Mises Showed in 1920 That Socialism Cannot Calculate a Single Price

Imagine you are put in charge of the whole economy. Congratulations. You have the best intentions, the smartest staff, a mandate from the people and a very large building. Your first assignment is small: build one railroad line between two towns. Should the bridges be steel or stone? Should the ties be oak or concrete? Should the line run around the mountain or through it?

You cannot answer. Not because you are stupid, and not because you are wicked, but because you have abolished the one tool that would have told you. That, in a paragraph, is the argument an Austrian economist named Ludwig von Mises published in 1920, and it is the most important argument against socialism ever made. It is not a moral complaint. It is a proof that the machine does not run. Let me walk through it.

The essay and the claim

In 1920 Mises published a long article in a German academic journal, the Archiv für Sozialwissenschaft und Sozialpolitik. Its title in English is Economic Calculation in the Socialist Commonwealth. He folded the argument into his larger book Die Gemeinwirtschaft in 1922, which appeared in English in 1936 as Socialism. The timing matters. The Bolsheviks had seized power in Russia in 1917, and socialist parties across central Europe were drafting plans to nationalize industry. Mises was not arguing about a distant theory. He was arguing about the next decade.

Define the terms first, because the whole case rests on them.

• Means of production: the things used to make other things. Land, mines, factories, machines, rail lines, raw materials.
• Socialism, in the sense Mises addressed: public, collective ownership of the means of production, with production directed by a central authority rather than by owners trading with one another.
• Economic calculation: the act of comparing the costs and benefits of alternative uses of resources in a common unit, so that you can tell whether a project adds value or destroys it.

Mises's claim was precise: under full socialism, rational economic calculation is impossible. Not difficult. Impossible.

The argument in five steps

Here is the reasoning laid out as premises, the way it deserves to be read.

1. Every production decision is a choice among alternatives. Steel used in a bridge cannot be used in a tractor. Labor spent laying track cannot be spent building houses. To choose well you must compare what you give up with what you get.
2. Those comparisons require a common unit. You cannot add tons of steel to hours of labor to acres of forest. Physical quantities do not sum. You need a single denominator, and in a market economy that denominator is money prices.
3. Prices for the means of production arise only from exchange. The price of a ton of steel or a lathe exists because owners of steel and lathes buy and sell them, each bidding according to what he expects to earn by using them. A price is the record of those competing bids.
4. Socialism abolishes that exchange. If the state owns every mine, mill and factory, there are no independent owners to trade them. The steel mill does not sell steel to the railroad; one ministry transfers it to another. A transfer between two departments of the same owner is not a market, and the number written on the transfer slip is not a price. It is a guess.
5. Therefore the planner has no way to know whether any project creates value or consumes it. He can measure physical output. He cannot measure economic cost. He is, in the phrase people later attached to the argument, flying blind.

Notice what Mises did not say. He did not say planners are corrupt, or lazy, or that people under socialism will not work. Grant all of that away. Assume angels run the ministry. Step four still holds, and the conclusion follows from it.

The railroad, worked through

A railroad is the classic way to illustrate the problem, and it is the right example because it is the kind of project planners love. It is big, visible and obviously useful. Let us make it concrete with a labelled hypothetical.

Picture a planning board choosing between two routes for a line. Route A goes around a mountain: it is 40 miles longer, uses more rail, more ties and more land, but needs no tunnel. Route B goes through: it is shorter but needs a tunnel that will consume a great deal of labor, explosives and specialized machinery for several years.

In a market economy, an engineer prices both routes. Rail costs so much per ton, land so much per acre, tunnel crews so much per day, machinery so much to rent, money so much in interest while the work is underway. Each of those prices carries information the engineer never had to collect himself. The price of steel tells him how badly shipbuilders, carmakers and builders across the country want the same steel. The interest rate tells him how much people value having things now rather than later. He adds the figures, compares the totals, and picks the cheaper route. If he guesses wrong, his firm loses money, and losses are the signal that resources were wasted.

