Mises Daily Articles
Proof That Socialism Cannot Work
Socialism, a socioeconomic system in which the means of production are publicly rather than privately owned, is a very old idea that has occurred in many guises throughout history. There correspondingly have been many objections to it. Most of these, it is probably safe to say, have focused on issues such as justice for property owners, or on the ability of socialism to provide sufficient incentives for economic actors, or on the dangers of State power, or even on the nature of man. While these objections are of tremendous importance, from an economic perspective, they really do not get to the heart of socialism’s problems.
It was not until 1920 that Ludwig von Mises put forth the first truly devastating critique of socialism: namely that, absent private ownership in the means of production, economic calculation—the comparison of anticipated revenues with prospective costs in a common (monetary) unit—is impossible. (See his article.) Even if one grants the socialists their best possible case—that they are well-meaning, farsighted individuals armed with complete knowledge of present and future conditions and have the best interests of consumers in mind, as opposed to being fools and psychopaths—from their own perspective, they cannot rationally compare costs and gains, or recognize profits and losses. This is an extremely deep argument the origins of which go far deeper than merely a topical concern with socialist economy (see Rothbard, 1991).
While Mises’s original argument was largely misunderstood (if not rejected out of hand) at the time by supporters and critics alike, events have borne him out. The economic chaos in the former Soviet Union in the wake of communism is now apparent for all to see, even if the causes are not widely acknowledged. In his superb postscript to Mises’s seminal article, "Economic Calculation in the Socialist Commonwealth," Salerno (1990) calls attention to a particular manifestation of this chaos. Why, in the Soviet Union, did tractors sit idle in the fields due to lack of fuel and labor for operating them, given such a great desire for food? How does this example illustrate the importance of economic calculation and its necessary precondition—a system of property rights in the means of production?
Consider the decision by the Central Planning Board (CPB) to produce wheat for consumers. Perhaps it knows with certainty how much property consumers will be willing to give in exchange for wheat as opposed to, say, shoes. Based on this knowledge, the CPB must then decide how to provide this wheat and how much of the wheat to produce relative to shoes. How can it make this decision, having decided upon wheat production in general?
Consider the case on a free market for producer goods. Of all the many technologically feasible ways of producing wheat, only those will be brought to bear the prices of which are less than the anticipated revenues they will bring. Of course, prices are a function of ownership, and under different conditions (patterns of ownership) these factors may be more or less expensive relative to the wheat they will produce. Thus, how much wheat will be produced depends on those factors of production being owned, as well as consumer tastes and values. If some factor owner will not surrender that factor for less than any prospective entrepreneur thinks he can get in exchange for wheat, then no wheat will be produced with that factor.
This avenue is unavailable to the CPB under socialism. At most, the CPB can only know which consumer goods are most valued. It cannot know how much to produce, because that depends on ownership of the potential factors of production. The CPB can arbitrarily decide to produce x tons of wheat, and can in principle choose those means best suited in a technological sense for this purpose. By doing so, however, the CPB runs the risk of exhausting technology in that endeavor at the expense of some other endeavor. That is, it may be physically and technologically impossible to meet some other end due to decisions to meet some wheat quota. Even producing a given means requires application of some other means, so one cannot guarantee even producing a particular end. Hence, the Russian situation of idle tractors due to the unavailability of complementary means.
The equations of neoclassical economics are of no use in solving this problem because they explicitly treat consumer utility as conceptually distinct from the factors of production for economic decision-making. That is, they treat the decision-making problem as one of maximizing consumer utility subject to constraints on the factors of production (due to physical scarcity/availability and technological ability) and budgetary constraints on consumers.
However, the problem cannot be so separated. The decision of what to produce (and how much) is inextricably linked to the ownership of factors of production. The only way consumers can obtain more of some good is either if (1) they are willing to pay sufficiently more (property) for it, so that it becomes possible for a factor to be profitably bid away from its owner, or (2) a factor owner becomes willing to exchange the factor for sufficiently less property than before, so that this factor can be used in the production of that good. Thus it is never a question of simply satisfying consumer demands, because in a free market these demands are constrained by the availability (stemming from ownership) of the means for meeting those demands.
The neoclassical equations, by their very nature, do not capture this aspect of the problem, apart from their other shortcomings (static nature, assumptions of perfect knowledge, cardinal utility, indifference analysis, etc.). It is clear socialism cannot solve this problem, either. It may be possible for the CPB to optimally assign factors to the technological production of some specified end. In an economic sense, though, this decision is completely arbitrary because the end may itself be arbitrary, and in fact the CPB may not be able to technologically meet all possible ends, or any at all. Whether it has perfect knowledge of time and place is irrelevant.
