4. The Debate about the Efficiency of a Socialist Economy

CHAPTER 4

The Debate about the Efficiency of a Socialist Economy*

Non-academic socialists have in general bothered little about how a socialist economy would work. Even Karl Marx preferred to attack capitalism rather than describe socialism. In popular thought, socialist production was to be organized very simply, with sole regard to the needs of the comrades.

The popular slogan “production for use, not for profit” overlooks the fundamental economic problem of scarcity. It is simply impossible to satisfy all of everybody’s wants fully. Without some indexes of the intensity and satiability of wants and of the scarcity of resources relative to usefulness, rational economic calculation is out of the question. Suppose ten more small radios could be produced at the sacrifice of one large television set. Would national output thereby be increased or decreased? Without some concept of “value,” the question is meaningless. Suppose a certain product could be produced either with ten units of land plus five of labor or with four units of land plus nine of labor. Which method of production is more economical? Without some concept of “value,” the question is meaningless.

When socialism became an immediate political issue at the end of World War I, the Austrian economist Ludwig von Mises spoke forth to deny that socialism is economically practicable. Mises’s main argument can be summarized—as the socialist H.D. Dickinson (1939, p. 111) has done—in three statements:

1. Rational economic activity requires the pricing of all goods, production goods as well as consumption goods.

2. Pricing requires the existence of a market.

3. A market requires the existence of independent owners of the goods exchanged.

Listen to Mises’s own words:

In any social order, even under Socialism, it can very easily be decided which kind and what number of consumption goods should be produced. No one has ever denied that. But once this decision has been made, there still remains the problem of ascertaining how the existing means of production can be used most effectively to produce these goods in question. In order to solve this problem it is necessary that there should be economic calculation. And economic calculation can only take place by means of money prices established in the market for production goods in a society resting on private property in the means of production. That is to say, there must exist money prices of land, raw materials, semimanufactures; that is to say, there must be money wages and interest rates. (1922/1981, pp. 141—142)

At about the same time that Mises’s famous article appeared in 1920, similar ideas came from the pens of Max Weber in Germany and Boris Brutzkus in—of all places—Russia. Brutzkus, for instance, wrote:

just as capitalism possessed a general measure of value in the rouble, so socialism would have to possess an analogous unit for the evaluation of its elements... Without evaluation any rational economic conduct, under whatever kind of economic system, is impossible. (1935, p. 15; italics in original)

Mises, Weber, and Brutzkus were not the first writers to question the economic efficiency of arbitrary planning. For instance, as early as 1902, the Dutch economist Nicolaas G. Pierson (1935, pp. 41—85) had emphasized that a socialist community would have to face the problem of value. But it was left for Professor Mises to revolutionize academic discussion. This Mises accomplished by his dogmatic insistence that rational economic calculation under socialism would be impossible. In Mises’s own words, “Every step that takes us away from private ownership of the means of production and from the use of money also takes us away from rational economics” (1920/1935, p. 104).

Ensuing discussion of how to avoid the pitfalls stressed by Mises became reminiscent of an article published by Enrico Barone in 1908. In his “The Ministry of Production in the Collectivist State,” Barone had applied Pareto’s system of equations to demonstrate that “all the economic categories of the old regime must reappear, though maybe with other names: prices, salaries, interest, rent, profit, saving, etc.” (1908/1935, p. 289). The Ministry of Production, through deliberate arrangement, would have to satisfy the two conditions that would result automatically from perfect competition, that is, equalization of prices with cost, and minimization of costs of production. The system of equations giving the correct allocation of resources and labor would be identical with the system reflecting the operation of free competition. However, Barone assumed that the Ministry would actually have to formulate and solve such equations, so he referred skeptically to “the laborious and colossal centralization work of the Ministry (assuming the practical possibility of such a system)” (p. 290).

The concept of “optimum conditions” is either explicit or implicit in the work of Barone and many post-Mises writers. If one makes a number of assumptions—such as that people’s preferences as workers and consumers are to “count” and that an ethically desirable income distribution can be achieved—then it is possible to deduce certain conditions which must prevail in a situation of maximum welfare as a situation in which no household could be made still better without some other household being made worse off in consequence. Following Vilfredo Pareto, J.R. Hicks, Paul Samuelson, Abram Bergson, M.W. Reder, Abba Lerner, Oskar Lange, and other recent writers have explicitly formulated sets of optimum conditions. The optimum conditions are the heart of modern welfare economics.

To save time, I shall not read a typical set of optimum conditions. The conditions are more easily followed when seen in print than when merely heard, anyway. Suffice to say that Professor Hayek has summed up several of the conditions very neatly: “the marginal rates of substitution between any two commodities or factors must be the same in all their different uses” (1948, p. 77).

