water. (A full tour of the failure modes gets its own lesson later; here’s the trailer.) The telegraph is a marvel, but it’s still a physical wire, and three things go wrong with it:

Some messages never get sent. Prices transmit only what trades can express. If a tin smelter poisons a river, the downstream village never gets to bill anyone — the pollution is an externality, a cost that shows up in no one’s private profit-and-loss, so it’s simply missing from the wire. The price of tin telegraphs the flooded mine perfectly and says nothing at all about the poisoned river. Silence on the wire is not evidence of nothing to say.

Some wires get cut. Remember price ceilings and floors from your supply and demand course? Reframe them now: a price control is signal jamming. Cap the price of tin below its clearing level and the telegram “use less, make more” is never transmitted — buyers keep demanding as if tin were plentiful, producers supply as if it weren’t wanted, and the gap becomes a queue. The control doesn’t repeal the scarcity; it just shoots the messenger.

Some messages are noise. Prices can carry error, too. In a bubble, buyers bid the price up because the price is going up — the signal starts reporting on itself instead of on scarcity, a feedback loop sometimes called reflexivity. The telegram still arrives crisp and authoritative; it’s just wrong. A marvelously efficient transmitter will transmit nonsense with the same fidelity as truth.

When to reach for this model

Reach for prices as signals whenever you’re tempted to say “somebody should gather all the information and decide.” Ask first: does a price already carry that information — and would your report be slower, less honest, and blind to the tacit parts? It also cuts the other way: when a price looks insane, ask what telegram is this transmitting, and from where? — a spiking price is usually reporting a real scarcity somewhere upstream, and jamming the signal (capping it) leaves the scarcity intact but unannounced. And before trusting a price fully, run the garble checklist: any unpriced harms? any controls on the wire? any self-feeding feedback?

Recap

You now know what the number says. Next lesson we watch it build something: a plain wooden pencil — cedar, graphite, lacquer, a metal ferrule from three continents — assembled by thousands of people, not one of whom knows how to make a pencil, and none of whom needed to. Nobody knows how; everybody’s prices do. The invisible hand, up close.