Consumers do not simply prefer, for example, 1.5 tons of wheat and 2,000 pairs of shoes to 2 tons of wheat and 1,000 pairs of shoes. Rather, they prefer to give some amount of property for some amount of wheat and some amount of shoes. However, it makes no sense to speak of how much they will give in exchange for wheat and shoes without reference to the available means for producing these things, which determines the availability/scarcity of wheat and shoes themselves. From the perspective of economic planning, the availability of these means depends on their ownership. Prices for consumer goods (like any other goods) are determined by both the demand for them and the supply of them. Thus the "prices" used in the neoclassical production equations are meaningless, because on the market, these consumer goods prices also depend on the prices of the means for producing these goods. These equations seek to determine how to "optimally" produce a given set of ends, but since these ends are arbitrarily chosen, there is no reason to think that the resulting equations will have any kind of meaningful solution, if at all.
On the free market, the price of any good is set by the marginal buyers and sellers. For example, suppose there is one buyer who will give no more than $5 for a case of beer, and another who will give no more than $4.50. Suppose there is a supplier who will give his case of beer for no less than $4.75, and one who will give for no less than $5.25. In this case, the marginal buyer is the $5 one, and the marginal seller is the $4.75 one, since that buyer will buy at a price up to $5 and that seller will sell at a price down to $4.75, and no other buyers and sellers can enter the market at these amounts. The price will be set somewhere in between, and one case of beer will be sold. The only way an additional case of beer will be brought to market is if the $4.50 buyer increases his maximum buying price or the $5.25 seller reduces his minimum selling price. In either case the decisions of property owners determine what is produced (and how much).
Of course, in reality these actual maximum and minimum prices are not known with certainty, if at all; some (or much) judgment about future conditions is required before a decision to produce something can be made. However, that is a different issue and does not affect the nature of price formation and economic decision-making. It may be the case (pardon the pun) that the CPB would know these maximum buying prices of the consumers. What it cannot know, however, is the minimum selling prices of producers, because no such prices can exist under socialism. It therefore cannot "know" the prices of these consumer goods, as that is determined by the interplay of supply and demand, and this is absent under socialism.
Socialism’s fatal flaw cannot be a lack of knowledge because there is nothing for it to know; supply is a reflection of ownership, not some fact of the world that could come to be known in an objective sense and "revealed" through prices. Thus in deciding, say, to produce two cases of beer, the CPB necessarily allocates resources to meeting ends that are suboptimal (in this case), even if it could select the technologically best way of making this amount of beer because that choice necessitates other ends being foregone.
Of course, it may be objected that under a capitalist system, entrepreneurs "know" which of the myriad technical ways are best via the price system, and that socialist planners would not possess this information. However, on the free market, what are better or worse production techniques change when the ownership of these factors changes. So, it cannot be said that a particular choice by the CPB is necessarily bad,because there might be some pattern of ownership on the market under which this choice is in fact good. But, by making this choice, the CPB necessarily affects all of the other choices it can make, to the extent that no choice can be successfully brought to fruition. This situation does not prevail on the market (abstracting from error), because choices there are always constrained by property ownership. No such constraint exists under socialism, so choice becomes arbitrary, at best. This constraint on any choice is independent of the knowledge of how to realize some choice.
Consider a more relevant example of the above. On a given market, consumers A, B, and C have maximum buying prices for some good of $6.00, $5.75, and $5.50, respectively. Means that could be used to produce this good are owned by producers D, E, and F, with minimum selling prices of $5.25, $5.40, and $6.10, respectively. (This is not meant to say that the same kind of means have different prices on the market; rather, different types of means which could produce the same result will in general have differing prices, making some economically feasible and others not.) It is possible for entrepreneurs to bid away the factors from producers D and E (for, say, $5.30 and $5.60) because they can sell their products to a consumer whose maximum buying price exceeds what they must pay for these factors (abstracting from discounting through time). The factors owned by F will not be bid away because the price that must be paid for these factors exceeds the maximum buying price of any consumer.
For it to be economically feasible to bring this factor into the production of the good in question, either some consumer will have to raise his maximum buying price to above $6.10, or producer F will have to lower his minimum selling price to below the maximum buying price of the consumer whose desire for this good is unmet. In either case, it is ownership that determines whether the factor will be so applied.
On the free market, entrepreneurs bid away factors of production to make a certain amount of consumer goods, because, in their judgments, there is a sufficient number of people whose maximum buying prices will enable the entrepreneurs to sell their wares and make a profit. In other words, based on their judgments of (future) consumer tastes, and due to the fact that exchange ratios (prices) exist on a free market, the entrepreneurs can calculate the amount of property needed to realize a particular production process. For a particular factor of production, an entrepreneur can determine how many consumer goods thus can be produced and, given the factor’s price, how much each good must be sold for to earn a profit. If the entrepreneur judges that there is (or will be) an insufficient number of consumers with a maximum buying price for that good to enable a profit, then that factor will not be so used and those goods will not be produced.
