The Economic Calculation Problem
A research lab on a deceptively simple question: can anyone, or any computer, coordinate an economy without market prices? Try it yourself, see why Mises and Hayek argued it cannot be done well, and weigh whether modern computing changes the answer.
A pencil has no single maker. If no one person knows how to produce even a pencil, how does an economy of billions of goods coordinate itself? And could a central planner do it instead?
Why this matters
This is not an abstract debate. The same question sits underneath central planning, central banking, central bank digital currencies, and the design of Bitcoin. Each is an answer to "who decides, and on what information?" Getting the question right is the difference between understanding those systems and just having opinions about them.
In 1920 Ludwig von Mises argued that a fully planned economy cannot calculate rationally, because without market prices for capital goods, planners have no way to compare the countless possible uses of resources. In 1945 Friedrich Hayek sharpened the point: the knowledge an economy needs is dispersed across millions of people and is often tacit, local, and changing. No central body can gather it in time. We will test that claim, not just assert it.
Try it: plan without prices
You are the planner. You have 100 workers to assign across three goods. People need all three, but you cannot see how much they need, because there is no market and no prices telling you. Make your best guess, then see how close you came.
Allocate 100 workers
Move the sliders. They do not have to sum to 100; the tool scales your mix to 100 workers when you run the plan. Then reveal what the people actually needed.
However you did, notice the core difficulty: you were guessing. A market does not guess. When bread runs short its price rises, which pulls workers toward baking and signals shoppers to economize, all without anyone issuing an order. The price is not arbitrary; it is compressed information about a scarcity only the buyers and sellers could know.
Prices as a signal: predict the shock
Hayek's claim was that prices move knowledge no planner could collect. Test your intuition on a concrete shock before reading the answer.
A drought cuts the wheat harvest in half. In an unhampered market, what happens first, and why is it useful?
How different systems answer the question
"Who sets prices, and where does the information come from?" separates these systems more cleanly than slogans do. The table is descriptive; each row has real tradeoffs.
| System | Who sets key prices? | Information source | Characteristic failure mode |
|---|---|---|---|
| Soviet central planning | Central planning board | Production targets and quotas | Chronic shortages and surpluses; no way to compare uses (see Kornai) |
| Modern central banking | Committee sets one key interest rate | Models and aggregate data | One rate cannot fit dispersed local conditions; boom and bust risk |
| CBDC | Central bank, programmable | State ledger | Control and surveillance by design; same calculation limits as planning |
| Bitcoin protocol | No one; issuance is a fixed rule, fees set by an open market | Open auction for block space | Volatile fees; coordinates money issuance only, not the wider economy |
Try it: set one interest rate for everyone
A central bank is a smaller version of the same problem. It sets one interest rate, the price of borrowing, for a whole country made of very different regions and sectors. Below, four parts of an economy each have a rate that would suit their local conditions, which you cannot see. Pick a single rate for all of them and see how well one number fits.
Set the national interest rate
One rate, four very different local economies. Set it, then reveal what each part actually needed. The wider the spread, the worse any single number does.
There is no rate that fits all four at once. Too low, and the overheating regions get cheap credit that fuels bubbles and malinvestment. Too high, and the struggling regions are starved of the borrowing they need. This is the calculation problem in miniature: one price, set centrally, cannot carry the dispersed and conflicting information that a thousand local prices would.
The strongest objection: can computers plan now?
The most serious challenge to Mises and Hayek is not that they were wrong in 1920, but that technology has changed since. This deserves its full strength.
The case that modern computing solves it
The economist Oskar Lange argued in the 1930s that planners could simulate a market by adjusting prices until shortages cleared (market socialism). Today the case is stronger: firms like Walmart and Amazon plan vast internal supply chains with enormous data and optimization, arguably out-planning small economies. Chile's Cybersyn project in the 1970s tried real-time economic coordination by computer. With modern machine learning and sensors, why could a planner not gather the dispersed knowledge Hayek described and compute the answer?