The optimum conditions may be restated briefly in terms of “costs.” The total cost incurred in the production of the optimum amount of any commodity must be a minimum, that is, the average cost at optimum output must be a minimum. The optimum output of each commodity is specified by the condition that marginal cost equal price. The “prices” used in valuing outputs and measuring costs need not be thought of as market prices; they can—in principle—be mere indexes inversely proportional to the common subjective marginal rates of substitution for households and the common marginal technical rates of substitution for firms.

The point of the discussion so far is that a socialist planning board could formulate a set of optimum conditions and translate these conditions into mathematical equations. If the planning board knew in complete detail the preferences of all people as workers and as consumers, if it knew the production functions of all productive processes, if it had detailed information on the stocks of all resources, and if it could decide on some scheme of income distribution, then—conceivably—it could determine the amounts of each and every sort of good and service which “should” be allocated to each and every use. If the planning board were omniscient, then—as Hayek remarks—the solution of the community’s economic problem would be a matter of pure logic (Hayek 1948, p. 77). For the board would embody its omniscience in mathematical equations. As Barone showed, there would be as many equations as unknowns. The equations are to be solved simultaneously, and the solution would be determinate.

Writers before the time of Mises’s famous article—notably Wieser, Pareto, and Cassel, as well as Barone—had used the concept of equilibrium determination through simultaneous equations as an expository device. But some socialist writers have envisaged the solution of simultaneous equations as the actual method of socialist resource allocation (for instance, Carl Landauer, and—at one time—H.D. Dickinson).

On this approach, the comments of Lionel Robbins are most pertinent. I quote from Robbins, The Great Depression (1934, pp. 150—151).

On paper we can conceive this problem to be solved by a series of mathematical calculations. We can imagine tables to be drawn up expressing the consumers’ demands for all the different commodities at all conceivable prices. And we can conceive technical information giving us the productivity, in terms of each of the different commodities, which could be produced by each of the various possible combinations of the factors of production. On such a basis a system of simultaneous equations could be constructed whose solution would show the equilibrium distribution of factors and the equilibrium production of commodities.

But in practice this solution is quite unworkable. It would necessitate the drawing up of millions of equations on the basis of millions of statistical tables based on many more millions of individual computations. By the time the equations were solved, the information on which they were based would have became obsolete and they would need to be calculated anew. The suggestion that a practical solution of the problem of planning is possible on the basis of the Paretian equations simply indicates that those who put it forward have not begun to grasp what these equations really mean. There is no hope in this direction of discovering the relative sacrifices of alternative kinds of investment. There is no hope here of a means of adjusting production to meet the preferences of consumers.

In his article “On the Economic Theory of Socialism,” Oskar Lange makes fun of Robbins and Hayek for worrying about whether socialists propose to solve simultaneous equations (Lange and Taylor 1938, p. 88). But before deciding whether Robbins and Hayek deserve this ridicule, lets see what Carl Landauer wrote in a book published as recently as 1944. The following lines are, to me, among the funniest in all economics:

Price formation on the market is a search for an equilibrium through trial and error. The sellers and buyers change their charges and biddings until a price is established which just equilibrates supply and demand. The planning board can carry this process out on paper with infinitely less cost and loss of time than it can be carried out by sellers and buyers in reality.

... it is possible to establish a system of simultaneous equations, in which the combined effects, in terms of utility produced, and the physical quantities of each element in each combination appear as the knowns and the unit values of the elements as the unknowns.

There is no difficulty at all in finding as many equations as we wish, since there are almost innumerable combinations of goods and we have always enough to calculate the unknowns.

Instead of following the rules of the mathematical text-book, we may systematically change all the elements of the equations until we arrive at magnitudes which will satisfy the conditions.

The planning board ... applies the trial-and-error process on paper... This form of experimental variation is an immense economy in time, effort, and material as compared with the experiments in steel and timber, copper and labor, selection of occupation and expenditure of consumer dollars, which is the prevalent form of approaching an equilibrium under the status quo.

... the equations will not reach a fabulous number, but it is quite possible that a few hundred thousand combinations have to be taken into account. Why this should exceed the power of algebraic analysis is difficult to see. It probably does not exceed the amount of calculation work which a hundred middle-sized engineering firms have to perform in a week.

It is a crude method to search for an equilibrium by experimentally varying all the determinants until they fit together, and, although it is infinitely more economical to carry out these variations on paper than in reality, a further great economy of effort might be achieved through the development of mathematical shortcuts. A future generation may look upon the trial-and-error process very much as a second-year schoolboy, knowing the use of multiplying and dividing techniques, looks at the abacus. But in the meantime, the abacus method serves the purpose of demonstrating that the problem is in any case soluble, practically as well as theoretically¹ (Landauer 1944, pp. 34—41)

Before proceeding to socialist schemes that assign a large role to prices, let’s examine the optimum condition that for every product, marginal cost should equal price. This is Abba Lerner’s famous instruction to managers of socialized enterprises. Its implications have been one of the hottest topics of debate in the field of welfare economics.

