Must Economies Be Rational?
"Is Socialism Really ‘Impossible’?"
Critical Review 16, no. 10 (2004)
Debate over Mises’s socialist calculation argument has been going on since 1920, and one might have thought that at this late date, it would be difficult to say something new. Bryan Caplan has done exactly that. He contends that although Mises showed that a socialist economy could not rationally allocate resources, this lacks the importance that Mises attributed to it.
Even if socialist calculation is impossible, Caplan says, it does not at all follow that socialism is impossible. "My [Caplan’s] thesis is that Mises, and the Austrian-school economists who have echoed his argument, lack any sound reasons for the extreme claim that socialism is ‘impossible’" (p. 35).
But does Mises in fact claim that socialism is impossible? Caplan fails to cite any use of just this word by Mises, but I think that he is in substance right. As Caplan points out, Mises thought that the calculation argument enables opponents of socialism to go beyond the argument from incentives. To the claim that socialism lowers productivity because of its malign effect on incentives, there was a response available, albeit one that few socialists would have the hardihood to make. A convinced egalitarian might say that even though socialism would lower the average standard of living, the goal of equality makes some degree of privation worth enduring.
As Mises saw matters, even a socialist prepared to "tough it out" could not treat the calculation argument in the same way. The inability of a socialist system to calculate would render production chaotic in a developed modern economy. No one but a misanthrope, intent on the destruction of humanity, could rationally choose socialism.
Here Caplan rises to object. He grants, and indeed explains very well, the substance of Mises’s case: a socialist economy cannot calculate. "If the state owns all of the capital goods, Mises reasoned, there will be no market for capital goods. With no market for capital goods, there will be no capital-goods prices. And without prices, there will be no numbers to run so as to determine the cheapest way to do things" (p. 34).
But how serious is this problem? Is this not a quantitative matter? "Just because less calculation leads to some degree of economic chaos does not imply that no calculation leads to complete economic chaos" (p. 39).
An analogy will clarify Caplan’s contention. A well-known argument shows that, other things being equal, a law that mandates wages above what employers find profitable to pay will cause unemployment. It does not follow from this, though, that a minimum wage law will always generate massive unemployment. The severity of the law’s effects depends on the particular circumstances of the case. Exactly the same is true, Caplan avers, for the calculation argument.
Caplan finds an ally, if an unwilling one, for his view in Mises himself. Does not Mises insist that praxeology cannot arrive at quantitative laws? How then can Mises claim to have proved that absence of economic calculation leads to chaos? Is this not just the sort of quantitative law he repudiated as impossible?
I cannot think that Caplan has accurately understood Mises. When Mises rejects quantitative laws, he has in mind laws that involve constant units. "[T]he impracticability of measurement is not due to the lack of technical methods for the establishment of measure. It is due to absence of constant relations" (p. 38, quoting Mises).
It hardly follows from this that Mises rejects all concepts that can be described, in rough terms, as more or less. When Austrians speak of a "depression" or a "cluster of entrepreneurial errors," are they violating some supposed methodological stricture by using quantitative concepts? Not at all: such terms of ordinary language do not assume the constant units that Mises rejected.
In like fashion, Mises does not violate his own rule by speaking of economic chaos. He would do so only if he claimed that there were precise units of chaos. If Mises had said, e.g., that each unit of increased socialization injects 1.5 units of chaos into the economy, he would be vulnerable to Caplan’s point. But of course he does not do so.
Caplan falls into error because he assumes that economic chaos must be a precisely quantified concept. Thus, he notes that Mises admits that in his simple economy Robinson Crusoe could make production decisions without monetary calculation. If so, Caplan asks, where do we draw the line? "Does Crusoe’s one-man socialism ‘completely break down’ when Friday shows up? What if a dozen people joined their isolated collective? . . . Mises has boxed himself in. Eventually he needs to draw a line. . . . But in drawing such a line he would violate his own strictures against quantitative economics" (p. 39). But why need we draw such a line? Do we not know chaos when we see it?
The point at issue extends far beyond economics. Someone with absolutely no hair on his head is bald. He is still bald if he has only one or two hairs. If we keep adding hairs, at some point the person will no longer be bald. But exactly where is the line to be drawn? Ordinary language offers no answer; but philosophical paradoxes created by this sort of question do not prevent us from judging that Michael Jordan is bald and that Bryan Caplan is not.
If Mises has not lapsed into methodological sin, though, it does not follow that he is right. How do we know that lack of calculation leads to chaos? But is not the answer obvious? How can an economic system function at more than a primitive level if it lacks a means to determine how to allocate resources efficiently? Caplan’s question seems analogous to the demand for a proof that social cooperation is more productive than an "economy" of isolated people.
Caplan disagrees, but I suggest that his failure to see the obvious stems from a radical misconception of the state of affairs Mises has in mind. He notes that some socialist countries have built large plants in complete disregard of what the calculations of efficiency require. However inefficient such plants, they nevertheless are able to function. Where is the chaos? Further, Mises himself admits that in cases where external costs are present, economic calculation is misleading. With obvious excitement, Caplan claims: "In fact, Mises raises doubts about the ability of economic theory to state unambiguously that the quantitative effect of calculation on prosperity is positive!" (p. 50).
But Mises’s problem differs entirely from the situations Caplan has mentioned. The calculation problem arises when there are no prices at all in an economy. If a centrally planned economy has access to world prices, the fact that it may disregard efficiency considerations in some instances is not the point. The case of external costs may be dealt with in exactly the same way. Is it not better to have somewhat misleading costs in some instances, so long as these exist within a price system, than to try to get by without any calculation at all?
Caplan has one final argument against Mises. Is it not "conceivable" that other factors will outweigh the problems caused by economic calculation? However severe these problems may be, perhaps technological progress will allow a socialist economy to function reasonably well. To this, I can only say that if Caplan wishes to conjure up a Utopia in which allocating resources efficiently is no longer needed, he is free to do so.
Much of Caplan’s article is devoted to the claim that the incentive problem, rather than difficulties of economic calculation, led to the collapse of Russian communism. As he himself notes, Mises did not claim that a socialist economy that could rely on capitalist world prices must be chaotic. Even if Caplan’s historical claim is correct, then, it leaves Mises’s argument untouched. n MR
Deirdre McCloskey has criticized economists for drawing quantitative conclusions from arguments that merely show that some phenomenon exists. See her The Secret Sins of Economics (Prickly Paradigm Press, 2002), pp. 42 ff.
Mises mentions such paradoxes in the essay "Profit and Loss."
As Peter Boettke and Peter Leeson note in their excellent Working Paper, "Socialism: Still Impossible After All these Years," the calculation argument does help explain the collapse of War Communism in 1921, a fact that was not lost on the contemporary Russian economist Boris Brutzkus. I have benefited greatly from this paper but do not agree with these authors’ view of what Mises meant by "impossibility."