Hayek on The Use of Knowledge in Society
I recently reread Hayek’s most famous paper, The Use of Knowledge in Society, written in 1945. I hadn’t read it in many years.
Page for page, TUKS is perhaps the most insightful economics paper ever written. Every undergraduate economics major should read it. I am not sure how many other papers written before 1960 are still essential reading.
Hayek’s basic point is that markets are essential for organizing economic activity because prices transmit information about the relative scarcity of different goods and give people an incentive to react to that information in a socially desirable way. If tin becomes more scarce due to a decrease in supply or increase in demand the price of tin rises, which gives everyone an incentive to economize on their use of tin. They do not need to know why tin became scarce. Prices just need to transmit the bare fact of increased scarcity, and let users adjust to this using their own knowledge of their purposes and alternatives.
The paper is short, only 12 pages. Remarkably, it could be much shorter. As usual, Hayek does not shy away from saying in a paragraph what could be said in a sentence, and much of the paper consists of Hayek’s speculative explanations for why his economist colleagues have not recognized the information transmitting/economizing role of prices. These speculations are interesting enough, but they are not really germane to the paper’s central point.
The paper does a better job making the case for markets than against central planning. At least, the case for markets is clearer. Interestingly, Hayek does not show that the market solution is efficient, or even close to efficient. He shows that changes in the use of resources that result from changes in conditions (shortages of tin, say) seem likely to be directionally correct: if the supply of tin goes down, its price goes up, and people all over the world have an incentive to economize on tin. This does not tell us much about the efficiency of the resulting allocation unless we know that the starting allocation was tolerably efficient. The tolerable efficiency of market allocation seems to follow from his analysis of how markets reallocate resources in response to changed conditions, but Hayek doesn’t say so explicitly, and you have to squint to see it.
If a student read this paper and nothing else, they would be able to explain why markets are useful for organizing economic activity. They would have a harder time explaining why central planning is unworkable.
As far as I can see, Hayek does not claim that central planners could not calculate an efficient allocation, even if they could collect all the dispersed information about “circumstances of time and place” that a sensible plan would rely on. What he does suggest is that the need for constant changes (as in the tin example) would overwhelm the ability of the planner to make adjustments in a timely way even if they could come up with a plan for a static economy.
No doubt both calculating an initial plan and updating it in a timely way are very serious obstacles to central planning. Still, there are important difficulties with central planning that Hayek’s analysis does not address. One well-known problem with planning is that firm managers do not honestly transmit information to the planners. Instead, they lie about their capabilities to get assigned easier targets. The planners anticipate this and try to make adjustments, but the information planners work with is highly distorted.
The severe computational and incentive difficulties of planning create an incentive for planners to have a smaller number of very large firms, and to limit trade to other centrally planned economies. A centrally planned economy cannot cope with the rapidly shifting demands generated by firms and consumers in a market economy. The result is pressure towards large firms and autarky.
Government ownership of firms under central planning also creates serious political problems. Weitzman and Kruse (I believe) point out one of the superpowers of capitalism is that it puts the authority to fire people in the hands of owners and managers who have a relatively strong incentive to dismiss workers who are unproductive or who become redundant due to efficiency improvements or falling demand. This is important because most of us dislike having to fire people, and no one likes being fired. Governments are responsive to worker demand for job protection, and, as a result, it is often difficult under central planning for managers to fire workers or planners to close firms. The result is that firms stay open even when they lose money (they have soft budget constraints), and labor and other resources are not reallocated to better uses. This is a purely political problem, layered on top of the informational problem Hayek focuses on.
Interestingly, Hayek does not emphasize the way the price system gives people an incentive to innovate, to generate new knowledge. He is primarily focused on how prices coordinate economic activity given current production possibilities.
Finally, there is nothing in this paper about the political dangers inherent in giving central planning authorities the powers they need to run an economy, which Hayek worried about in other work, notably The Road to Serfdom.
“[one of the superpowers of] capitalism [is that it] puts the authority to fire people in the hands of owners and managers who have a relatively strong incentive to dismiss workers who are unproductive”
“Governments are responsive to worker demand for job protection”
In capitalism a worker is considered unproductive if the worker is efficiently making stuff that nobody wants to buy. Since firing people reduces demand, the owners and managers power creates an unstable feedback mechanism.
Government ability to maintain demand is a useful stabilizing mechanism.
Some people think that continuous growth (baby boom, women entering the workforce, tech bubble, housing bubble, immigration) can provide the necessary counter, but I think we need a slightly larger government.
This is a “two (or three) things can be true at once” situation. Hayek can right about the essential role of markets, capitalism can have an advantage over a system in which government owns the means of production in part because capitalists have an incentive to get rid of unproductive/redundant workers that the government lacks, and government may have an important role to play in keeping aggregate demand at full employment levels.
How does this gibe with the fact that companies rely on internal cross subsidies in their operations? If one part of a company is producing tin but running into a processing problem, a company producing a product that relies on a reliable supply of tin will bury the cost increase lest it interfere with overall profitability. The entire point of the combined operations of a company is to enable central planning, ignore inconvenient pricing fluctuations and still produce a profit.
Obviously, if a company ignores important signals for too long, it could go bankrupt, but sometimes particular essential divisions will be cost centers while others, just as essential, are profit centers. Hayek makes an argument that prices provide useful information, but much of capitalism is based on the ability of companies to plan centrally and ignore or internally hide price information for the purpose of overall efficiency.
This has puzzled me for a long time. Somehow, prices are important for consumers and small businesses, but should be ignored within a broad range by larger organizations. If I were to channel Joan Robinson, I’d say that paeons to pricing are merely apologies for the excesses of big business. Since I am not channeling Joan Robinson, I’ll agree that prices are useful, but there’s a reason even Hayek didn’t come up with an argument against central planning. It’s too important in our economic structure.
(I’ve long been skeptical of all those arguments that central planning is bad because it requires solving a massive optimization problem. After all, companies solve massive central planning problems all the time, and no one accuses them of anything but rapacious capitalism. Remember, one doesn’t need an optimal solution, just a good enough solution.)