The Economic Calculation Problem- Why Control Economies Fail

Economics, Long Reads
Reading Time: 8 minutes

A modern, diverse, industrialized economy can be expressed in little more than a maths problem- simultaneous equations with a billion equations and a billion unknowns. This is the problem presented to decision-makers when the process of ‘economic calculation,’ or the allocation of resources, the setting of prices and ultimately the determination of what is consumed by whom, is carried out on a daily basis.

The ‘Economic Calculation Problem’ was first described by Ludwig von Mises in his work Economic Calculation in the Socialist Commonwealth. Aimed as a criticism of centrally planned, Socialist economies, it outlines several reasons why central planners are inherently inferior at Economic Calculation, and why efficient calculation is impossible with communal ownership. Given a scenario in which various interdependent productive enterprises are centrally owned, due to a complete lack of competitive exchange and therefore a lack of prices which are set by individual actors every step along the line and accurately reflect the relative worth of production goods, the calculation of various crucial metrics such as profit margins is extremely difficult. If a producer wished to compare the resources necessary to produce 800 barrels of wine as opposed to 800 barrels of oil, a pricing system provides him with an easy mechanism by which relative cost is computed. When all goods are transferred internally without pricing, a manager can only make rough estimates on these figures.

In a competitive decentralised market, with many vendors and purchasers, prices vary until they reach an equilibrium which reflects the supply and demand of the goods, commodities, and labour being traded. On the one side, consumers translate their own subjective desires into an objective measurement- the prices they are willing to pay. Profitable ventures are those where the value on the market of the resources used for a process is lower than the value and return of the goods produced; as a result, individuals seeking profit take the initiative to invest capital into ventures that are efficient in this regard, usually through the mechanism of financial markets.

On the other side, producers set the lowest price at which they are willing to sell, translating the numerous factors determining the ‘cost’ of a process- availability of resources and the cost of production goods and labour- into another objective measurement. Not only does this theoretically prevent the over-use of a resource, but it incentivises consumers to and fund processes which are themselves efficient to carry out, through the mechanism of supporting that venture by providing it with profit. Mises thus argues that without accurate pricing it is impossible to determine the most efficient allocation of resources and the configuration of an economy with near-infinite permutations.

Socialist critics of central planning also note its inefficiencies. As argued by Trotsky in The Soviet Economy in Danger: ‘in reality, the bureaucracy errs frightfully in its estimate of its spiritual resources.’ A planning board completely lacks the sensory bandwidth possessed by millions of individual producers and consumers- Gosplan, the Soviet agency responsible for planning, regularly made balance sheets for nearly 20,000 different commodities. Its information is also of far lower quality and resolution because of what Hayek called the ‘local knowledge problem’- the fact that those who are best able to efficiently plan a certain area or venture economically are those who are ‘closest’ to it, that is to say, those who have the greatest amount of knowledge about it. In the vast majority of cases this is not a central planning board but the individuals and groups who operate in that sector daily, those who would usually carry out economic planning for that venture.

According to Hayek, who later developed Mises’ theory in his 1945 work The Use of Knowledge in Society, this vast network of individuals creates a cohesive web of productive forces because of the rapid and effective communication possible through prices. Rather than needing to compare the various factors that went into the production of two goods, a buyer can simply compare two prices. Even if a competent central planning board were able to process information and plan accordingly, the significant lag in decision-making would result in a vast amount of lost productivity. The Soviet Union practice of shturmovschina involved frantic overtime work near quota deadlines due to significant delays in the delivery of orders because of the time inefficiency of central planning. The result of this involved significant declines in productivity and output. 

Gosplan set prices and production at the beginning of each year- retailers could not change prices in response to fluctuations in demand on the ground. The massive, unnecessary shortages and surpluses that this created have been since extensively documented, particularly by Hungarian economist Janos Kornai in his 1980 book The Economics of Shortage. Mises also criticised the use of the Labour Theory of Value as a benchmark to set prices, because it failed to reflect the factors of technology, productivity and the availability of resources and commodities, which ultimately also determined the opportunity costs of producing certain goods compared to others.

Recent developments in data science and information technology, according to computer scientist Paul Cockshott and economist Allin Cottrell in their 1993 book Towards a New Socialism, will provide central planning with the means to overcome this problem. Moore’s law successfully observed for several decades that the numbers of transistors in a dense integrated circuit doubled every two years. Whilst from 2015 according to former CEO of Intel Brian Krzanich this has slowed to two and a half years, sustained growth in computer power seems inevitable.

A supercomputer, or a network of computers, could theoretically, given reasonably correct inputs and programming, manage an entire economy. Two notable examples of attempts to achieve this in the past- one by Soviet computer scientist Victor Glushkov in 1962 which ultimately lost the support of Soviet leaders possibly due to the threat it posed to their control, and another from 1971 in Chile, named ‘Project Cybersyn,’ which supposedly achieved a few limited successes but ultimately was cut short in 1973 by the military coup. As discussed by Leigh Phillips and Michael Rozworski in 2019 in their book The People’s Republic of Walmart, certain large organisations such as Walmart, the Pentagon and Amazon already use computer processing to handle large portions of operations.