First, just why should marginal cost and price be equal? The marginal cost of commodity A is the sum of the prices of the additional factor-units needed to make an extra unit of commodity A. The price to a producer of commodity A of each factor-unit equals the factor’s marginal value productivity elsewhere. If consumers will pay more for commodity A than its marginal cost, this means that the necessary additional factors will produce more satisfaction (as measured by what consumers are willing to pay) in the production of commodity A than in the production of anything else. If, on the other hand, consumers will not pay as much for commodity A as its marginal cost, this means that marginal factors will produce more consumer satisfaction elsewhere. Now, an additional unit of any factor should have the same marginal value productivity—that is, the same marginal productivity of consumer satisfactions—in all its uses. Therefore, production of commodity A should be expanded or contracted to the point at which marginal cost and price are equal.

If average cost of a commodity decreases throughout the relevant range as output expands, marginal cost is less than average cost. The total amount paid for the commodity on the basis of marginal-cost pricing will thus fall short of total costs. But since price should equal marginal cost, the industry should run at a loss, and the government should make up the loss out of general taxation.

Alfred Marshall had already in his Principles (1890/1920, pp. 469—473) championed subsidies to decreasing-cost industries in the old-fashioned language of “consumers’ surplus.” In recent years Harold Hotelling (1938, pp. 242—269) and Abba Lerner (1947, esp. pp. 194—199) have insisted vigorously on the proposition; and though the conclusion is startling, the Hotelling-Lerner logic, summarized above, seems impeccable.

The disconcerting thing is that another chain of reasoning leads to a seemingly opposite conclusion. R.H. Coase (1946, pp. 169—182, esp. p. 172) states two principles of optimum pricing:

1. For each individual consumer the same factor should have the same price in whatever use it is employed. Otherwise consumers would not be able to choose rationally, on the basis of price, the use in which they preferred to have the factor employed.

2. The price of a factor should equate its supply and demand and should be the same for all consumers and in all uses.

From these principles Coase draws the implication that the amount paid for a product should be equal to the total value that the factors used in its production have in another use or to another user. In other words, the price of a product should be equal to its full cost.

Coase argues that if certain factors of production can be obtained free in one use (because they do not enter into marginal cost), but have to be paid for in another use (because they do enter into marginal cost), then consumers may choose to employ these factors in the use in which they are free, even though they would in fact prefer to employ them in some other way. If the Hotelling-Lerner solution were adopted, there would be only one way out of the difficulty. That would be for the state to decide whether or not each consumer should be supplied with the particular good in question. This would be done by estimating whether or not each consumer would be willing to pay the full cost of supplying him if he were called upon to do so. Coase further argues that no government could estimate individual demands accurately; that if all pricing were on a marginal-cost basis, there would be less information available by which such an estimate could be made; and that the incentive to correct forecasting would suffer if there were no subsequent market test of whether such estimates of individual demand were correct or not.

Coase’s second objection to Hotelling-Lerner pricing is that it would redistribute income in favor of patrons of decreasing-cost industries. It is not easy to imagine how such a redistribution might be considered ethically desirable. For instance, Harry Norris (1947, pp. 54—62) imagines two countries which are identical except that one is lighted by a constant-cost gas industry and the other is lighted by a decreasing-cost electric industry. In each country ten percent of the taxpayers are backwoodsmen living beyond the range of utility service. In the gas-lit country, each household pays for its lighting in full. But according to the Hotelling-Lerner solution, the backwoods taxpayers in the electric country should subsidize the lighting of the city dwellers. Yet this arrangement could hardly be defended on ethical grounds.

A third objection to Hotelling and Lerner is that the taxation necessary to raise subsidy money would have the familiar disincentive effects where imposed.

A fourth set of objections to marginal-cost pricing is mentioned, strangely enough, by a champion of the Hotelling-Lerner solution and opponent of Coase’s proposals. William Vickrey (1948, pp. 218—238) points out that in some instances the application of Hotelling-Lerner pricing might involve serious political and sociological consequences. Vickrey may have in mind the dangers of any new excuse for raids on the United States Treasury.

R.H. Coase (1946, pp. 169—182, esp. pp. 173ff.) seeks to reconcile the implications of his own and the Hotelling-Lerner reasoning by the ingenious device of multi-part pricing. Patrons of decreasing-cost industries are charged a lump sum or series of lump sums which are supposed to cover intra-marginal costs. Each consumer is then allowed to obtain additional units of product at the marginal cost. The advantage of multi-part pricing is that consumers can be asked to pay a total amount which is equal to the total cost. Therefore, it is possible to discover whether consumers value the total supply at least the total cost of supplying them. (Under a pricing system, whether consumers are willing to pay an amount equal to total cost can be discovered only by actually asking them to pay this amount.) At the same time, additional units are supplied at additional cost, and so the right output can be obtained; that is, the Lerner allocation of factors can be achieved.

