Why Soviet shoppers stood in line, and the calculation problem behind it

The Soviet Union built rockets that reached orbit, a nuclear arsenal that frightened the world and a chess school that dominated a generation. It could not reliably supply its citizens with shoes that fit. That is not a joke. It is the central fact of the twentieth century's biggest economic experiment, and the explanation is one of the most important ideas in the social sciences.

The usual story blames the people involved: corrupt officials, lazy workers, an ideology that went wrong. Those things were real. But they are not the root. The root is a problem that would have defeated the cleverest, most honest, most hardworking planners in history. It is called the economic calculation problem, and it explains why a country can master physics and still stand in line for bread.

The claim, and the terms

Here is the thesis. Central planning fails not because planners are stupid or wicked, but because without market prices, produced by private ownership and voluntary exchange, they cannot know what anything is worth, and therefore cannot know whether any use of resources is an improvement or a waste.

Define the terms. A market price is the rate at which people actually trade a good, formed by many buyers and sellers each acting on what they know and want. Central planning is the system in which a government agency decides what is produced, how, and in what quantities, and sets prices by decree. Economic calculation is the process of comparing the costs and benefits of alternative uses of scarce resources, in a common unit, so that a decision-maker can tell which use is better.

The argument has four premises.

One. Resources are scarce and have many possible uses, so every economy must decide which uses to pursue.

Two. To decide rationally, someone must compare the value of what is given up with the value of what is gained.

Three. In a market economy, prices that arise from voluntary exchange between private owners of resources carry that information. A high price of copper tells every user of copper to economize and every producer to produce more, without anyone needing to understand why.

Four. In a system where the state owns all the means of production, there is no exchange between owners of capital goods, and so there are no genuine prices for them. Without prices, the planners are, in effect, working in the dark.

If the premises hold, the conclusion follows: a fully planned economy cannot calculate, and will waste.

Mises, 1920

The argument was put forward in 1920 by the Austrian economist Ludwig von Mises, in an essay translated into English as "Economic Calculation in the Socialist Commonwealth." The date matters. The Bolsheviks had seized power in 1917, and during the civil war and its aftermath they attempted to replace markets with forced requisitions and direct orders, a period known as War Communism. The results were catastrophic, and in 1921 Lenin retreated into the New Economic Policy, which allowed private trade and small business to return. Mises wrote at the moment when the idea of abolishing markets was fashionable and untested.

His point was not that socialism would be unpleasant or immoral, though he thought some of that too. It was that it could not work as an economic system once it had abolished the institutions that make calculation possible. A socialist manager could be told to build a railway, but he could not know whether the railway was worth building compared with a canal, a road, or nothing at all, because the prices of steel, labour, land and time would not reflect real scarcities.

Soviet history provided a long test of this claim. Seventy years of it.

Hayek, 1945: the knowledge problem

A quarter century later, Friedrich Hayek sharpened the point. His 1945 article in the American Economic Review, "The Use of Knowledge in Society," is one of the most cited essays in economics. Hayek's insight was that the problem is not only that planners lack prices; it is that the knowledge an economy needs is scattered among millions of people, much of it local, temporary and impossible to write down. The shopkeeper knows which items are moving. The foreman knows which machine is about to fail. The farmer knows how this year's rain has affected this particular field.

No committee can collect all that. It is not that the information is hidden. It is that most of it is not even available in a form that could be reported, because the people who hold it have not yet needed to articulate it. The price system, Hayek argued, solves the problem by condensing that dispersed knowledge into a signal everyone can act on. When the price of tin rises, you do not need to know whether a mine has flooded or demand has surged. You only need to know that you should use less tin.

A planner who replaces that signal with a decision has not made the economy rational. He has deleted its nervous system.

How Gosplan actually worked

The Soviet planning agency, Gosplan, was established in 1921 and became the centre of the system when Stalin launched the first Five-Year Plan in 1928. The method was to set output targets, quotas, for factories and farms, and to set prices by decree.

What does that produce, in practice? The first thing is that quotas are expressed in what can be counted. A shoe factory told to produce a certain number of shoes will produce the cheapest shoes it can in the quantity required, whether or not anyone wants them. A factory told to meet a quota by weight will make heavy products. A chandelier factory, the old joke went, would produce chandeliers so heavy that they pulled down ceilings. I offer that as an illustration of the incentive, not as a documented case; the mechanism is the point.

The second thing is that prices set by committee do not move with scarcity. If the official price of a good is below the price that would clear the market, the quantity demanded exceeds the quantity supplied, and the good vanishes from the shelves or is rationed by waiting. That is the queue. In a market, a price rises until supply and demand meet. Where the price is fixed, the adjustment happens in time: you pay with hours of your life. The Soviet shopper's most precious currency was time, and she spent a fortune of it.

The third thing is that the planners could not know what the right quantities were, so they used the previous year's output as a base and added a percentage. That meant continuing to make what had been made, whether or not anyone needed it, while the things people actually needed fell short. The result was chronic shortage alongside piles of unwanted goods. The Soviet economy, in the language of its own commentators, was an economy of shortage, in which scarcity was not an accident but a built-in feature.

The fourth thing is human. When goods are scarce, those with connections get them first. A system that promised equality produced a world of special shops for officials, favours, informal exchange and a black market, because people will always find a way to trade when the official system fails them. Equality of outcome was promised. What arrived was a hierarchy of access.

