Introductory comments,The following essay was written in an attempt to understand the economic arguments presupposed by Hayek's The Road to Serfdom. My then-complete ignorance of economic theory, and the fact that its main influences are the texts in footnote 2 (though I didn't find the Rothbard article until after finishing my paper), may explain why the paper exhibits such an uncritical acceptance of the new emerging consensus about the Mises-Hayek argument. The recent defenders of the Mises-Hayek argument claim that it was misinterpreted and underestimated in the 1930's because economists didn't realize how deeply the arguments undermined the consensus of the era. I agree, but now suspect the same is true of his defenders, who have accepted his premises and welcomed his conclusion, without cross-checking them against his methodological assumptions. When we do, it is not clear that his whole argument hangs together.
With the old misunderstandings out of the way, Hayek's argument now provokes two main reactions: supporters adopt the argument and use it as another stick to beat socialism, critics admit that the argument has scored a serious blow against centrally planned economies, but (typically) offer tu quoque responses: competitive markets aren't so good either: What about pollution? What about monopolies? What about bad advertising? A less hackneyed critique might reevaluate the idea that `prices are signals' in light of the Austrians' most characteristic claims:1) methodological individualism: the claim that, whatever else we say about complex social phenomena, they must be explainable in terms of individual actions; 2) subjectivism - the emphasis on the subjective origin of value, and the importance of idiosyncratic differences in beliefs and knowledge; 3) no "perfect competition" - a deep suspicion about the practical relevance of neoclassical analysis of the conditions of perfect competition.Consider 3: perfect competition, like `frictionless surface' in mechanics, is an ideal type, useful for illustrating the forces underlying the complex phenomena in question. When seen for what they are, these assumptions are harmless and helpful; but the Austrians are right to reject them as the basis for practical policy proposals. However, Mises' claim that "prices are signals", while certainly true even in the absence of perfect competition, becomes problematic: during imperfect competition, the quality of the price signal degrades: delays, misinformation, changing preferences, incomplete markets, etc. all make the utility of the price signal suspect. What does this do to Hayek's `markets are conveyors of information' claim? Could it be that the plausibility of Hayek's anticentraliztion argument depends on reading it against a background of perfect competition assumptions that Hayek rejects?
Or consider subjectivism. If we take seriously the subjectivity of values and the tacitness of some kinds of knowledge, then it is no longer clear if or how we can use the concepts of `cost' and `efficiency'. Both involve reference to values, but when (to borrow the title of one of Hayek's key articles) we view "Competition as a Discovery Procedure" which both reveals and creates knowledge and preferences, then ex ante judgments about cost in terms of "foregone alternatives" may become impossible.The Problem
HAYEK ON SOCIALIST CALCULATION
This paper is about Friedrich Hayek's argument that centrally-planned economies are necessarily such inefficient users of essentially dispersed knowledge that they will be far less efficient than market economies at producing the goods necessary to support a large population.
This argument interests me for three reasons. First, Hayek's account of the market as a medium of knowledge places epistemology on the ground-floor of political philosophy. Thus, unlike common criticisms of socialism (e.g. Nozick's), Hayek's argument does not turn on the alleged immorality of non-capitalist means, but on what he claims are their unavoidably disastrous consequences. Second, it has important political consequences: if correct, Hayek's argument shows that traditional socialist programmes won't work. Third, despite its important insights and clear exposition, the standard response to Hayek's argument, where it is recognized at all, has been exactly the opposite of what it should be. In one way its reception is the most interesting thing about Hayek's arguments: if this is the best political and economic analysts can do, then why bother?
Before discussing this there's much stage-setting to do. The paper's main purpose is to say what Hayek's argument was and put it in a context which explains why it was so widely misunderstood. The paper has four main sections; the first sketches standard economic analyses of capitalism, the next discusses socialist criticisms of capitalism, the third is on the Socialist Calculation Debate, and the last raises some interesting and disquieting questions about the possibility and relevance of ex ante political philosophy.
In brief, I argue that the standard neoclassical and socialist market analyses are based on absurdly unrealistic assumptions, and that by drawing attention to some of these assumptions, Hayek has identified a fatal flaw in non-market socialism and continuing difficulty with "market socialism".