Now give the same choice to the planning board. It knows Route A needs 40 more miles of rail. It knows Route B needs a certain number of worker-years and a certain tonnage of explosives. How does it compare 40 miles of rail with three years of a tunnel crew? It cannot ask what the rail would have fetched elsewhere, because nobody bids for rail. It cannot ask what the crew's time is worth, because the crew is assigned, not hired. It cannot ask what waiting three years costs, because there is no capital market setting an interest rate. Every number it writes down is administrative. It can pick a route. It cannot know whether it picked the better one, and it will never find out, because there is no profit and loss to tell it.

Multiply that one decision by every bolt, field and factory in a continental economy, and you have the problem.

The best reply: Oskar Lange

The argument started a fight that economists still call the socialist calculation debate, and the socialist side deserves its fairest statement. Its best champion was the Polish economist Oskar Lange, whose essay On the Economic Theory of Socialism appeared in 1936 and 1937.

Lange's reply ran roughly like this. Prices, he said, are just ratios at which goods trade, and a central planning board can set them. Let the board announce prices for the means of production. Instruct every state factory manager to follow two rules: produce where cost equals price, and choose the combination of inputs that minimizes cost at those prices. Then watch the warehouses. If steel piles up, the price was too high; lower it. If steel runs short, the price was too low; raise it. By trial and error the board can mimic the market and reach the same efficient result, while still owning everything and distributing income fairly. Lange even suggested that the board could do this better than real markets, because it could avoid monopoly and the swings of the business cycle.

That is a serious argument made by a serious man, and it is how most economists in the 1930s and 1940s thought the debate had ended. Socialism, they concluded, was possible in principle.

Why Lange's answer fails

It fails on four counts, and the fourth is the one that decides it.

One: the board cannot see what it is supposed to adjust. Shortage and surplus are visible in a few standard goods. They are invisible in millions of specialized ones: a particular grade of bearing, a spare part for one model of pump, a type of glass used by three factories. Lange's method needs the board to watch and reprice every one of them, continuously. A real economy has more distinct goods than any office could list.

Two: the managers have no reason to follow the rules. Lange told managers to minimize costs. Why would they? They do not keep the savings. They are judged by meeting quotas. Under that incentive the sensible manager hoards inputs, understates his capacity so his targets stay easy, and reports whatever keeps the board off his back. The information the board relies on comes from people with every reason to shade it.

Three: the knowledge does not exist in one place. Friedrich Hayek, Mises's younger colleague, made this point in a 1945 article, The Use of Knowledge in Society. Much of the knowledge that matters is local and fleeting: that a warehouse has spare space this week, that a supplier is about to fail, that one machine runs better on a different alloy. It lives in the heads of millions of people and changes daily. Prices let people act on it without anyone collecting it. A board cannot collect it at all.

Four: Lange's prices are still not prices. This is the decisive point, and it is Mises's original insight restated. A market price emerges because independent owners risk their own property bidding against one another. An entrepreneur who thinks rail is underpriced buys it and bets his capital on that judgment; if he is right he profits, if he is wrong he loses, and either way the price moves. Under Lange's scheme nobody owns anything, so nobody can bid, so nobody bears the loss of a bad guess. The board can raise and lower numbers forever. Those numbers are the board's opinions, not society's revealed valuations. Lange answered the problem of calculating with prices by assuming the prices into existence.

Lange himself, in a paper published after his death in 1967, described the market as a computing device of the pre-electronic age and suggested that electronic computers could now do its work faster. That only restates the hope. The difficulty was never arithmetic. It was that the data the equations need are created only by the exchanges socialism forbids. There is more on that thread in Nobody Designed the English Language, and Hayek Called That Spontaneous Order.

The Soviet test

Theory is one thing. The twentieth century ran the experiment, at a cost no one should want repeated.

The Soviet Union built its planning apparatus around Gosplan, the state planning committee, which set output targets in physical units: tons of steel, meters of cloth, numbers of tractors. Mises's argument predicts what such a system will do. It will produce the measurable quantity rather than the useful good, and it will produce chronic surpluses of things nobody wants alongside chronic shortages of things everybody needs. That is what happened.

The famous illustrations, told in the Soviet Union's own satirical press, are about targets set by weight or count. Set a nail factory's target by weight and it makes a few enormous nails. Set it by number and it makes millions of tiny ones. Whether every particular story is literally true matters less than the mechanism, which is exactly what the calculation argument describes: with no prices to say what the nails were worth to the people using them, the only available measure was the physical one, and managers met it.