The response, and where it lands: three gaps remain. First, much knowledge is tacit and never written down, so it cannot be uploaded. Second, preferences are subjective and constantly changing, so the target moves faster than any model. Third, and most telling, large firms plan internally precisely because they buy their inputs at market prices set outside the firm; they ride on the very signal a whole-economy planner would have abolished. Walmart knows what a truck costs because there is a trucking market. Remove all external prices and the optimizer has no values to optimize against. That is the part computing has not closed. It is a real and live debate, not a settled rout in either direction.
Dispersed knowledge, from pencils to Bitcoin
Leonard Read's essay "I, Pencil" (1958) makes the point vividly: no single person knows how to make a pencil from raw materials, yet pencils are abundant and cheap. Graphite miners, cedar growers, machinists, shippers, and retailers coordinate through prices without any of them grasping the whole. The knowledge is in the system, not in any head.
Bitcoin is a smaller example of the same shape: coordination without a coordinator. Miners, node operators, wallet developers, exchanges, and self-custodians each act on local information and incentives, and the network holds together with no one in charge. Notice this is a claim about coordination design, not a claim that Bitcoin runs an economy. It coordinates one thing, the issuance and settlement of a money, by rule.
Try it: the fee market as a price signal
Here is a price signal working inside Bitcoin, with no planner setting it. Each block holds only so much space, so when more transactions are waiting than will fit, users bid fees for priority. The fee is a price for block space, and it rises and falls with demand exactly as Hayek would expect, coordinating who gets in next without anyone in charge.
Bid for block space
Set how much demand is waiting (in multiples of one block of space), then set your fee bid. The market clears at whatever fee fills the next block from the highest bidders. See whether you get in.
Notice no one decided the clearing fee. It emerged from everyone's bids against a fixed supply of space, and it told you, in one number, how urgently to bid or whether to wait. That is a market price doing exactly the job Mises and Hayek described, inside a system with no central authority to set it.
Claims to handle with care
"Bitcoin solved the economic calculation problem."
RealityThe calculation problem is about allocating an entire economy's resources without market prices. Bitcoin does not do that; it coordinates the issuance and settlement of money by a fixed rule. It is an example of decentralized coordination, not a solution to socialist planning. Conflating the two overstates the case.
"Central planning just needs smarter people and better technology."
RealityThis is the strongest objection, and it has a real answer (above): the binding constraints are tacit knowledge, changing subjective preferences, and the fact that internal planning still depends on external market prices. Better tools help within those limits; they have not removed them.
"Prices are arbitrary numbers set by sellers."
RealityIn a competitive market a price is the meeting point of countless buyers' and sellers' local knowledge and willingness. It is closer to a measurement than a decree, which is exactly why suppressing it (price controls) destroys information and tends to produce shortages.
💭 Reflection questions
No answer key. Argue them in both directions.
1. In the planning tool, what information would have made your allocation better? Where in a real economy does that information live, and who has it?
2. Make the strongest case that Walmart's internal planning disproves Hayek. Then explain why Walmart still needs outside markets to do it.
3. If a price is information, what exactly is lost when a government fixes a price by law? Can you name a case where the loss might be worth it?
4. A central bank sets one interest rate for a whole country. In what sense is that a smaller version of the calculation problem? In what sense is it different?
5. Bitcoin coordinates money issuance without a planner. What does it deliberately not try to coordinate, and why might that limit be a feature?
6. What experiment, even a thought experiment, would convince you that a computer could plan a whole economy as well as markets do?
Sources and claim ledger
Last reviewed June 17, 2026.
- The calculation problem: Ludwig von Mises, "Economic Calculation in the Socialist Commonwealth" (1920).
- The knowledge problem: F. A. Hayek, "The Use of Knowledge in Society," American Economic Review (1945).
- Market socialism (the planning-can-simulate-markets case): Oskar Lange, "On the Economic Theory of Socialism" (1936 to 1937).
- Shortage as a systemic feature of planned economies: János Kornai, Economics of Shortage (1980).
- Dispersed knowledge in production: Leonard Read, "I, Pencil" (1958).
- Real-time computer coordination attempt: Project Cybersyn, Chile (1971 to 1973), historical record.
- Price controls and shortages (illustrative modern evidence): Rebecca Diamond et al. on rent control, American Economic Review (2019).
- Bitcoin coordination by rule (issuance, fee market): Bitcoin protocol and Bitcoin white paper (2008).