William Vickrey (1948, pp. 218—238) raises a number of practical objections to multi-part pricing. But—to put it mildly—a number of practical objections could be leveled against Hotelling-Lerner pricing also. In summary, I believe that Coase’s multi-part pricing scheme remains at least as intellectually respectable as marginal-cost pricing.

Writers who were unhappy about the prospect of actually having to solve millions of simultaneous equations formulated the so-called “competitive” or “trial-and-error” solution to the problem of socialist resource allocation. There is a large measure of agreement in the ideas of Oskar Lange and Fred W. Taylor (1938), A.C. Pigou (1937), R.L. Hall (1937), H.D. Dickinson (1939), and Burnham P. Beckwith (1949).

According to the “competitive solution,” households have freedom of choice in regard to jobs and consumer goods. Households receive and spend actual cash. The central authority instructs the manager of each production unit to operate in accord with two basic rules. For any given scale of output, he must combine the factors of production in such a way as, at the established prices, to minimize average cost per unit of output. Secondly, he must fix output so that marginal cost of the product equals its established price.

The question arises: How do the prices of consumer goods, intermediate goods, and productive resources get “established”? Answer: the central authority does it. (Incidentally, the prices for all things except labor and consumer goods need not be market prices; they can be mere accounting prices.) The central authority decrees a price for each good. If the managers of productive units follow the rules, they will—like entrepreneurs under perfect competition—regard the prices as parameters. Of course, the prices decreed by the central authority will not be correct at first. There will be shortages of some things and surpluses of others. Consequently, the authority will raise the prices of things in excess demand, and lower the prices of things in excess supply. By constant experimenting—by trial-and-error—the central authority is supposed to make everything work out all right.

Similarly, the central authority could make capital freely available to the socialist enterprises at an established rate of interest. Interest would be reckoned among the elements in cost. In many socialist blueprints, the central authority would arbitrarily fix the total amount of liquid capital available. In any case, the authority could conceivably equate the demand and supply of capital by manipulating the interest rate, just as it would manipulate other prices. Incidentally, Enrico Barone had already mentioned the idea of setting the interest rate by trial and error in his essay of 1908 (pp. 268—269).

The main feature of the Competitive Solution, then, is that some Board, rather than the market, adjusts prices to bring supply and demand into line.

Several questions about this Lange-Taylor “solution” are obvious:

1. How would the Board force managers to obey the “rules”? For instance, enterprises that were large in relation to their market might profit by taking into account the effects of their actions on the Board’s price-setting decisions. The managers might restrict output in much the same way as monopolists do under capitalism. Even assuming away all questions of monopolistic motives, there still remains the problem of whether managers should be allowed to act upon their anticipations of price changes by the Board.

2. How would the Board judge the efficiency of managers?

Of course, in coping with those two difficulties, the Board might look into the books of the individual production units. But if carried to any length, this practice would conflict with the essential aim of decentralizing decisionmaking.

3. How would the Lange-Taylor solution apply to the prices of commodities that cannot be standardized, such as large units of capital equipment which must be made to order? In all such cases there would be no basis for centralized fixing of prices so as “to equalize demand and supply.” Socialist writers simply ignore the various complications that would arise.

4. If the price-setting authority perceives, for example, that the demand for rubber exceeds the supply, how is it to know whether this indicates too low a price for rubber or too high a price for tires? If the demand for sheet aluminum is less than the supply, is the price of aluminum too high, or is the price of plywood too low? If the demand for gasoline falls short of the supply, is this due to too high a price for gasoline or too high a price for automobiles? The point is that the central authority would have to work with Walrasian, and not merely Marshallian supply and demand functions. The authority could not change a single price without changing the equilibrium prices of other goods. Trial-and-error would not be simply a matter of making particular prices higher or lower; it would be a problem involving at least many millions of possible patterns of price increases or decreases. And for each increase or decrease there would be the problem of how much.

Furthermore, continual shifts and changes in supply and demand functions themselves would blur the central authority’s view of the effects of its trials-and-errors.

Still further comments on Lange-Taylor socialism are in order. Like some of the foregoing points, they are stated most cogently in the writings of Professor Hayek (1948, chaps. IX and IV). For instance, by the very nature of administrative decisions, such price changes as were made would occur later than if the prices were determined on a free market. Secondly, the central price authority would differentiate much less than a free market would between the prices of commodities according to differences of quality and the circumstances of time and place. This means that managers of production would have no inducement, and even no real possibility, to make use of special opportunities, special bargains, and all the little advantages offered by special temporary or local conditions. Society could not make full use of the sort of dispersed knowledge that cannot be collected in the form of statistics, for example, knowledge of vacant space in the hold of a tramp steamer about to sail, knowledge of a machine that is not being fully used, knowledge of a particular person’s skill that might be better utilized.

