Chapter XII

CHAPTER XII OTHER SOLUTIONS, MARGINAL COSTS AS CRITERIA THE EconomicJournalof December, 1936, carried an article by E. F. M. Durbin, "Economic Calculus in a Planned Economy", in which the author attempted to show that there was no theoretical or logical difficulty in having an exact pricing system in a planned economy. In his opinion there was no reason to abandon the search for economic or utilitarian calculus in an industrial system controlled by a Central Authority. He agreed with Professors Robbins and Hayek that the mathematical solutions and any system involving simultaneous equations were impracticable, but asserted that he could point to an alternative system, thus refuting the arguments advanced by Mises, Halm and I-Iayek in CollectivistEconomic Planning. Dr. Durbin based his argument on "the well-established fact of economic scarcity". The consumers must be consulted about what they want "unless democratic doctrine is rejected root and branch". If there is no wastage of resources, the costs of one commodity must be assessed in quantities of sacrificed alternative goods. "There must, therefore, also be a free market for the factors of production".

Now, says Dr. Durbin, it is the best known truism of the theory of value, that perfect competition, including perfect foresight, will secure the right adjustment of production to the preferences of consumers. 1 This basic doctrine has been advanced in three separate expositional forms of very different value for solving the problem of a planned economy. There are 11arshallian supply and demand curves, the Austrian School's solution by way of marginal products, and the equational systems. Dr. Durbin considers that it is "almost certain" that the .second (the marginal product method) can be equally well used in a planned society, as in one of laissezfaire. \Vere the central authority to instruct the various producing units (I) to calculate the marginal productivity of all mobile resources and (2) to move all mobile resources to the positions of highest calculated product, there would-according to Dr. Durbin-seem to be no reason to suppose that the distribution of resources would be any different in such a community than it would be in one with complete competition, "since all logical, theoretical and accountancy problems are the same in both systems".

IDr. Durbin adds in a footnote that he has expressed his doubts about the validity of this doctrine and refers the reader to his article, "The Social Significance of the· Theory of Value", in the EconomicJournal, 1935. 144 MARGINAL COSTS AS ·THE CRITERIA 145 Dr. Durbin grants that it might be very difficult to calculate the marginal products, but the technical difficulties will be the same in both societies. There are problems that can only be solved after· a comprehensive sociological and psychological analysis, but in the realm of economic theory the marginal productivity solution would seem adequate. The ability to discover marginal products does not depend upon the existence of a particular set of social institutions and certainly not on those of "IaiJJcz1airc",he thinks. On the other hand, Dr. Durbin is not satisfied with a solution exclusively based on marginal products. He admits that the estimates of marginal value products are extremely liable to error. Demand curves will have to be calculated. In either calculation the margin of error is likely to be great, and in the resulting value product doubly so.

There can be no complete solution of the problem, unless the process can be submitted to another check and assessment. This raises the question of applicability of the English cost analysis to the problems of the planned economy. Then Dr.. Durbin discusses the position of a socialized trust and assumes that there will be competition between trusts, in which connection he mentions a rate of interest, established in the free market, for new capital. He also discusses the possibility of inaccuracies occurring and what a trust could do, if demand sinks. The treatise also criticizes Professor Hayek's views on the question of depreciation in CollectivistBcono.wit Planningand here the author of this book is largely in agreement with Dr. Durbin. Dr. Durbin concludes by saying that in his opinion he has refuted the main charge offered by economists against a planned economy, that it has no method for the rational calculation of relative produc tions. He thinks the real arguments for and against a planned economy lie elsewhere. He does not specify where, but thinks that not economic theory, but only social science, can hope to find an answer to them.

Many of Dr. Durbin's comments are interesting and some of them help to throw light on the problem, but it is difficult to agree with his assertion that he has solved the problem. He has managed to maintain· the discussion on a high theoretical level, but in his anxiety not to "dogmatize on practical questions" he has overlooked the crux of the whole problem, namely, how the data on which the socialist trusts are to base their calculations are to be obtained. It is all very well to say that competition and market prices are preferable to mathematical computation of thousands of questions, but that is no solution. The great questions are: to what extent can market prices be obtained in a socialist community, and how far are 146 ECONOMIC CALCULATION IN THE SOCIALIST SOCIETY competing trusts compatible with the central authority which Dr. Durbin assumes to exist? On these questions he does not touch. He speaks of "interest rates, established in the free market, for new capital", but never says who is to offer this new capital. If he means that it is to come from the savings of the members of the community, this would involve a number of questions that ought to have been discussed (see Chapter VIII). If, however, he means that the new capital is to be supplied by the central authority and that the various trusts shall compete for it, this, too, raises interesting problems that must be solved before one can accept his assumption of a free interest market.