A problem faced by every group of people is the coordination of individuals with different but partially overlapping ends in a world where individuals can't get everything they want. One typical formulation asks: What is the best way to reconcile limited altruism with scarcity? I'll call this (contentiously, no doubt): The Problem.
One solution to the conflicts arising from rivalry for scarce goods would be for rivals to selflessly surrender their claims to scarce goods. Traditionally, however, scarcity has been the focus of the conflict between limited altruism and scarcity. There are several reasons for this: pessimism about the malleability of human nature, moral concerns about reengineering human desires, and the optimism about reducing scarcity created by great increases in production since the Industrial Revolution.
The Socialist Calculation Debate (SCD) concerns the viability of one means of solving The Problem: can an economy work when the means of production are socially owned and centrally directed? Historically, the SCD occurred in the context of classical and neo-classical analyses of capitalism. These analyses are important for three reasons: 1) they form the theoretical basis of the socialists' economic assumptions, 2) Hayek's criticism of centrally-planned economies is equally a criticism of neo-classical economics (NCE), and 3) the dominance of NCE helps explain why Hayek's criticisms were misunderstood and not addressed by either socialist or NCE. So, before proceeding to Hayek's argument, a brief summary of classical and neo-classical economic theory.
[SECTION ON HISTORY OF ECONOMIC THEORY DELETED. BASIC POINT: Adam Smith's Invisible Hand account says that markets partially solve The Problem by using self-interest to motivate people to make and do what others want, with mutually beneficial results. The result, in effect if not intention, is self-interested altruism. There are three main unintended economic consequences of the competition for profits: supply tends to equal demand, price tends toward cost, and cost tends to its minimum. The bulk of mainstream economics is the ever-more sophisticated analysis of these three trends.]
The Socialist Calculation Debate
Of course, greater production and the tendency of price to approach cost aren't the only consequences of competition. Advocates and critics both claim that economic arrangements are more than just socially neutral means. Advocates appeal to the utilitarian virtues of greater consumption, and to the indirect social benefits from rewarding virtues like diligence, effort and alertness, as capitalism is alleged to do. Critics have many targets too. The standard criticism of utilitarianism applies: general happiness may be increased at the expense of individuals' happiness. Because the market does not necessarily help people, but consumers, profits will be made while needs go unmet.
The details of a socialist economy follow from its political goals. Broadly speaking, socialism opposes the competition and unequal distribution of capitalism with the goals of solidarity and egalitarianism. Additionally, socialist economists have been influenced by Marx's charges that capitalism is inherently alienating and exploitive. Unlike socialism's production for use (in which the ends being served are recognized by all), capitalism's production for exchange (in which commodities are produced for purposes their creators need neither know nor approve) causes two forms of alienation: workers are alienated from both the products of their labour, which are not owned by them and may even harm them, and from their labour time, which is directed by and for the benefit of their capitalist bosses. Capitalism exploits workers because it pays them less than their labour is worth. According to Marx's labour theory of value, the value of a commodity is the amount of socially necessary labour time `congealed' in it or required to reproduce it. Capitalists profit because the labour time embodied in the commodity is greater than the labour power they pay for, and they keep this "surplus value" from commodities which rightfully belongs to the workers.
So, a good socialist economy minimizes alienation and exploitation, while promoting solidarity and equality. How is this best done while still addressing The Problem? Marx himself was nearly silent on the organization of a post-capitalist society. His few remarks on it, in "Critique of the Gotha Programme", say that communism requires three conditions: an end to the division of labour, labour has become the workers' medium of free self-expression, and productive capacity has grown so much that scarcity is not a serious problem (1875:531). Given that the division of labour is one of the chief means of increasing production, Marx was clearly hoping for a technological solution to the problem of scarcity. In the meantime, what is to be done?
Horribly, this `planning vacuum' was filled by the rhetoric of economic ignoramuses like Lenin. Here is one of Lenin's typical accounts of how to run an economy:it is quite possible, after the overthrow of the capitalists and the bureaucrats, to proceed immediately, overnight, to replace them in the control over production and distribution, in the work of keeping account of labor and products, by the armed workers, by the whole of the armed population. (The question of control and accounting should not be confused with the question of the scientifically trained staff of engineers, agronomists, and so on. These gentlemen are working today in obedience to the wishes of the capitalists and will work even better tomorrow in obedience to the wishes of the armed workers.)What could be easier? All that's needed are cashiers. Has Lenin mistaken competitive markets for self-stocking super-markets? This would be laughable if it hadn't contributed to the immiseration of millions. The Socialist Calculation Debate concerns a more intelligent proposal.