Furthermore, theorists of the Lange-Taylor school err in regarding cost curves as “given.” One function of capitalist price competition is to reduce costs to a minimum. Under socialism, the new man with the new idea is not able to enter an industry and undercut old producers unless the central authority approves his projects und provides him with capital.

Further comments on Lange-Taylor socialism apply to an even more decentralized type of socialism as well. I will save these comments until the discussion of that other type.

Professor Pigou, for one, recognizes the difficulties that would beset any attempt to apply his Competitive Solution to the problem of socialism. He says: “Evidently ... the practical difficulty of working such a process will be enormous... Far-reaching errors are almost inevitable.” Nevertheless, Pigou congratulates himself on the fact that his “analysis shows that the allocation problem is soluble in principle” (1937, p. 115). This brings me to remark that I cannot understand the common charge that critics of socialism base their case on “merely practical” objections. As Lionel Robbins observes, “it is one thing to sketch the requirements of the plan. It is another thing to conceive of its execution” (1934, p. 150). Of course the allocation problem is soluble in principle. But it is just as possible, in principle, to breed winged elephants. Anyone who doubts this can find out how from me later.

Some socialist writers, notably Maurice Dobb, find it hard to take seriously the suggestion that plant managers should “play an elaborate game of bidding for capital on a market, instead of transmitting the information (about productivities) direct to some planning authority” (1946, p. 302). Dobb and his school take refuge in a large measure of central planning of the Soviet type. Paul Sweezy, also, favors comprehensive planning; indeed, he agrees with H.D. Dickinson that centralization is all but inevitable under socialism. Yet Sweezy (1949, pp. 232—239) cites Lange’s On the Economic Theory of Socialism “as having finally removed any doubts about the capacity of socialism to utilize resources rationally,” and he seems not to realize the inconsistency of his own position.

Since the centralized-planning approach is unquestionably vulnerable to the criticisms advanced long ago by Mises, I find it a peculiarly uninteresting form of socialism. Therefore, let’s go on to an extremely decentralized form of socialism.

Abba Lerner, in The Economics of Control (1947), and Lerner and Oskar Lange, in a pamphlet published in 1944, advocate “free enterprise.” By this, Lerner and Lange mean that both government and private entrepreneurs should be free to enter any line of business not reserved to the government. Perhaps this arrangement would be the famous “mixed economy” rather than full-blown socialism. Franco Modigliani (1947, pp. 441—514) advocates a decentralized socialism in a lengthy article published in 1947, though he does not go nearly as far towards so-called “free enterprise” as do Lerner and Lange.

Lerner and Lange recognize that

what the manager of every factory needs is some simple indication of the usefulness for alternative production of each of the goods that he might use. Such an indication of the alternative productivity of each factor is provided by its price. For this we must have markets for the factors in which the price equates the supply to the demand, with appropriate rules governing the demand for the factors of production by the various managers in charge of the public enterprises. (1944, p. 18)

Incidentally, I cannot help but interpret that quotation from two of the most competent socialist writers as anything less than an outright concession of at least two-thirds of Professor Mises’s original argument. It is also noteworthy that Lange has abandoned the position he held in On the Economic Theory of Socialism.

Of the various instructions which Lerner, Lange, and Modigliani address to the managers of production, the most important is our old friend: equate marginal cost and price. But the central planning board now seems to have lost its job of setting prices by trial and error.

In The Economics of Control, Lerner writes:

In each market, whether for factors or for products, prices are raised whenever the demand for any product or factor is greater than the supply and lowered when the supply is greater than the demand until a set of prices is reached in which each demand is equal to the corresponding supply. (p. 63)

It is not clear who is to do the raising and lowering. Lerner seems to leave the task to a market rather than to a Board. But if prices are not parameters either established by atomistic competition or by some Board, then the rule “equate marginal cost and price” is not unambiguous.

Professor Morgner has suggested to me that perhaps the idea is to have each manager operate at the output and sell at the price indicated by the intersection of his average revenue and marginal cost curves. Some remarks by Paul A. Samuelson suggest that this interpretation may be correct:

the decentralized operators in a planned society should refrain from a literal aping of atomistic, passive, parametric price behavior. Instead of pretending that demand curves are infinitely elastic when they are not, the correct shape of that curve is to be taken into account. This does not mean that the decentralized operators should take account of their influence on price as a monopolist would. (1948, p. 232)

Assuming that the equating of marginal cost and average revenue is what Lerner, Lange, and Modigliani have in mind, let’s see how this system would work:

1. Barring extreme coincidences, different enterprises would be trying to charge different prices for the same product. There is no reason for assuming that the marginal-cost-average-revenue intersections of different enterprises would be at the same price.

2. If managers receive prestige, power, or bonuses according to the prosperity of their enterprises, they would have an incentive to take advantage of imperfect competition by overcharging buyers and underpaying suppliers. If the central authority tried to enforce rigorous adherence to the “rules,” there would have to be duplication of management and so perhaps centralized planning after all.

