In this post I will critique Ludwig von Mises' classic polemic against socialist planning, Economic Calculation in the Socialist Commonwealth, published in 1920. In this text Mises introduces the Economic Calculation Problem (ECP). I will quote segments of the text and highlight what effects 100 years of advance in the field of computation has had on the Mises' arguments. A copy of the text is available here from archive.org.
I have broken down this critique on a chapter-by-chapter basis.
1. The Distribution of Consumption Goods in the Socialist Commonwealth
In this chapter Mises correctly identifies that "who is to do the consuming and what is to be consumed by each is the crux of the problem of socialist distribution". These two problems correspond to remuneration and measuring/predicting demand.
It is characteristic of socialism that the distribution of consumption goods must be independent of the question of production and of its economic conditions.
This statement is wrong, partly because distribution of both final and intermediate goods happens on the same transportation network. In addition, what final goods that can even be offered depends on the state of the productive forces. Thirdly, there is an overlap between the two groups of goods. Flour is both a consumer good, and a means of producing bread in a bakery. Fuel is both a consumer good, and a vitally important resource in the distribution process.
For, as we shall show, it lies in the very nature of socialist production that the shares of the particular factors of production in the national dividend cannot be ascertained, and that it is impossible in fact to gauge the relationship between expenditure and income.
This is equivalent to saying technological coefficients cannot be measured. We know this to be false.
Then comes a few sections implying that the State decides how goods are to be distributed, and that remuneration would be done with ration coupons, based on some a priori distribution. To Mises' credit it is not out of the question that certain goods might be subject to rationing, in particular goods that are scarce and in high demand, for example fossil fuel. But for all other goods there is no need to do this. In Cockshott and Cottrell's model for example, an estimate of final demand is computed based on sales in publicly run stores. If the estimate is wrong then sales data provide feedback to the planning system to increase or decrease production of some final good. A small excess can also be explicitly planned for, say 1-5%, on the grounds that overproduction is not as bad as underproduction.
Mises is correct that calculating the price of production goods in terms of money is impossible in a socialist system. This is a consequence of calculation in kind.
Further in Mises points to the limited fungibility of labour, among other things asserting that:
Yet one cannot allow the laborer who had put in an hour of the most simple type of labor to be entitled to the product of an hour’s higher type of labor.
But such "higher" labour is itself the product of labour, namely teaching and studying. A medical professional will have absorbed more labour before they begin working than the average worker will have absorbed. Therefore medical workers are more valuable. This doesn't necessarily mean medical workers would be paid more, only that it costs society more to train them. We could choose to pay medical workers more, but we we could also choose to pay people for studying medicine.
I also don't subscribe to the notion of there being "higher" and "lower" forms of labour. There are merely different forms, all of which require different amounts of training to perform. Proficiency is a vector value, not a scalar.
Mises ends the chapter with the following:
For, over and above the actual labor, the production of all economic goods entails also the cost of materials. An article in which more raw material is used can never be reckoned of equal value with one in which less is used.
I wonder whether Mises had actually read any of the classical economists before writing this, since none of them suggest that one can ignore the cost of raw materials. Maybe Mises real argument is getting lost in translation?
It is very easy to imagine two goods that have the same value, are made of the same raw material but one uses more of it than the other. One example of this is copper ingots vs copper tubing. Ingots are easier to produce than tubing, and so have less social labour embodied per kilogram than does tubing. Therefore a mass X of ingots and Y of tubing, where X > Y, may indeed have the same value.
If Mises is talking about equivalent goods here, say copper ingots from two different foundries, then of course equal amounts have equal value (unless you get your copper ingots from Ea-nasir).
2. The Nature of Economic Calculation
Here Mises confuses use-value with value, a trait he shares with neoclassical economists. It is obvious that different people have different wants, that use-value is in the eye of the beholder. Mises then points out that it is exchange-value that is useful in trade, not use-value. This is of course correct. Mises then says that because market relations play out a certain way, then that way is "the appropriate employment of goods". An obvious counter to this is to point out that short-sighted employment of goods like petroleum are most definitely not "appropriate". On top of this there is the problem of market manipulation.