Accounting and control that is mainly what is needed for the "smooth working", for the proper functioning, of the first phase of communist society. All citizens are transformed into hired employees of the state, which consists of the armed workers. All citizens become employees and workers of a single countrywide state "syndicate". All that is required is that they should work equally, do their proper share of work, and get equal pay. The accounting and control necessary for this have been simplified by capitalism to the utmost and reduced to the extraordinarily simple operations which any literate person can perform of supervising and recording, knowledge of the four rules of arithmetic, and issuing appropriate receipts. The Economic Basis of the Withering Away of the State (Ch.4)
An Objection: Ends before Means
First, I want to address one preemptive objection. Studying means may seem to put the cart before the horse. Instead of debating means, shouldn't we first decide what sort of ends are best? A typical way to put it goes like this:reasoning about means is merely instrumental reasoning, and that can't tell us anything about which ends we should prefer; but all actions presuppose choices made with ends in mind, so questions about ends (say, justice) are conceptually prior to questions about means. (i.e. Justice before economics)I think there are several reasons this is mistaken. Most importantly for this paper, it simply assumes the ends in question are achievable. That's usually a fair assumption, but in cases where there are no means to an end, it makes no difference how desirable that end is. This is simply modus tollens from the premise that "ought implies can":
If ought, then can;
therefore `not ought'.
We may, of course, choose to debate such ends as thought experiments or just for the fun of it. But if there is a compelling case for "not ought", then, among other benefits, we will know that debating the value of those ends really is just a mental exercise -- nothing rides on it.
Hayek's Critique of CPB-Socialism
The Socialist Calculation Debate (SCD) concerns the viability of one means of solving The Problem: social ownership of the means of production directed by central planning. If feasible, replacing competition by planning would meet socialists' goals by not only avoiding exploitation and alienation and the unequal distribution of goods caused by competition, but could also produce more goods by avoiding the inefficiencies of uncoordinated competition.
The standard account of the SCD found in most economic textbooks goes like this:
- 1908, Enrico Barone showed that central planners' equilibrium equations are formally similar to the market mechanism;
- 1920, Ludwig von Mises argued that rational economic decisions are impossible in a socialist state;
- 1935 Hayek edited a volume which repeated and buttressed this criticism of socialism;
- 1936, Oskar Lange decisively refutes Mises and Hayek;
- 1940's, Hayek retreats, arguing that rational planning is not logically impossible, but only "practically impossible";
- 1968, Hayek decisively refuted again by Lange, who argued that the practical difficulty of calculation is eliminated by computers.
Below, I'll argue that this account is mistaken in almost every detail.
Barone's 1908 article is peculiar: both sides cite it as support. Socialists say that Mises and Hayek overlooked Barone's "proof" that the formal similarity between markets and a Central PlanningBoard (CPB) means that whatever the market can do the CPB can do too. Hayek (1985:58-59) claims that Barone showed that there is a formal similarity, but denied that the CPB could perform their part. Who's right?
Barone's article makes three claims relevant to the SCD: 1) a CPB's equilibrium equations could simulate a perfect market mechanism; 2) these equations would require: a) "gigantic" but not impossible work to solve (§56), and b) very difficult but "not inconceivable" (§56) collecting of the information about goods and needs; but 3) "it is frankly inconceivable" that the CPB could set up the equations without experimentally testing "on a very large scale" whether their production functions are the right ones (§57), which would require production to be substantially similar to market production (§58). In short: Barone said it is not impossible to set up and solve equilibrium equations, but without market-like competition it is impossible to rationally pick which equations to set up.
Mises deserves more attention because he originates the economic calculation debate, and his claim that prices convey information essential to rational decision-making is central to Hayek's critique.