3. Managers might not know their own marginal cost curves, even if determinate. Most important, managers almost certainly could not know their average revenue curves for the simple reason that such curves would not exist as independent entities. To be specific, the socialized enterprises would almost certainly be oligopolies, and the outstanding characteristic of oligopoly is absence of any determinate average revenue curve. Notice that I am not talking about “mere practical difficulties” of measurement. My point is that neither the most powerful statistical techniques nor the fact of government ownership could provide an oligopolistic firm with something nonexistent, that is, with a determinate average revenue curve.

Thus it appears that the Lerner-Lange-Modigliani rules do not yield any determinate price-output situation at all. The socialist economy would simply have to wallow in chaos.

Several observations apply both to Lange-Taylor socialism and to Lerner-Lange-Modigliani socialism:

1. The adoption of competition or quasi-competition means giving up whatever advantages centralized planning might afford (if those advantages were considered worth the price in terms of grave disadvantages). For example, competitive socialism has no sure cure for the business cycle. Dickinson, Modigliani, and others who discuss the problem pin their faith on what are essentially the sort of Keynesian fiscal policies that capitalism could adopt. But Beckwith has the most elegantly simple device for getting rid of unemployment: just cut wage rates to whatever extent may be necessary. With regard to wages, Dickinson and Beckwith declare themselves for piece-rates and the other trappings of Taylorism. Incidentally, few socialist price-theorists try to pretend that their systems hold any place for unionism as we know it today. To my knowledge, only Lerner and Lange cannot bear to stop mouthing the shibboleths of contemporary unionism.

2. It is questionable whether the diligence and the decisions of socialist managers of production would be governed by suitable incentives. As Mises points out, the capitalist entrepreneur

does not just invest his capital in those undertakings which offer high interest or high profit; he attempts rather to strike a balance between his desire for profit and his estimate of the risk of loss. He must exercise foresight. If he does not do so then he suffers losses—losses that bring it about that his disposition over the factors of production is transferred to the hands of others who know better how to weigh the risks and the prospects of business speculation.

Capitalists and speculators cannot be expected to act as mere agents of the community, for

the function which capitalists and speculators perform under Capitalism, namely directing the use of capital goods into that direction in which they best serve the demands of the consumer, is only performed because they are under the incentive to preserve their property and to make profits. (1922/1981, pp. 140—141)

Furthermore, there is no reason to think that the recklessness of some socialist managers would compensate for the over-cautiousness of others. Two wrongs don’t make a right.

3. As I have already implied, the schemes of price-theory socialists have an extreme static bias. We may well join with Professors Hayek, Mises, and Robbins in asking: What is to be the independent business unit? Who is to be the manager? What resources are to be entrusted to him? How is his success or failure to be tested? On what principle is the control of productive facilities to be transferred from one manager to another (Hayek 1948, esp. pp. 172, 196—197)? The idea of instructing the controllers of various industrial units to act as if they were capitalist entrepreneurs ignores the fundamental problem of a dynamic economy, the problem of deciding what resources should be risked in what ventures under the control of what men. As Robbins says,

For competition to be free the entrepreneur must be at liberty to withdraw his capital altogether from one line of production, sell his plant and his stocks and go into other lines. He must be at liberty to break up the administrative unit. It is difficult to see how liberty of this sort, which is necessary if the market is to be the register of the varying pulls of all the changes in the data, is compatible with the requirements of a society whose raison d’etre is ownership and control at the centre. (1934, pp. 153—154)

Mises remarks that when socialist theorists assume stationary conditions, the essential function of economic calculation has by hypothesis already been performed (1949, pp. 137ff.).

R.L. Hall worries about how anyone could tell whether a socialist monopoly were as efficient as it might be. In Hall’s words,

If a world champion has no one against which to measure himself, how can he tell of what he is capable? A runner can run against himself by means of a stop-watch: will a state concern be prepared to make continuous efforts to reduce its own costs? If it does not do so, no one else can. In a capitalist state Nemesis is always waiting for the lethargic monopoly, as the calamity of the internal combustion engine fell upon the railways. But it is doubtful whether anyone in a collectivist state would have any hope of starting a rival department. (1937, pp. 138—139)

4. Few blueprints of socialism make much specific provisions for economic development. Even Franco Modigliani’s meager remarks are unusual in the literature (1947, pp. 463—465). Modigliani faces squarely the disadvantage that decisions about new products would have to be made by an administrative body. He perceives the twin dangers of inertia due to lack of a profit incentive and of recklessness due to lack of a loss-penalty. (Modigliani feels that the tendency toward inertia is probably more characteristic of bureaucracy.) To avoid these dangers, Modigliani proposes a Research Commission to decide on new big investments and on new products requiring new plants. The Research Commission and the various industrial managers are to share the responsibility for mere product modifications, and the managers are to receive bonuses for successful innovations.