After a comparison to the fictional Robinson Crusoe, Mises makes this claim:
That [taking into consideration the intersubstitutability of goods] is only possible in very simple conditions is obvious. In the case of more complicated and more lengthy processes of production it will, plainly, not answer.
Anyone who wishes to make calculations in regard to a complicated process of production will immediately notice whether he has worked more economically than others or not; if he finds, from reference to the exchange relations obtaining in the market, that he will not be able to produce profitably, this shows that others understand how to make a better use of the goods of higher order in question.
This is an admission by Mises that he doesn't know how to compute the effects on the economy of using some new productive process. But as I have shown previously here, computing the material effects of new technology is relatively straightforward. Mises on the other hand must resort to comparing exchange values, and claims that the result of this is always the best use of resources.
Mises makes the case that all economic activity is in fact exchange (circulation), in stark constrast to the classical economists, especially Marx. By making circulation central, rather than production, or the combination of production and circulation, Mises again shares traits with the neoclassicals.
Mises does make this very good point:
No single man can ever master all the possibilities of production, innumerable as they are, as to be in a position to make straightway evident judgments of value without the aid of some system of computation.
In other words, the sheer number technical coefficients combined with the limitations of the human brain lead to a problem. To the modern reader there is a simple solution: use computers.
Mises makes a second useful observation: monetary calculations can be dispensed with in household economies. The reasoning for this is that household economies are small enough that a single human can understand them. But loyal readers should know that we can offload this process to computers. This leads to a realization, for which I thank Mises: socialism can be understood as a worldwide household economy! Planning can be viewed as "householding" (hushålla med, being thrifty with) our limited resources.
Mises also goes into a quaint argument against in natura planning (calculation in kind), stating that it is impossible to compute how much intermediate goods need to be produced in order to arrive at some set of final goods. This is Mises saying that solving Leontief's (I - A)x = d is impossible, an argument that was invalid even in his day, and one that has been rendered completely moot by modern advances in computing.
One amusing thing about this chapter is that it was written prior to the establishment of Gosplan which, despite its many faults, refuted Mises' claim one year later, using pen, paper and mechanical adding machines. That's not to say that Gosplan was optimal, even by 1920's standards, only that it refuted Mises' notion that planning couldn't work.
3. Economic Calculation in the Socialist Commonwealth
Here Mises indirectly makes the point that private enterprises are planned internally, that each branch of an interprise is a household economy. Mises again claims that exchange is necessary for a rational economy, for economic calculation to take place. As my readers should now, this is not actually the case.
Mises does raise a good point in this chapter:
When the “coal syndicate” provides the “iron syndicate” with coal, no price can be formed, except when both syndicates are the owners of the means of production employed in their business. This would not be socialization but workers’ capitalism and syndicalism.
This is not a uniquely right-wing critique. The same point is raised by Amadeo Bordiga here, especially in the section The Economic Function:
To clarify our point, let us imagine that the organisation of bread production, and of all other wheat-based products, is entrusted to the “Bakers’ Union”, with analogous arrangements for all other trades and industries. [...] These organisations would need to make decisions about how to utilise the entire product (in our example: bread, pasta etc,) in such a way as to receive, from other parallel organisations, not only what their members require for their personal consumption, but new raw materials, instruments of labour, etc, as well. Such an economy is an exchange economy.
[At the higher level of the organisation] the system of exchange remains a mercantile one, that is, it requires some law of equivalence in order to equate the value of the stocks of one syndicate with another; and we can easily suppose that these syndicates would be very numerous, and just as easily suppose that each of them would need to separately negotiate with all the others.
[...] Let us be so “liberal” as to think it possible that the various equivalent values could be peacefully determined through a spontaneously arrived at equilibrium. A measuring system of such complexity couldn't operate without the age-old expedient of a general equivalent, in other words, money, the logical measure of every exchange.
What I want said with these two quotes is that it is impossible to do away with money unless all productive forces fall under the same umbrella, that for economic purposes, they are all part of the one single commune. If they're not then exchange must be used, and different communes/syndicates/firms/whatever must haggle with each other, production will be haphazard and mediocre, labour and resources will be wasted and so on. I may go into more detail on this point in a future post, but in short segmenting the economy in this way means keeping technical coefficients secret, which can be shown to lead to crises similar to a capitalist economy.