Before 1920, most socialists followed Marx in not developing accounts of socialist economics. Because so little had been written about them, Mises's critique was directed at measures consistent with Marx's ideals: moneyless economies, the labour theory of value, and the abolition of private property. By concentrating on these basics, Mises's arguments achieved a generality which allowed them to anticipate (and "pre-refute") many later socialist economic blueprints.
Mises concluded that "Socialism is the abolition of rational economy" (1935:110), because "Where there is no free market, there is no pricing mechanism; without a pricing mechanism, there is no economic calculation" (1935:111). Mises's argument is based on three main claims: rational choices require commensurable scales, money is the only possible scale for rational calculation of all relevant factors, private property is necessary individuals to make and act on rational plans.
For Mises, rational decisions attempt to satisfy the (individual's) most important needs given the options available. When options are both few and discrete then rational decision-making is fairly easy. But economic decisions, particularly those involving production factors, are rarely so straight-forward:Picture the building of a new railroad. Should it be built at all, and if so, which out of a number of conceivable roads should be built?... Where one cannot express hours of labour, iron, coal, all kinds of building materials, machines and other things necessary for the construction and upkeep of the railroad in a common unit it is not possible to make calculations at all. (1935:108)Mises's first main claim is that rational, non-arbitrary, decisions require common units for comparing different options. This is a really important point, so I'm going to repeat it using the best short simple summary of it I've found:Our problem is this. It has been decided to make useful products which we will call `widgets'. We are informed by our engineering and design experts --our technologists-- that there are three different ways of making widgets, three technical processes of production or recipes, which I will call methods A, B, and C. All three methods require inputs of rubber and wood, but in different quantities. To make one widget:Method A requires 5 lbs of rubber and 5 lbs of wood.Instead of rubber and wood, we could use bricks and cement, or paper and plastic, or --for the mathematically gifted-- x and y.... we have set up a problem where the only basis for choice between production processes is the different quantities of rubber and wood used up in production... we at once perceive that both Method B and Method C must be better than Method A...
Method B requires 5 lbs of rubber and 4 lbs of wood.
Method C requires 4 lbs of rubber and 5 lbs of wood.
So we confidently eliminate Method A. We now have to choose between B and C. There seems to be no sensible reason for preferring B over C, or C over B. B uses up less wood than C, but then, B uses up more rubber than C. We can state our problem like this: We know that A costs more than either B or C, but we have no way of telling whether B costs more than C, or C costs more than B, or (a third possibility) whether B and C cost the same. (Steele, 1992:5-6)
Pop Quiz, hot-shot: B or C? In the absence of a common scale on which to compare them, won't any choice between B and C be arbitrary?
Mises's third point is that price is a guide to use-value only when there is exchange based on private ownership: "the crucial point. Exchange relations between production-goods can only be established on the basis of private ownership of the means of production" (1935:112). I am less clear about Mises's reasoning here, but here are two possibilities. First, prices are indices of exchange, and without private property there is no exchange. Full stop. The second possibility is more convoluted: successful action requires that we take account of other's interests in our decisions affecting them --- not because this is a moral or considerate way to act (though it is), but because people will act to promote their interests and this may upset our own plans; second, people's interests are nowhere more clearly revealed (to themselves as well as others) than in their decisions regarding private property; so learning about others' interests is best facilitated by a system of private property.
Mises's thesis that socialist calculations are irrational follows from these three claims: "monetary calculation... affords us a guide through the oppressive plenitude of economic potentialities" (1935:101), and "The human mind cannot orient itself properly among the bewildering mass of intermediate products and potentialities of production without such aid. It would simply stand perplexed before the problems of management and location" (1935:103). But in a socialist economy, "because no production-good will ever become the object of exchange, it will be impossible to determine its monetary value... Calculation in terms of money will here be impossible." (1935:92), so that "we have the spectacle of a socialist economic order floundering in the ocean of possible and conceivable economic combinations without the compass of economic calculation."
Oskar Lange was a Polish economist and one of the originators (with H.Dickinson, F.Taylor and A.Lerner) of "market socialism". In 1936-37, Lange published two articles which attempted to rebut Mises's criticism that rational economic calculation is impossible where there is no market for production factors. Lange apparently accepted the need for money and prices, but thought both could be had without private ownership of the means of production.