Doubts about the progressiveness of socialism do not rest on any supposed lack of incentives for scientists. As Brutzkus writes,

Scientific discoveries, it is true, are not made out of a desire for profit but in answer to humanity’s unquenchable search for truth. In the case of inventions the scientific interest recedes in favour of practical motives. But neither scientists nor even inventors are directly responsible for economic progress; it is the organizers and practical men who stimulate development.

Even supposing ... that the highest posts were filled in the best possible manner, there would still remain the danger that each innovation could only be tested in a definite place... .

If ... the socialist organization succeeded in assuming stable forms it would be distinguished by immense indolence and conservatism. It would offer nothing which could be compared to the unceasing movement of economic life under capitalism. (1935, pp. 67—69)

In his essay On Liberty, John Stuart Mill uses his arguments for freedom of expression as a case for freedom of enterprise:

The management of purely local business by the localities, and of the great enterprises of industry by the union of those who voluntarily supply the pecuniary means, is further recommended by all the advantages which have been set forth in this Essay as belonging to individuality of development, and diversity of modes of action. Government operations tend to be everywhere alike. With individuals and voluntary associations, on the contrary, there are varied experiments, and endless diversity of experience. (1859/1929, p. 131)

Modigliani’s Research Commission would hardly be an adequate substitute for Mill’s “diversity of modes of action,” even if it were staffed entirely by scientific and technical experts. The very essence of innovation is that it embodies ideas divergent from prevailing thought. Progress cannot be completely plotted and blueprinted in advance; that is why it is progress. Often discovery can be identified only in retrospect (cf. Harper 1949, pp. 72—76). If the government Research Commission decided not to “waste” resources on some new idea, it would be ruled out; that’s all. But in a competitive enterprise system, an idea discarded by ninety-nine companies still has a chance with a hundredth company. And no company can afford to be too rash in rejecting innovations, for fear that its competitors will “get the jump” on it. In a competitive system, a research worker who feels frustrated in one job can take his ideas to another employer. Under socialism things would be different.

With regard to invention, R.L. Hall writes:

there does not seem to be any criterion by which the Socialist state can decide the amount of resources which it is proper to spend: so that they cannot rationally create a profession of inventors. On the other hand, if these matters are left to the chance on which they depend the path of the innovator will be even harder than it is in the capitalist state. In the interests of economy it is necessary to discourage cranks; but it is probable that the leaders of industry will consider that all innovators are cranks. In the capitalist state the inventor is free to devote any resources which he inherits, or can earn or can wheedle from patrons, to his researches. In a socialist state he will get nothing which he does not earn and he may find himself in a labour camp if he neglects his work. The progress of spontaneous invention may well be slowed unless care is devoted to preserving a receptive frame of mind in the higher officials. But if they are too receptive there may be dissipation of the national resources in grandiose projections doomed to failure. It is difficult to follow a middle course when no one knows where the middle is: the socialist state will have no certainty, nor can any other form of organization give it. (1937, pp. 192—193)

In summary, innovations start out in a progressive economy as the whim of the few and are adopted only later by the inert masses. Innovation would be slowed in an economy which did not allow the widest possible scope for variety, whims, even eccentricities.

It maybe objected that the foregoing considerations are political objections to socialism, but do not constitute an economic argument. I do not think this is so. Whether an economic system is efficient or not is very largely a matter of whether it is progressive or not. The mere fact that we cannot handle questions of economic development by the precise and elegant techniques of price theory does not mean that such questions fall outside the scope of economics.

The political case against socialism is quite different. It would emphasize, among other matters, the danger that pressure groups could sabotage progress. Suppose, for instance, that the automobile had not yet been invented, but that some men had ideas for developing “horseless carriages.” Now which pressure group would have the ear of the government Research Commission—the buggy makers, with their thousands of votes, or the would-be automobile makers, with their mere handful of votes?

This paper does not handle the political case against socialism or the question of freedom or serfdom under socialism. I’ll leave these matters for discussion afterward. But I do want to record my conviction that political and cultural considerations about socialism rival in importance the purely economic considerations. It does seem futile to worry about whether a socialist government could manage the economy efficiently and in accord with people’s wishes, when the more immediate question is whether powerful rulers could at all times be forced to want to rule in the interests of all the people.

It seems to me that the history of socialist literature is a history of continual attempts to get rid of the difficulties in socialist blueprints. And as these difficulties are eliminated, socialism comes to look more and more like competitive enterprise. Therefore, it is not without some justification that Professor Mises, with his characteristic dogmatism, now claims final victory in the debate over the efficiency of socialism. In his latest book Mises writes:

It is ... nothing short of a full acknowledgment of the correctness and irrefutability of the economists’ analysis and devastating critique of the socialists’ plans that the intellectual leaders of socialism are now busy designing schemes for a socialist system in which the market, market prices for the factors of production, and catallactic competition are to be preserved. The overwhelmingly rapid triumph of the demonstration that no economic calculation is possible under a socialist system is without precedent indeed in the history of human thought. The socialists cannot help admitting their crushing final defeat. They no longer claim that socialism is matchlessly superior to capitalism because it brushes away markets, market prices, and competition. On the contrary. They are now eager to justify socialism by pointing out that it is possible to preserve these institutions even under socialism. They are drafting outlines for a socialism in which there are prices and competition.