Thus, without prices as a tool of economic calculation, the CPB may very well decide to produce a number of this good that is either too much or too little relative to consumer tastes, if not physically impossible to realize. It could in principle pick just the right number. However, this decision would be completely arbitrary and better described as a lucky guess. Of course, choosing one end implies choosing some means to attain it, and quite often these means require some other set of means for their realization. Thus the means-ends framework exists at all stages of production. It is fantasy to think that the CPB could get lucky at every stage, but lacking ownership in these means, at every stage it must make an arbitrary decision as to what and how much to make. Thus, the chaotic Russian situation under communism is not surprising.
Even with perfect knowledge of all its circumstances, the CPB’s decisions are basically a coin flip. It is important to note that prices as exist under capitalism do not impart any knowledge regarding correct economic decision-making; errors in judgment about the future can and do occur under capitalism. Rather, they provide a tool for making coherent judgments about means and ends. But they can only do this because of their origin in ownership, in their nature as an amount of money property that must be used to acquire some factor that can be compared with the amount of money property that can be obtained from its proceeds. It does not matter how "dispersed" the knowledge of some production technique is. Only those techniques will be applied the prices of which are less than the revenues they are anticipated to bring. (See Hülsmann  for a brilliant discussion of the nature of economic calculation and the fundamental role of property.)
It is true that a particular allocation could be economically feasible under some pattern of ownership of production factors. In the example above, the pattern of factor ownership and consumer valuations meant that two of the consumer good would be produced. Under a different owner, the remaining factor needed to produce a third consumer good might become available. So while the CPB’s decision to produce three consumer goods with these factors is economically wasteful under the existing pattern of ownership, it might not be under another set of conditions.
However, by applying these factors to a particular line of production, they cannot by necessity be applied to a different line. Under a free market, diversion of a factor to satisfy one need over another can only be reversed if consumers are willing to pay more (property) to satisfy that need or if producers are willing to accept less for the means needed to satisfy that need. Unless that happens, less of that need will be met. Thus it is not enough for the CPB to mandate that x amount of some consumer good be produced. They must also mandate that x’ of another good be reduced. Now this decision may also be consistent with some pattern of ownership on the free market. However, there is no guarantee that the patterns of factor ownership implied by allocations x and x’ are consistent. That is, they may not necessarily reflect the valuations of the same set of owners on a hypothetical free market that the CPB is trying to imitate.
Consider the previous example of selecting possible means for a given end. By introducing another consumer good, for which the consumers have maximum buying prices of, say, $4.00, $5.00, and $7.00, we must also consider means for attaining that end. Consider on the free market possible means, whose owners have minimum selling prices of $6.00, $5.20, and $5.30. Given this pattern of ownership, only one such consumer good can be feasibly produced, since only one consumer will be willing to give a sufficient amount of property to render that venture profitable. The introduction of more consumer goods and associated producer goods poses no fundamental problem for capitalist entrepreneurs, because the same decision-making process applies here as with a single consumer good.
For the CPB, matters are quite different. It must decide on a number of the first good to produce, and a number of the second good to produce, and there is no guarantee that any given decision will be even technologically feasible. Means applied to good 1 cannot be applied to good 2. On the free market, by bidding away factors for one endeavor, an entrepreneur necessarily affects the prices of all other factors, because now the amount of means for other ends has become more limited. For example, by bidding away some means needed to produce one unit of good 1, the prices of the means for producing a unit of good 2 will be changed, so that now it may become infeasible to produce any of that good. Economic calculation will reveal this, however (pending a judgment about future conditions, of course). The CPB has no recourse to this approach, though. It must blindly forge ahead with its plan, even if it is in fact technologically infeasible and even if the CPB has perfect foresight of future consumer valuations.
In other words, on a free market, the valuations of factor owners are integrated to produce some amount of consumer good 1, consumer good 2, etc. Under socialism, the decision to produce some amount of good 1, good 2, etc., is completely delinked from any notion of factor ownership. There is no guarantee that the various pieces of the "plan" will fit together. Even with the case of a single consumer good, the decision to produce x of that good implies the need to produce some amount of means toward that end, so a decision must be made at that stage as well. These means may require further means for their realization, necessitating another stage of decisions. Without ownership in this means, no way exists for determining (via economic calculation) how much of the means can be produced, so the decision of how much to make is arbitrary, and that decision may render it impossible to realize other constituent means and ends.
The conclusion is clear: Only a system of private property and ownership in the means of production permits coherent economic decision-making, where the question of what to produce is linked to the question of how to produce. Any system that moves away from a free market is a recipe for economic chaos.