There are three main groups in Lange's economy: workers, managers, and the central planning board (CPB). First, the CPB announces tentative prices of all commodities. Next, within each industry and every factory, production managers choose how and how much to produce. Since they can't, by definition, aim to maximize profit, Lange offers two rules managers must follow when guesstimating their input and output: use production factors which minimize cost, and produce that quantity at which marginal cost equals the price set by the CPB (1938:94-95). After following these rules, managers submit their estimates to the CPB, which tallies all estimates and resets all prices, lowering the price on goods which do not sell, and raising the price on those which do, with prices eventually set so that all markets clear. Meanwhile, individuals have a free choice of employment and consumption. It is assumed there will be no conflict between these individual choices and managerial and CPB decisions.
Lange believed that his system could not only avoid three of capitalism's main defects --greatly unequal wealth, negative externalities, and stagnation due to monopoly-- but also that it could duplicate capitalism's good features, and even out-perform the original:The market process with its cumbersome tatonnements appears old-fashioned. Indeed, it may be considered as a computing device of the pre-electronic era. (1967:402)Lange's system is clearly designed with NCE equilibrium assumptions in mind. In fact, the whole process is a slower, paper-driven version of Walras' `auctioneering' account of markets, with the CPB as the auctioneer and managers replacing consumers as the bidders:I refuted the Hayek-Robbins argument by showing how a market mechanism could be established in a socialist economy which would lead to the solution of the simultaneous equations by means of an empirical procedure of trial and error. Starting with an arbitrary set of prices, the price is raised whenever demand exceeds supply and lowered whenever the opposite is the case. Through such a process of tatonnements, first described by Walras, the final equilibrium prices are gradually reached. These are the prices satisfying the system of simultaneous equations. It was assumed without question that the tatonnement process in fact converges to the system of equilibrium prices. (Lange, 1967:401)This is a peculiar boast, since not only is it a hallmark of Austrian economics to deny the empirical content of static equilibrium models, but Mises's original article explicitly said so:The static state can dispense with economic calculation... But this is only conceptually possible... a static state is impossible in real life, as our economic data are for ever changing, so that the static nature of economic activity is only a theoretical assumption corresponding to no real state of affairs (1935:109)Static equilibrium analysis is a common part of neo-classical economics, and it is heuristically useful because it helps reveal otherwise obscure market tendencies. But real markets are never static nor in equilibrium: innovations, accidents and changing desires keep markets in constant flux. This doesn't vitiate equilibrium analysis, but it's a serious problem for Lange, because with no equilibrium there are no equilibrium equations, and so there are no paint-by-the-numbers decisions for Lange's mangers (Steele 1992:148). If managers can't base their decisions on static equilibrium reasoning, Lange's model and his conclusion collapse.
Conclusion: Why This Matters
Hayek's main arguments follow Mises: there is a need for prices which accurately represent agents' interests, and Lange's CPB-created prices do not do this. Hayek offers a host of other objections to Lange's proposal, among them: there are no incentives for managers to perform well when they can own no special share in their business; there are no capital markets, and therefore no market prices indicating our best guesses about future demands; judging the quantities of unsold consumer-goods can tell us nothing about which components are the most valuable; consumers have too limited an effect on CPB decisions about what to produce when profit is not considered; if the CPB lowers prices when goods do not clear, and managers are required to produce goods "even if it should involve plants or whole industries in losses" (1938:96), then what tells the CPB to take something off the market?
Hayek agrees with Mises that CPB socialism can't make rational economic decisions without money, markets, and private property; but he adds a further point: centrally planned economies can't efficiently coordinate knowledge. Hayek's main argument is that centrally planned economies are necessarily such inefficient users of essentially dispersed knowledge that they will be far less efficient producers of goods necessary to support a large population. Hayek is offering what is essentially a transcendental argument: he claims that the efficient generation and coordination of knowledge are necessary conditions for the success of large economies, and that this entails that every economic system is subject to epistemic constraints which drastically limit the extent to which such systems can be intentionally guided; because centrally planned economies do not meet these constraints they will fail.