Next Mises makes another confused argument around value, stating that two commodities P and Q that both embody 10 hours social labour where P needs two units of raw material a but Q only needs one unit of a, that P is more valuable than Q. This is the same argument that I replied to using the copper analogy earlier.
Then is another argument about the limited fungibility of labour, which has already been addressed. For us the fact that some people are adept at different things, that only a certain fraction of the population have the traits necessary to become say a surgeon, is merely another set of constraints to be satisfied.
In the last section of this chapter Mises claims that the labour theory of value is not indispensible to socialism. This might be true, if unlikely. Either way this doesn't mean social labour isn't a useful thing to optimize on, and unlike Mises I claim that we could in fact optimize on anything. One example of non-labour optimization in planning is the proposal of some technocrats to optimize on energy. In discussions with local technocrats I've conceded that we could choose to optimize on say a linear combination of labour and energy. I just happen to believe that minimizing labour is easier to "sell" to people, since most people want more free time. This is a point I've alluded to in previous posts, most explicitly here:
Once computed, it is in everyone's interest to work according to the plan, since doing so means plan goals are met with a minimum of labour.
It is also the case that the set of feasible plans is much larger than the set of labour-minimizing or energy-minimizing plans.
4. Responsibility and Initiative in Communal Concerns
In this chapter Mises claims that socialists are in favor of a system where workers are renumerated regardless of the amount of social labour performed, something that has never been the case, at least in the Marxist camp. This passage is an example of this claim:
The obvious objection that the individual is very little concerned whether he himself is diligent and enthusiastic, and that it is of greater moment to him that everybody else should be, is either completely ignored or is insufficiently dealt with by them. They believe they can construct a socialist commonwealth on the basis of the Categorical Imperative alone.
Mises is however correct, that merely hoping that people do the right thing, is utopian. He also touches on the problem of initiative and innovation, which admittedly was a problem in the USSR, if I understand its history correctly. This is not an insurmountable problem however.
5. The Most Recent Socialist Doctrines and the Problem of Economic Calculation
Here Mises points out that the nationalization and amalgamation of all banks into a single central bank, coupled with abolishing markets, means that the central bank seizes to be a bank. This is correct, and this critique is not dissimilar to Marx' critique of the Owenites' labour money. It also seems that Mises anticipates the ad-hoc nature of Gosplan, which I think is also a fair critique.
Mises also has a bit of fun here poking at Lenin's move towards the New Economic Policy (NEP). Mises also points out that Lenin does not properly distinguish between monetary calculation and calculation in kind. This is also a correct observation, as far as I can tell.
In the conclusion Mises reiterates that only private property and exchange can be economically rational. This supposed rationality stands against the wasteful use of resources in the capitalist mode of production. Fossil fuel use is again an excellent example of this, along with the bourgeois states dragging their feet on the environmental issue.
Mises again returns to exchange as being useful for people to gauge their relative desire for certain goods, but conveniently ignores that whoever has the most economic muscle has the most to say in exchange relations. References to petty consumption of alcohol or tobacco ignores the reality of the class system, the immense autocratic power held by the haute bourgeoisie.
Postscript: Why a Socialist Economy is “Impossible”
There is a postscript by Joseph T. Salerno in this version of the text. Among reiterations of the points in the main text, Salerno is upset at F. A. Hayek for framing Mises' arguments in a way that they may be refuted.
One thing near the end of the postscripts stands out in particular:
This is true, for example, of environmental regulations that prohibit development activities for the vast majority of Alaskan land and along much of the California coastline as well as of recent calls for suppressing development of Amazon rain forest and coercively maintaining the entire continent of Antarctica forever wild. Needless to say, thoroughgoing and centralized land use regulations, which some fanatical environmentalists are calling for, is tantamount to the abolition of private property in national resources and business structures.
Free market apologists like Salerno offer no solution to the environmental question. They do not suggest any way we might "household" with our limited resources, any way that we might prevent a runaway greenhouse effect. Therefore the solution is exactly what Salerno fears: the abolition of private property and the global planning of land use. To keep free market economics going at this point is at best utopian and childish, and at worse deliberately genocidal.