Monetary policy and the knowledge problem.

• Author: O'Driscoll, Gerald P., Jr.

The curious task of economics is to demonstrate to men how little they really know about what they imagine they can design.

--F. A. Hayek (1988: 76)

The knowledge problem in economics is most closely associated with Friedrich Hayek, who articulated it in analyzing the role of prices in markets, the socialist calculation debate, and monetary policy. In tins article, I summarize and apply Hayek's analysis to contemporary monetary policy debates. I also connect Hayek's work with that of Milton Friedman, who articulated his own version of the knowledge problem in his monetary work. Friedman's views in this area are underappreciated.

Hayek argued that knowledge is inherently dispersed and localized across the population of economic agents. It is not possible to assemble the totality of knowledge existing in society in any one mind or place. Individuals may reveal their localized knowledge by their actions, but only if incentivized. Moreover, knowledge is often tacit and cannot be articulated. What the totality of individuals knows far exceeds what any policymaker can know, no matter his or her expertise and wisdom.

Hayek on the Knowledge Problem

Hayek presented his analysis of the knowledge problem in society in the course of what came to be known as the socialist calculation debate (Hayek [1935] 1975a). (1) He developed the argument further in a series of lectures and articles in the 1930s and 1940s. They were made more accessible by being reprinted in one place (Hayek 1948).

Early supporters of socialism supposed they could dispense with economic problems. Some thought of societal resource allocation as an engineering problem. Hayek (1975a: 5) pointed out that engineering problems involve a singleness of purpose. Resource allocation involves competing uses of resources, and accounting for opportunity costs is necessary (Hayek 1975a: 6-7). What is being introduced here is the basic knowledge problem of a diversity of actors with different preferences. The difficulty for socialists wanting to abolish private property and markets was "how in the absence of a pricing system the value of different goods was to be determined" (Hayek 1975a: 27). He supported Mises' argument that only a system of money prices, including for factors of production, could produce a rational solution to resource allocation. That system necessitated private property, including in capital goods.

Hayek's essays bookended the 1935 volume. His concluding chapter dealt with various responses to Mises' original critique. These were the "alternative socialist systems which differ more or less fundamentally from the traditional types against which the criticism was directed" (Hayek 1975a: 202). The debate took many twists and turns, but I will follow the knowledge argument. Hayek reiterated that knowledge is dispersed, and for any variant of central planning, it must somehow become concentrated in one mind or those of a very few experts. Much knowledge is not preexisting, but it is created only in the process of adapting to change (Hayek 1975a: 210-11). (2) "Every passing whim of the consumer is likely to upset the carefully worked out plans" of the central planning authority (Hayek 1975a: 214).

Hayek (1975a: 227) emphasized the importance of disruptive change and the dynamic nature of the economic problem. Economists ignored the importance of change because of "the excessive preoccupation with the conditions of the hypothetical state of stationary equilibrium" (Hayek 1975a: 226). In response to suggestions for marginal cost pricing in a socialist state, Hayek (1975a: 229) argued that "the competitive or necessary cost cannot be known unless there is competition." What are the assumed "givens" or data of an economic model are information to be discovered in the real world in a competitive process. That is the knowledge problem in a nutshell.

The socialist calculation debate ended inconclusively for many economists, "a draw" as Caldwell (2004: 338) phrased it. But Hayek continued working on the knowledge problem. "Economics and Knowledge," reprinted in Hayek (1948), was originally a 1936 lecture, published in Economica in 1937. It thus follows temporally on the socialist calculation debate, though I will soon connect it logically to a different part of Hayek's work.

First, Hayek (1948: 34) linked the equilibrium concept to having correct foresight. He also observed that the concept of equilibrium has "a clear meaning only when confined to the analysis of the action of a single person." Analyzing the interactions of different individuals introduces "a new element of altogether different character" (Hayek 1948: 35). Actions for a single individual are in equilibrium only if they are part of "one plan" (Hayek 1948: 36). We can speak of a state of equilibrium for a moment in time only if the plans of all individuals are "mutually compatible." Any change in data would upset such equilibrium (Hayek 1948: 41).

In this seminal article, Hayek introduced the concept of equilibrium as plan coordination. To achieve it, the economist must assume perfect foresight or introduce assumptions about learning (Hayek 1948: 42, 45-46). Both moves are fraught with difficulties. Equally important as his equilibrium concept was Hayek's analysis of the "division of knowledge," the fact that the knowledge needed to achieve equilibrium is dispersed among all the individuals in the economy. Equilibrium could be brought about "by deliberate direction only by somebody who possessed the combined knowledge of all those individuals" (Hayek 1948: 51).

I agree with Caldwell (2004: 337) that "Economics and Knowledge" is a "key document," but I don't agree with him that it "marked a change in the direction" of Hayek's thinking. Caldwell elaborates by saying that "Hayek came to have doubts about the ability of static equilibrium theory to capture certain essential features of a free market economy." But Hayek harbored such doubts long before this article, and the article is neither the first nor even the most complete statement of Hayek's views on equilibrium over time. Those descriptions belong to a 1933 lecture, "Price Expectations, Monetary Disturbances, and Malinvestments."

Hayek delivered that lecture in Copenhagen. It was published in German in 1935 and next in French in 1935. It was not available to English speakers until translated and included in Hayek ([1939] 1975b). So Hayek's 1937 article was novel at the time for an English audience. (3) But modern researchers should not be confused by the chronology. (4)

Hayek (1975b: 137-38) began by clarifying drat he was explaining "dynamic phenomena"--that is, processes that take place over time. He focused on the problem of economic fluctuations, and explored the self-reversing character of expansions financed by money creation. Intertemporal equilibrium entails correspondence between the multiperiod spending plans of consumers with "the separate and independent decisions" of entrepreneurs to supply the desired consumption goods at all relevant future dates. Correspondence among all these plans is what is entailed when economists say that "savings are equal to investments" and there is "an equilibrium rate of interest" (Hayek 1975b: 153-54).

Playek's analysis presented a challenge for all who would extend the concept of the equilibrium of an individual, or static equilibrium, to dynamic equilibrium in a world of monetary disturbances and changing expectations. The essay's title reflected the focus on monetary policy. If the achievement of intertemporal equilibrium is a challenge, then so, too, is the implementation of an optimal monetary policy. Hayek's articulation of the knowledge...