When any criticism of socialism --except the `doesn't-go-far-enough' variety-- is sufficiently "reactionary" to provoke a spate of unflattering name-calling, a few preliminary comments are in order. First, this is not a critique of socialist ideals, motives or goals. Rather, as Hayek says:
socialists are wrong about the facts... I am prepared to admit that if socialist analyses of the operation of the existing economic order, and of possible alternatives, were factually correct, we might be obliged to ensure that the distribution of incomes conform to certain moral principles, and that this distribution might be possible only by giving a central authority the power to direct the use of available resources, and might presuppose the abolition of individual ownership of means of production. ...however... there is no known way, other than by the distribution of products in a competitive market, to inform individuals in what direction their several efforts must aim so as to contribute as much as possible to the total product.Hayek has published moral arguments against socialist ideals, but this isn't one of them. Hayek's contribution to the SCD is to argue that there are no known means to achieve socialist goals, however worthy.
however... those who demand a deliberate arrangement of human interaction by central authority based on collective command over available resources [make] a factual error... about how knowledge of these resources is and can be generated and utilized. (1988:6-7)
Second, obviously, Hayek neither shows nor attempts to show that capitalism is without fault. Capitalism may also suffer from a self-defeating fatal flaw (though it's not this one). As such, this is a defense of market economies, but not more: Hayek is not a laissez faire liberal.
Two Kinds of Subjectivism: Value and Knowledge
In 1871, Carl Menger and William Jevons independently proposed the law of diminishing marginal utility. It is based on an explicitly subjectivist account of value. Where classical economists had argued that value is an objectively determined quality of a good, ultimately dependant upon labour or land, Menger and Jevons said that value depends on utility, and that utility is best understood as `desiredness'. Utility is not an intrinsic quality of things, but "a circumstance of things arising out of their relation to man's requirements" (Jevons 43); thus, a commodity's utility is relative to the degree it serves one's purposes. If value is relative to purpose, then it is relative to all the things purposes depend upon, among them: individual needs, desires, and circumstances (including, but not limited to, quantities of goods already owned). Radical though that may have been, it is only part of the Austrians' subjectivist revolution, the second half was waiting for Hayek.
Desire is only one part of intentional explanation, equally important is belief. What Hayek called his "most original contribution" to economics (1994:79) was his emphasis on the subjective nature of belief and knowledge. The subjective nature of desire is obvious: with the possible exception of needs for food and shelter, human desires vary as much as individuals. The subjective nature of knowledge is harder to see. There are several reasons for this. First, the paradigm case of knowledge is knowledge that can be expressed in propositions (e.g. "5+7=12"), which leads to an overemphasis on book knowledge over tacit knowledge or `knowing that' rather than `knowing how'. Second, on a naive epistemology, knowledge is `justified true belief' and truth is one; so knowledge is readily (but fallaciously) assumed to be the same for everyone. Third, the overemphasis on the general over the particular reinforces the disregard of individual knowledge of particular facts, circumstances and opportunities. Fourth, while truth is important, "most of the objects of social or human action are not `objective facts' in the special narrow sense in which this term is used in the Sciences... So far as human actions are concerned the things are what the acting people think they are" (1979:44). When combined with the NCE assumption of perfect knowledge, it is easy to see why economists overlooked the many subjective differences in belief. But these features of knowledge and beliefs are important for understanding one of the market's main virtues: the coordination of knowledge.
This is important because, briefly, the division of labour applies to decision-making as well as more overtly physical forms of labour;
there is here a problem of the division of knowledge, which is quite analogous to, and at least as important as, the problem of the division of labor. But, while the latter has been one of the main subjects of investigation ever since the beginning of our science, the former has been as completely neglected, although it seems to me to be the really central problem of economics as a social science. (1948:50)
By adopting Mises's view of prices as conveyors of information about countless exchanges, and analyzing these exchanges in terms of the subjective desires and essentially dispersed knowledge which cause them, Hayek developed a view of the market as a kind of information system which does what no individual mind could ever do, upon which:
[t]he whole market process then became understood as a process of transfer of information enabling men to use, and to put to work, much more information and skill than they would have access to individually (Hayek, 1988:97).