Nor does Dr. Durbin say exactly whether it is vertical or horizontal trusts he has in mind. Even under the assumption that there will be markets with several buyers and sellers of all intermediary products, there will be only one supplier of natural resources (obtained from the soil) and the fixing of prices in such a market would be as imperfect, as though the capital was merely supplied by the central authority. When assuming the simultaneous existence· of a planning authority and competing trusts, as Dr. Durbin does, one has at least the right to expect an indication of the extent of the trusts' autonomy and of whether they will be able to expand or diminish their activities, even though this ran counter to the plan, or not. Had Dr. Durbin embarked on a discussion of these points, he would presumably have seen that his solution gets us no further, not even on paper. The free right of disposition and competition in order to achieve maximum marginal productivity, production according to a central plan, and production according to the consumers' preferences, are three things that cannot be combined. Dr. Karl Polanyi, whose attempted solution, to which we shall return, also assumes the existence of competition, fully realizes the implications of such a proposal and says straight out that calculation is out of the question in a community directed by a central authority.

Dr. Durbin's idea, that special institutions are not required to determine marginal products, at any rate not those found in a laissezfaire society, is interesting, and it is to be regretted that he neither pursues it, nor gives his reasons. Perhaps he thinks that it is not within the scope of a treatise on theoretical economics to give reasons, but in that case the opinion should never have been expressed. Dr. Durbin's tendency to advance views on matters that he considers to lie outside the scope of the economist, appears again in his conclusion where he has beautiful things to say of the "better and juster type of society" called a Planned Economy, but declines to justify his views. In actual fact, much of Dr. Durbin's treatise does not touch on our problem: the possibility of economic calculation in a socialist MARGINAL COSTS AS THE CRITERIA 147 society, but on an extraneous and in itself interesting problem, namely, the production policy that autonomous monopolistic trusts in such a society ought to conduct.

One solution that commands respect is that of Dr. A. P. Lerner, who has made many important contributions to the discussion. His criticism of the solutions of others has at times been so severe that one might almost think that he himself doubted the possibility of economic calculation in the socialist community; but this is not .the case. His' views were, however, somewhat difficult to grasp and it was not till Dr. Maurice Dobb accused him of being an "elusive" and "invisible opponent" that he made them clearer. In his Economic Theory and Socialist Economy: A ReplY he treats first the difficulties encountered in other solutions: "Where there are thousands of products and thousands of factors, being combined in thousands of different ways in millions of different productive units, and where a reshuffling of factors may have to be of a most complicated kind, it seems to me that it would not be so easy to find the technical expert who knows all that is going on everywhere."

and he continues: "If every producer so regulates his production as to make the marginal productivities of factors proportional to their prices on a market, and if the prices are moved so as to equate the producers' demands to the supply, the problem can be solved without waiting for the super-technician."l ·One must agree with what Dr. Lerner says here, but it is difficult to see how this represents a solution of the problem. True enough, with both markets and prices there is no need for the super-technic~an, but it is just the assumed absence of markets and prices that has called forth the proposals for mathematical solutions, which in their turn call for the technical expert who knows everything that is going on everywhere. With both markets and prices the problem ceases to be a problem. The question is: How are these to be obtained? Of that Dr. Lerner has nothing to say. Considering that Dr. Lerner with wit and acuteness has demanded clarity in the discussion, and considering the light he himself has shed on the problem, one hesitates to think that he could have overlooked the importance of how markets and prices are to be achieved in the IOn the preceding page he writes: "And by a price system I do mean a price ~ystem. Not a mere a posteriori juggling with figures by auditors, but prices which will have to be taken into consideration by managers of factories in organizing production."