The queues are the larger proof. Soviet shoppers stood in lines for bread, meat, shoes and soap for decades, not because the country lacked land, labor or engineers, but because prices set by decree did not match what people wanted and could not adjust. The Hungarian economist János Kornai, who lived inside a planned economy, gave the pattern a name in his 1980 book Economics of Shortage: shortage was not an accident of bad planners but a normal, built-in feature of the system. We tell that part of the story in more detail in Why Soviet shoppers stood in line, and the calculation problem behind it.

Even socialist intellectuals came around. When the Soviet bloc collapsed, the economist Robert Heilbroner, a man of the left who had spent his career sympathetic to planning, wrote in The New Yorker that the debate was over and that Mises had been right. That is as close as the profession gets to a formal concession.

And the natural experiments keep running. East and West Germany started with the same language, the same culture and the same war damage, and ended with a wall built to keep one side's people from leaving. North and South Korea tell the same story with satellite photos of electric light. See Two Germanys and two Koreas, the natural experiments that socialism lost.

What it means for the policy debates in front of us

You will rarely meet anyone who proposes full Soviet planning. The calculation argument still applies, because it scales down. Every time government replaces a price with a decree, it recreates a small version of the problem.

• Price controls. Cap the rent on apartments, and the price no longer tells builders where housing is needed. The predictable result is fewer new units and worse upkeep of existing ones.
• Single-payer schemes. When one agency sets every reimbursement rate, the rates are administrative numbers, not prices. They will be wrong somewhere, and the error shows up as waiting lists or as services that quietly disappear.
• Industrial planning. When a government picks which industries or technologies get capital, it substitutes a committee's guess for investors' bids. Sometimes the guess is right. The system has no reliable way to know, and no automatic way to stop funding the guesses that were wrong.

None of this means every public program is futile. Roads get built and courts get run. It means that the more an economy's decisions are made by command rather than exchange, the less anyone can tell whether those decisions add value. The burden of proof belongs on the planner. The eight socialism myths tested against the countries that actually tried it post tests the most common modern claims against the record, and The Value of a Thing Is Not the Labor in It, and Menger Answered Marx covers the theory of value that Mises built on.

The Founding saw the moral half of this

The calculation argument is economics, but the Founders had grasped the moral principle beneath it more than a century earlier: that free people own and dispose of their property, and that the protection of property is one of government's first purposes. James Madison wrote in Federalist No. 10 that the diversity in the faculties of men, from which the rights of property originate, deserves protection, and that this protection is "the first object of government." The Fifth Amendment forbids taking private property for public use without just compensation.

The Founders built a system of dispersed ownership because they believed dispersed ownership protects liberty. Mises showed something they did not need to know: dispersed ownership is also what makes rational economic life possible at all. Liberty and prosperity turn out to rest on the same foundation. A government that owns everything can tell you what to do, and it cannot tell itself what is worth doing.

That is why the argument still matters. Socialism keeps returning with new branding: democratic, market, green, participatory. Ask each version one question. Who owns the means of production, and if no one can buy or sell them, where do the prices come from? If the answer is a board, you already know the ending. The satirical version of where that road leads is in A Thought Experiment, the Nation That Planned Breakfast From One Office.

The case in brief

1. Mises argued in 1920 that rational economic calculation requires money prices for the means of production, and those prices arise only when independent owners trade them.
2. Socialism abolishes private ownership of the means of production, so it abolishes those prices, and the planner cannot tell whether any project creates or destroys value.
3. Oskar Lange's reply, that a planning board could set prices by trial and error, assumes away the problem: numbers set by a board are opinions, not bids by owners who bear the loss of being wrong.
4. The Soviet economy's chronic shortages, measured-by-the-ton output and endless queues are what the argument predicted, and later socialist economists conceded the point.
5. The lesson scales down: every time a decree replaces a price, a smaller version of the same blindness follows.

What you can do: before you support any program that sets prices by law, ask your representative in writing what signal will tell the agency when its numbers are wrong, and look up their votes on the scorecard.