It is Hayek's claim that markets make fuller use than centrally planned economies of the division of knowledge by diffusing the authority to make decisions and the power to act upon them (1988:77). This is worth stating because the market's role as an efficient user of diffuse knowledge is obscured by the NCE assumptions about perfect competition and perfect knowledge:
The assumption of a perfect market, then, means nothing less than that all the members of the community, even if they are not supposed to be strictly omniscient, are at least supposed to know automatically all that is relevant for their decisions. It seems that the skeleton in our cupboard, the "economic man," whom we have exorcised with prayer and fasting, has returned through the back door in the form of a quasi-omniscient individual. (1948:46)
Is the criticism right?
Hayek's focus on knowledge has two big advantages over other anti-planning arguments. First, while there is some desperate plausibility to the hope that benevolent and wise planners could know everyone's interests (either through personal contact in a small society, or for all via a True Theory of Human Nature), there is zero plausibility to the suggestion that managers and central planners could know, communicate, and efficiently and correctly evaluate as much knowledge as decentralized markets process. Second, Hayek's argument avoids tempting counter-objections to structurally similar anti-centralization arguments. For instance, unlike raw materials like ores and electricity, whose quality and quantity can be roughly but reliably extrapolated from current levels, new knowledge appears unpredictably --as Popper says, "we cannot anticipate today what we shall know only tomorrow"-- thereby thwarting any kind of planning.
I think Hayek is clearly right to claim that the best centrally planned economy cannot be as efficient a generator and coordinator of knowledge as the best market economy. The reasons for this are by now familiar: nobody's omniscient; people know different things and so the more people contributing to market-shaping decision-making the more information is taken into account; individuals can use but not communicate tacit knowledge, so each step up the decision-making hierarchy of a planned economy has access to less of the relevant knowledge available at lower levels.
However, if we compare imperfect markets with partially planned economies the difference will not be so clear. Here we might want an estimate of the degree of inefficiency of a CPB economy. But Hayek cannot provide this. He would claim that this is not a defect of his arguments, but a consequence of their subject matter: individual components of complex orders cannot be analyzed and judged in isolation from the rest of the system, because the nature of a complex order limits the possibility (and maybe the intelligibility) of more than approximate and qualitative judgments of the contribution of that element to the whole. Hayek recognizes this as a consequence of the complexity of the economic order:most of the phenomena in which we are interested, such as competition, could not occur at all unless the number of distinct elements involved were fairly large, and that the overall pattern that will form itself is determined by the significantly different behaviour of the different individuals so that the obstacle of obtaining the relevant data cannot be overcome by treating them as members of a statistical collective.
For this reason economic theory is confined to describing kinds of patterns which will appear if certain general conditions are satisfied (1967:35)
If this is correct (and I think it is), then economics cannot make the kind of quantitative estimates of degrees of inefficiency that we would need in order to have an air-tight argument for or against planning.
But if that's true, then how could Hayek ever show, as he sometimes claims he has, that a centrally planned economy is "logically impossible" (1988:7)? Strictly speaking, there is no inconsistency in the idea of an efficient centrally-planned economy. We misunderstand Hayek's argument --and he sometimes overstates its force-- if we hold that he is arguing that central planning is logically impossible. Hayek's argument turns on the impossibility of human omniscience, and this is not a logical but a practical impossibility.
If this is a "practical impossibility" then it must be decided by practical judgement:The practical problem is not whether a particular method would eventually lead to a hypothetical equilibrium, but which method will secure the more rapid and complete adjustment to the daily changing conditions in different places and different industries. How great the difference in this respect would be between a method where prices are currently agreed upon by the parties of the market, and a method where these prices are decreed from above is, of course, a matter of practical judgment. But I find it difficult to believe that anybody would doubt that in this respect the inferiority of the second method would be very great indeed. (1948:118)So the real question is whether we can imagine a centrally-planned economy working well without magic. If by "working well" we mean to include (as do all the authors discussed above) consumer satisfaction over a wide range of goods, then I agree with Hayek that the answer is "no": the more the CPB exercises its power the less responsive its producers can be to consumer demand, and the more responsive they are the less planning they can do. However, practical judgments can and do differ.
One reaction to this essay might be: `Of course. This is all old-news, nobody argues for centrally-planned economies anymore.'