148 ECONOMIC CALCULATION IN THE SOCIALIST SOCIETY socialist society. Unfortunately, one's doubts are not removed, but strengthened by a later article of his,1 criticizing Dr. Durbin's article. He himself describes it as a protest against "the developing tradition, in approaching the problems of socialist economics, of starting from the consideration of competitive equilibrium, instead of going direct to the more fundamental principle of marginal opportunity cost. This approach is not only subject to methodological criticism as indirect and cumbersome, but is a fertile source of actual error deriving from unrealized implications of the static nature of competitive equilibrium." There is no doubt that the method is cumbersome, that it makes mistakes possible, and that it is natural to assume static conditions when discussing competitive equilibrium, but it is difficult to see how one can avoid consideration of underlying conditions and difficulties, if the task is to elucidate the problem before us.

Dr. Lerner's treatise is first and foremost a criticism of Dr. Durbin's solution. 2 In his opinion it is more practical than Dickinson's and "also refutes anew the well-known thesis of Professors Mises, Hayek and Halm that a socialist economic calculus is impossible." This "also" is rather surprising in view of Dr. Lerner's almost deadly criticism of Dickinson's solution. There are many valuable points of view in Dr. Lerner's treatise, but it does not mention what, in our opinion, is the greatest weakness in Dr. Durbin's proposal, namely, the lack of any indication of how the necessary data for calculation are to be obtained. This is where Dr. Lerner's own solution, too, falls down. As far as Dr. Lerner is concerned, the matter is simple. He writes: "If we so order the economic activity of the society that no commodity is produced unless its importance is greater than that of the alternative that is sacrificed, we shall have completely achieved the ideal that the economic calculus of a socialist state sets before itself."

This immediately and obviously raises the question of how the "importance" of products is to be determined. Later Dr. Lerner writes: l"Statics and Dynamics in Socialist Economics", in Economic fournal, June, 1937. 2Dr. Durbin has himself in "A Note on Mr. Lerner's 'Dynamical' Propositions", Economic fournal, September, 1947, accepted some of Dr. Lerner's objections and refuted others. He also repeats that he and Dr. Lerner agree on the most important thing of all, "the applicability of the logic of the theory of value to the circumstances of the Planned Economy".

MARGINAL COSTS AS THE CRITERIA 149 "We must therefore aim directlY at our real object, the most economicutilization of resources. What does this imply? . . . If we assume that the members of the society, in spending their income, do not take. into account the effect of their individual purchases on the prices of consumption goods, we can take the ratio between the prices at which goods sell freely on the market as measuring the ratio between the marginal significance of the commodities. This is because every individual, in using his income to the best advantage, will purchase more significant shillingworths. . . until all the shillingworths have . . . the same marginal social significance. " This means that the central authority is to take its indications of the importance of products from a free consumers' market. If that is the case, then one ought to go into the question of whether alterations in prices are to be made by the central authority or by autonomous retail shops, and to clarify the conditions under which "ratios of prices"

can be said to give reliable indications. Dr. Lerner says in a footnote: "This assumes that individuals choose best for themselves. Whenever this is not considered to be the case, others-normally in the form of the State-can choose for them either wholly or partly (i.e. by influencing particular prices by taxes or bounties). These others are then the consumers and the whole scheme formally remains the same." There are two reasons why this is a noteworthy statement. First, because Dr. Lerner had earlier said that to let the central authority determine the scales of needs would be irreconcilable with democracy and socialism (see footnote on p. 71). In the second place, because -and this is the main point-this is of decisive importance for assess ing the possibility of calculation. Dr. Lerner has rightly said that the question of the extent of the free choice of goods and occupation the community shall have is not one that can be settled by an economist as such. This, however, does not mean that it is not of great economic consequence and of essential importance in solving the problem of calculation, which of these alternatives is chosen. This will be dis cussed in greater detail in Chapter XIV, but it can be said here that the valuation of capital goods (that is, the obtaining of the necessary cost data) is another-and easier-task, when the valuation is to be made by the same individual (or individuals) who determine '\vhat the members of the community are to have, than when the valuation is made by a different method. The problem is quite different, if pro duction is, for example, to be adapted to (and the ,means of production I 50 ECONOMIC CALCULATION IN THE SOCIALIST SOCIETY valued according to) the constantly shifting needs of the. members of the community. It is thus wrong in principle to discuss the question on the basis of both assumptions simultaneously.