The 1962 Novocherkassk protests

The picture of a docile population is false, and the best proof is Novocherkassk. In June 1962, the Soviet government announced increases in the retail prices of meat and butter, at the same moment as workers at a locomotive plant in the city of Novocherkassk saw their pay rates cut. The two decisions together touched off strikes and protests on June 1 and 2. Troops were sent in, and on June 2 soldiers fired on the crowd. Officials later put the death toll at a little over twenty, and the event was suppressed from public knowledge for decades.

Notice what the episode shows. The price increases were not the whim of a cruel official. They were a response to the planners' difficulty in balancing supply and demand at fixed prices, with agricultural problems and subsidized food prices that did not cover costs. The state attempted to repair the imbalance by moving a price, and the result was an uprising that it answered with bullets. That is the nature of a system in which prices are political decisions: every price change is a political crisis.

The strongest objection: computers can solve it

Now to steelman the other side. In recent years, some argue that the calculation problem was a problem of the technology of the 1920s and 1930s, and that the situation has changed. A modern company like Amazon or Walmart plans the movement of an enormous number of goods internally, without market prices inside the firm. With powerful computers, big data and machine learning, the argument says, a central authority could gather all the relevant information and compute an efficient allocation. Soviet scientists themselves explored this: the mathematician Leonid Kantorovich developed linear programming for optimal allocation and shared the 1975 Nobel Prize in economics, and Soviet cyberneticists in the 1960s proposed a nationwide computer network to manage the economy. Today, the argument goes, we could finish the job.

It is a serious argument, and here is the answer, point by point.

One. A firm plans inside a market. Amazon and Walmart rely on prices at every edge: what they pay suppliers, what they charge customers, what wages and rents are in the market. Their internal planning is made possible by external prices that tell them what things are worth. Remove the market around them and the planning loses its anchor. A firm is an island of planning in an ocean of prices, and it cannot be scaled up to be the ocean.

Two. The data is not just hard to collect; much of it does not exist until people act. The relevant knowledge is not a pile of facts sitting somewhere waiting to be gathered. People discover what they want through choosing among actual alternatives, and entrepreneurs discover what is possible by trying. A person who has never tasted a food cannot report a preference for it. A computer can optimize with known data; it cannot know what has not yet been revealed by choice.

Three. Incentives remain. Even if an optimal allocation could be computed, someone must carry it out, and a planner with no stake in the outcome has little reason to report accurately or to act on the plan. The managers of Soviet enterprises had every reason to understate their capacity so that quotas would be easier to meet. A computer fed false inputs gives false outputs.

Four. Innovation is the problem, not just allocation. An economy is not a static puzzle to be solved once. It changes constantly as new goods and methods arise. A plan fixes the future in advance, and a market discovers it. The reason a smartphone exists is not that a planner decided it should, but that thousands of competing people were free to try.

Five. The test has been run, repeatedly. If the knowledge problem is merely a data problem, the nations with the most capable technology should have been the ones who prospered by planning. Instead, the planned economies fell behind the market economies on living standards, and the gap was visible in a simple comparison. Two Germanys and two Koreas, the natural experiments that socialism lost sets out two cases where the same people were split by a border and given two systems, and the results speak for themselves.

What the queue teaches

The queue is a monument to a mistake. It is what a price looks like when it has been turned into a decree. The Soviet citizen paid for her bread in time, in connections, in favours, in humiliation, rather than in rubles alone. That cost never appeared in any statistic, but she bore it.

None of this means that markets are perfect, or that government has no role. Conservatives do not claim that. It means that when a government sets out to replace prices with commands across a whole economy, it has taken on a task no human institution can perform, and it will fail in the same ways whatever the intentions of its leaders. The right question for any policy that interferes with price, whether a rent ceiling, a wage floor or a price cap, is what information the price was carrying and where that information will now go. For the small-scale version of the same story, Rent control's record from Stockholm to San Francisco is the same story is a good place to start, and profit is a signal, not a sin, and what a loss tells a company shows how the signals work in an ordinary firm. For the school of thought that identified the problem, see The Austrian school in plain words, from Carl Menger to Friedrich Hayek.

Measure it against the Founders

The Founders did not know the phrase "calculation problem," but they knew its moral cousin. They distrusted concentrated power because no ruler could know enough or be good enough to run other men's lives. James Madison, in Federalist No. 10, wrote of the "diversity in the faculties of men, from which the rights of property originate," and treated the protection of that diversity as the first object of government. A free economy is the economic form of that principle: it lets millions of people use their own knowledge for their own purposes, and it lets the results be sorted out by voluntary exchange rather than by a decree from a distant desk.

The great lesson of Novocherkassk is that when the state sets every price, the people have nowhere to appeal except the street. The great lesson of the Soviet queue is that no plan can replace the knowledge spread among a whole people. And the great lesson of the American system is that when government protects property, enforces contracts and keeps its hands off the price, it harnesses a computer far greater than any in a Moscow ministry: the free judgment of a free people.

The case in brief

1. Central planning fails because without market prices, produced by private ownership and voluntary exchange, planners cannot calculate whether any use of resources is an improvement, as Mises argued in 1920.
2. Hayek added in 1945 that the knowledge an economy needs is scattered, local and often unwritten, and that prices are how it is gathered and used.
3. Gosplan's quotas and fixed prices produced chronic shortages, the queue and a hierarchy of access, and the 1962 Novocherkassk protests showed that in a planned system every price change becomes a political crisis.
4. The claim that computers can now do the planning fails because firms depend on outside prices, much knowledge does not exist until people choose, incentives still distort data, and a plan cannot discover the new.
5. The Founders' insight, that no ruler can know enough to run other people's lives, is the same insight in political form.

What you can do: read Hayek's 1945 essay, which is short and available free online, then pick one local price control or licensing rule in your own city and ask your city council what information it is suppressing.