There are two reasons I think this reaction would be a mistake. First: people do argue for centrally directed economies and we ignore these arguments at our peril. For instance, real inefficiencies in the market's use of resources, including knowledge, foster the belief that properly located experts could make economic decisions which avoid these mistakes. But this misses half the point of Hayek's epistemological analysis: no subaltern intelligence can have the knowledge necessary to make impeccable economic decisions because much of the information required is both necessarily diffused and a product of the economic process. Moral: ignorance-based market "failures" are inherent and ineliminable aspects of the market.
The second reason this `of course' response is mistaken is that, in fact, it wasn't of course, and the very fact that economists and socialists ignored and misunderstood the epistemological critique is itself important. First, it should raise disturbing questions about the objectivity and credibility, and hence the existence of (academic) political economy. Consider this: in ending a 1991 lecture explaining why he now believes market socialism can't work, Janos Kornai, an exception to the NCE consensus, spoke about why he changed his mind about these issues:Thinking is strongly affected by metarational factors: values, sentiments, prejudices, and hopes. These act like gates, or at least like filters...
I read Mises and Hayek thirty years ago and rejected their objections to market socialism. Later I read them again in a different frame of mind, and suddenly I became receptive to their arguments. The resistance was gone...
What changed many of our minds was a series of political traumas and disillusionments. With professional experts like economists, the decisive blow was not dealt in many cases by negative experiences in their own areas of competence.... First, the foundations of their philosophy of life collapsed, usually under the influence of some earthshaking event: the sight of Russian tanks in Budapest, Prague, or Afghanistan, or the experiences a friend relates on being released from prison. Once this enlightenment has happened... as a result of psychologically searing experience, the mind immediately opens to the rational arguments as well... Works whose ideas had bounced off the walls of prejudice suddenly appear convincing. (Kornai 1993:62)
I don't doubt that this is how Kornai came to change his mind; but we should ask ourselves: if the right arguments, clearly stated and timely put and addressed to experts and laypersons alike, did not carry the debate, then why bother doing ex ante political or economic analysis? It is depressing to think about how much wasted debate (and human suffering) would have been avoided if Kornai et al had come to this conclusion when Hayek and Mises did. Why didn't they? Meta-rational factors are one reason. Kornai suggests another:The warnings of a Mises or a Hayek about market socialism are brilliant guesses, but they are not scientifically proved ex ante. A vision was confronted by a guess, not a scientific proposition by a scientific repudiation of it. An ex post position has now been reached... economists of countries where an experiment was made in applying market socialism are now in a position to make statements based on firsthand experience. (Kornai 1993:63)In short: the Mises-Hayek argument wasn't a scientific proof, but now we have the experience which confirms their guesses. Quite apart from whether "scientific proof" is an appropriate standard for social studies, there's an obvious question begged by Kornai's reference to "firsthand experience". The most charitable thing I can say about this is that its admirable commitment to empirical observation is undermined by its selective memory of socialist economic experience: what socialists euphemistically call the "events of 1989" were by no means the first evidence of socialist economic failure. As Friedrich Hayek, then himself a recently disillusioned ex-socialist, wrote sixty years ago:The breakdown of "war communism" occurred for exactly the same reasons, the impossibility of rational calculation in a moneyless economy, which Professors Mises and Brutzkus had foreseen. The development since, with its repeated reversals of policy, has only shown that the rulers of Russia had to learn by experience all the obstacles which a systematic analysis of the problem had revealed... Officially the blame for nearly all the difficulties is still put on the unfortunate individuals who are persecuted for obstructing the plan by not obeying the orders of the central authority or by carrying them out too literally. But, although this means that the authorities only admit the obvious difficulty of making people follow out the plan loyally, there can be no doubt that the more serious disappointments are really due to the inherent difficulties of any central planning. (1935:206)This brings us full-circle. The questions "Why weren't economists and socialists convinced by the Mises-Hayek arguments?" and "Why does this matter?" both concern the division of knowledge. Central planners from the 1920's to today need to be convinced because, like most human beings, they don't learn from other people's experience. The arguments failed to persuade because --as we intellectuals (reenacting the hubris that is the main moral of this story) know but hate to admit-- as an antidote for false hopes, arguments are a weak substitute for harsh experience. The socialist calculation debate matters because our best chance for avoiding these mistakes is to understand the ideas that failed to ward it off the first time.
Copyright © 1997 John Upper