That Dr. Lerner assumes the community to be dynamic is a step forward. He calls his treatise Statics and Dynamicsin SocialistEconomics and points out that his hypothesis includes variable marginal costs, at the same time taking Dr. Durbin to task because the relationship between short-period and long-period problems "is left beautifully vague in the simplified directives given by Mr. Durbin to the managers." He considers that Dr. Durbin's expression "in the fullness of time", which is the only clue he gives of when and how quickly a new factory is to be built, "must surely be the nearest English expression for the untrans latable word by which the citizens of such countries as Russia and Spain show their refusal to recognize time as an economic factor." As the community is assumed to be dynamic, it would have been natural to discuss the effect on demand of alterations in, for example, the size, age-composition and tastes of the population. This is not the only point on which Dr. Lerner, to use his own expression, is "beautifully vague". He is vague, too, in regard to data on costs.

Besides the general directions, which have already been quoted, he employs two rules. The second, which he mentions first, "suggests that if all the officers of the economic administration equalize their marginal revenues to their marginal cost-and this is what they would do if each is simply enjoined to maximise the profits of the enterprise under his control-.,--this will suffice to set in motion all the forces necessary to achieve the equilibrium. We may call this rule, suggested by the second method, Rule Two .... If the application of Rule Two results in a deviation from this norm, the officers may be instructed to subordinate Rule Two to another rule, which we may call Rule One, derived from the first method and calling for the equalization of price to average cost." His final word is: "Price must be made equal to marginal cost. This is the contribution that pure economic theory has to make to the building up of a socialist economy."

Of the definition of marginal costs and marginal productivity Dr. Lerner writes,· that instructions should be issued that: MARGINAL COSTS AS THE CRITERIA I 5I "the use of every factor is to be extended up to the point where the marginal physical product multiplied by its price is equal to the price of the factor. Or, in other words, up to the point where the price of the factor needed to produce another unit of product multiplied by the product is equal to the physical quantity of any price of the factor. This value, which has to be equated to the price of the product, we shall call the marginal cost." And in a footnote to this extract he writes: "The concentration on the price of a factor is achieved under conditions of perfect competition because then the price happens to be equal to the extra cost of buying one unit more. This has very aptly been called the" parametric function of prices under perfect competition. See O. Lange, 'On the Economic; Studies of Socialism, Part One,' Review of Economic Studies, October, 1936, p. 59. With our rule we need not rely on the conditions of perfect competition being present, and we are not upset if, because of the growth of the unit of production .or for any other reason, the parametric function of prices breaks down."

The views expressed in this footnote will be discussed later, but it must be pointed' out that Dr. Lerner, although using the term "marginal. physical product") is really operating with prices and values. In one place he says that marginal costs result from multiplying with the price and elsewhere he says that in evaluating the greater or lesser marginal (physical) productivity of the various factors, the price of the factors must be taken into account. Apart from the obvious objection that Dr. Lerner omits to say how he will arrive at these prices, his argument can also be criticized for circular reasoning. As Professor Fritz Machlup wrote in "On the Meaning of the Marginal Product": 1 "By measuring units of factors in terms of their market value, marginal productivity analysis is, to my mind, reduced adabsurdum. One must bear in mind that marginal productivity analysis as a part of the theory of distribution is to serve as explanation of the market values of factors or services. To define these services in terms of their market values is to give up the task of explaining them."

It is also of importance to point out that" all forms of marginal consideration assume the possibility of choice and of a selection of goods, and that the more imperfect the markets are, the more valueless lSee Exploration in Economics. Notes and Essays Contributedin Honour of F. W. Taussig, New York, 1936.

I 52 ECONOMIC CALCULATION IN THE SOCIALIST SOCIETY become the decisions taken. This is a question that Dr. Lerner might well have discussed, but has not.! Even though we accept, as Dr. Lerner does, the importance of supply and demand and the importance of relating prices to costs, it is obvious that his directions do not solve the problem of economic calculation in the socialist community. In his positive solution he moves within a narrow field, where he-without even stating it assumes all the relevant data to be given, both those needed for determining costs and those required for determining demand. In other words, he assumes that the most difficult part of the problem has already been solved. lIt is no elucidation that Dr. Lerner has elsewhere written: ". . . the competitive price system has to be adapted to a socialist society. If it is applied in toto we have not a socialist but a competitive society". See "Economic Theory and Socialist Economy", in Review of Economic Studies, 1934-35, p. 55.