However, it is unclear to what extent long-term economic decisions are actually left to the discretion of these machines. At the moment, the concept of a computer running an entire economy, a ‘Laplace’s demon’ if you will, is confined to science fiction, even appearing in Asimov’s I, Robot. Terminator’s Skynet deserves an honourable mention (and we all saw how that turned out!). Even if this were possible, the Economic Calculation Problem is not only a mere complexity issue that perfect analysis can solve. Mises and Hayek distinguish between technological knowledge, being the statistical knowledge of the various factors which affect production, and information which ‘by its nature cannot enter into statistics,’ the subjective desires and choices of consumers.

Hayek goes on to write that the categorization of consumer preferences into statistical data is extremely difficult without price signals, which act as a common basis for comparison – individuals may not realise the importance of the information they have or often have little incentive to transmit correct information about their preferences, if at all. Consumer preferences constantly change as conditions shift- without the periodic self-adjustment of markets, information gathering could not be reliable over extended periods and would, therefore, have to be a sustained and expensive but unreliable process. 

Salerno writes in defence of Mises in 1994: ‘The Miseian demonstration of the logical impossibility is not predicated on the central planners’ incapacity to perform tasks that can conceivably be carried out by individual human minds… Rather it is concerned with the lack of a genuinely competitive and social market process in which each and every kind of scarce resource receives an objective and quantitative price appraisal in terms of a common denominator reflecting its relative importance in serving (anticipated) consumer goods.’ Market processes naturally tend towards an equilibrium through an entrepreneurial process of competition which Mises writes ‘again and again reshuffles exchange ratios and allocation of the factors of production.’

Rival entrepreneurs and managers who have access to this subjective information can rationally allocate resources in a way any computers for the foreseeable future cannot. These data ‘are intimately tied to the institution of private property and the market process and do not come into existence in the absence of this process.’ Poor economic planning is punished by the loss of business, meaning that a market is self-correcting in a way that any government is not. A computer may theoretically have such perfect knowledge that it understands the thought processes of all individuals at once, but such a task is many orders of magnitude more difficult and therefore more unrealistic than computing problems at the moment only considered in the realm of science fiction. Even if such a computer were possible, fundamental issues arise when it comes to its programming. Presumably, its priorities, its goals and its framework for decision making would be set by the same board of central planning which we have determined is so inefficient. Would it truly be autonomous? Or would it merely be a digital copy of its predecessors?

Oskar Lange’s response to the Austrian duo in 1936 was that rational calculation using value terms was still possible without private property ownership and capital markets. The Lange model, which so far has never been implemented, presents a system in which the means of production are publicly owned but pricing mechanisms still exist, referred to by Lange as ‘market socialism.’ A central planning board (or supercomputer) allocates investment based on information provided by the choices of workers and consumers, who are allowed to choose the prices at which they sell their labour and pay for products, with a trial-and-error process eventually resulting in Pareto efficiency. It is claimed to be superior to capitalism because enterprise managers supposedly no longer have to maximise profits for business owners and therefore can set prices to be equal to marginal costs, keeping availability as high as possible. Finally, the state’s ability to manipulate resource prices means that externalities can be more easily accounted for and factored into the eventual cost of the good which creates the externality in the first place.

Milton Friedman criticised Lange’s model on the basis that it was completely based on abstract principles and logic, without consideration of real-world factual evidence whatsoever, along with a wide range of other methodological flaws. However, even if Lange is taken at his best and a pricing system can exist, a crucial mechanism by which capitalist economies attain efficiency is through entrepreneurship. Entrepreneurs play a critical role in supplying unfulfilled needs in various markets, taking advantage of situations where an efficient, productive business could be built and eventually fulfilling the need for customers. Under socialism, entrepreneurs lack the profit motive as an incentive to carry out this kind of activity and bear the risks of failure. In many cases, entrepreneurship was outlawed completely in socialist societies.

Assuming that humans in both scenarios are motivated to a great extent by self-interest (a cultural and psychological phenomenon which we can presume exists for the foreseeable future), as opposed to the profit motive pushing individuals to carry out market interactions which benefit both parties, in a socialist system the ultimate road to the promotion of one’s interests is the accumulation of political influence and status. At the managerial level, this is largely achieved by the fulfilling and exceeding of quotas and overperformance compared to other competing departments and state enterprises. This achieved through the securing of resources and funding, and the political influence necessary to acquire such means, eventually creating a cycle in which officials and bureaucrats can move up the ranks. The consequence of this was the inefficient reporting of information, with the tendency of Soviet Union managers to bribe quality control officers and union leaders discussed by Heinzen in his book The Art of the Bribe in 2007.

Another aspect of the problem extensively discussed by Hayek was the incompatibility of central planning with a democratic government. Hayek partially presupposed Arrow’s Impossibility Theorem in demonstrating that even when the process of calculation was fully democratic it was incapable of satisfying the preferences of a majority of consumers in many cases, given the existence of preferences as usually a ranked list of different options (the Theorem’s main claim that with a fair democratic system no ranked electoral voting system can successfully convert ranked preferences into a Pareto efficient policy). Whilst the link between Arrow’s Theorem and economic decision-making has been challenged, Hayek makes the further point that coercion on a large scale (through central planning) requires the rarest of things: a democratic consensus. That is, however, a matter of political science as opposed to economic science.

The ideology of central planning is in the present moment either based on the hubris of the claim that a single group of humans can direct a whole economy, or fanciful, unrealistic and flawed faith in a supercomputer capable of playing God. A system where each individual holds his small piece of the puzzle is far better in creating a coherent final product that fits together to form a prosperous economy.

Leave a Reply