What these neosocialists suggest is really paradoxical. They want to abolish private control of the means of production, market exchange, market prices, and competition. But at the same time they want to organize the socialist Utopia in such a way that people could act as if these things were still present. They want people to play market as children play war, railroad, or school. They do not comprehend how such childish play differs from the real thing it tries to imitate. (1949, pp. 702—703; footnote omitted)

Perhaps the culmination of the free-enterprise trend in socialist thinking is James E. Meade’s Planning and the Price Mechanism (1948). Meade calls himself a socialist, but his proposals amount to scarcely more than a program for the reconstruction of competitive capitalism, including a moderate amount of government enterprise.

The implication of my remarks should be clear by now. If economists are bent on seeing socialism substituted for capitalism, they should leave the task to their colleagues who are better qualified for the job, namely, to the lexicographers.

REFERENCES

Barone, Enrico. “The Ministry of Production in the Collectivist State.” 1908. In Hayek 1935.

Beckwith, Burnham P. The Economic Theory of a Socialist Economy. Stanford, Calif. : Stanford University Press, 1949.

Brutzkus, Boris. Economic Planning in Soviet Russia. London: Routledge &Sons, 1935.

Coase, R.H. “The Marginal Cost Controversy.” Economica (August 1946): 169—182.

Dickinson, H.D. The Economics of Socialism. London: Oxford University Press, 1939.

Dobb, Maurice. Political Economy and Capitalism. London: Routledge, 1946.

Hall, R.L. The Economic System in a Socialist State. London: Macmillan, 1937.

Harper, F.A. Liberty. Irvington-on-Hudson, N.Y.: Foundation for Economic Education, 1949.

———. Individualism and Economic Order. Chicago: University of Chicago Press, 1948.

Hayek, F.A., ed. Collectivist Economic Planning. London: Routledge & Kegan Paul, 1935.

Hotelling, Harold. “The General Welfare in Relation to Problems of Taxation and of Railway and Utility Rates.” Econometrica (July 1938): 242—269.

Landauer, Carl. Theory of National Economic Planning. Berkeley and Los Angeles: University of California Press, 1944.

Lange, Oskar, and Fred M. Taylor. On the Economic Theory of Socialism. Edited by Benjamin E. Lippincott. Minneapolis: University of Minnesota Press, 1938.

Lerner, Abba. The Economics of Control. New York: Macmillan, 1947.

Lerner, Abba, and Oskar Lange. The American Way of Business. Washington: National Education Association, 1944.

Marshall, Alfred. Principles of Economics. 1890. 8^th ed. London: Macmillan, 1920.

Meade, James E. Planning and the Price Mechanism. London: Allen & Unwin, 1948.

Mill, John Stuart. On Liberty and Other Essays. 1859. New York: Book League of America, 1929.

Mises, Ludwig von. “Economic Calculation in the Socialist Commonwealth.” 1920. In Hayek 1935.

———. Human Action. New Haven, Conn.: Yale University Press, 1949.

———. Socialism: An Economic and Sociological Analysis. 1922. Translated by J. Kahane. Indianapolis: Liberty Fund, 1981.

Modigliani, Franco. “Lorganizzazione e la direzione della produzione in un’economia socialista.” Giornale degli Economisti (September/October 1947): 441—514.

Norris, Harry. “State Enterprise Price and Output Policy and the Problem of Cost Imputation.” Economica (February 1947): 54—62.

Pierson, Nicolaas G. “The Problem of Value in the Socialist Community.” 1902. In Collectivist Economic Planning, edited by FA. Hayek. London: Routledge & Kegan Paul, 1935.

Pigou, A.C. Socialism versus Capitalism. London: Macmillan, 1937.

Robbins, Lionel. The Great Depression. London: Macmillan, 1934.

Samuelson, Paul A. Foundations of Economic Analysis. Cambridge, Mass.: Harvard University Press, 1948.

Sweezy, Paul. Socialism. New York: McGraw-Hill, 1949.

Vickrey, William. “Some Objections to Marginal-Cost Pricing.” journal of Political Economy (June 1948): 218—238.

---

*Presented in the Department of Economics staff seminar, Texas A&M University (then College), November 1949. Printed here unchanged except for standardization of the format of references.

¹In a footnote Landauer cites the work of Henry Schultz to show that actual demand curves can be found.

In the book quoted, Landauer, an old-time socialist, is ostensibly arguing for “planning” rather than for “socialism.” However, it would be mere quibbling about the use of words to deny that Landauer ‘s proposals are socialism.