At the same time, engineering services are urgently needed all around the country. Engineers’ time and expertise is needed to design and build irrigation systems, mines, harbor installations, and thousands of other systems that will improve the country’s productivity. If you use a great deal of engineering building your railroad, the country will wait precious months for those other systems and their services.
So there is your problem: If, when you build your railroad line, the alternative uses of steel—the girders, pots and pans, vehicles, surgical instruments, etc.—are more important and pressing to the nation than the systems and services that need engineering, then you should put the engineers to work building the railroad through the mountains. That way you would save on the relatively scarce and precious steel, and use more of the relatively abundant engineering. By contrast, if the other uses of engineering—the irrigation systems, mines, harbor installations, etc.—are more important and pressing than the other uses of steel, then you should run the line around the mountains, to save on the relatively scarce and important engineering services and use more of the relatively abundant steel.
What would you decide?
* * * * *
When I ask this in lectures, most of my audience has no answer. And that’s the right answer.
Think what you would need to know in order make the judgment: In order to know the value of steel in other uses, you would need detailed information on the value of the uses to which it would be put. Consider a new hospital that might be built with new steel girders. What is its value? In order to answer that question, you would need to know what the various doctors and nurses and hospital administrators know about currently available space, the state of repair of existing hospitals, the benefits of the new location, and so on. In order to know the value of steel in making pots and pans, you would need to know what various householders and restaurateurs know about the condition of their existing pans, their expectations of need for more pans or pans of different sizes, their preferences for steel pots and pans as opposed to copper, and so on. You would need similar kinds of knowledge to assess the value of steel for manufacturing different kinds of cars and trucks.
Think how dispersed and interconnected is the information you would need with respect to even one vehicle. In order to know the value of, say, a new truck, you would need to know what the trucker knows about the value of the shipments he could make with the truck. In order to know the value of particular shipments, however, say, of produce to grocery stores in the region, you would need to know what the grocer knows about the value of fresh groceries on his shelves. In order to know that, you would need to know what the customers know about the value of the dinners they mean to make with those groceries.
The same kinds of considerations hold on the engineering side. What would you need to know in order to assess the value of engineering used to construct, say, an irrigation system? You would need to know what the farmers know about how much the yield of their fields would increase with irrigation. Of course, to know the value of that increased yield, the farmers need somehow to know what consumers know about the importance to them of the additional food produced.
In short, in order to make a sound assessment of whether steel or engineering is more important in other uses, and accordingly whether to build your railroad line from City A to City B through or around the mountains, you would need an overwhelming amount of detailed knowledge held by thousands, nay, millions of people throughout your society. How would you get that information? Would you send out surveys? As Commissar, you hold absolute power; you may execute anyone who does not tell you promptly what you ask.
But do people even know what they prefer until they are faced with an actual choice? Often they don’t, so they might not even be able to answer survey questions accurately. And how would you aggregate all this information, if you could get it all? How much time would it take to get answers back? And by the time you got it all and aggregated it, wouldn’t conditions have changed? Your information would constantly be going out of date. Isn’t it clear that you simply could not get the information you need in a manner timely enough to make it useful?
Furthermore, even if you could get complete and timely information about what everyone knows that’s relevant to the use of steel and engineering, you would still need to deduce from it where to build your railroad. How could you possibly know what all that information means for your decision? How would you begin to make sense of all that data?
If you could not say which route you would choose for the railroad from City A to City B, you gave the right answer. It is simply not possible to decide on any rational basis. In the words of Ludwig von Mises, who first pointed out this problem to the socialists, you would be “groping in the dark.” You would have to guess. Even though you have absolute power as Commissar of Railroads, and even though you have the best will in the world and a burning desire to make communism outperform capitalism, you would be simply unable to determine the best route, because you could not possibly know which route would be less costly to society overall. The knowledge you would need is too vast, too dispersed, too specialized, and too changeable; and too much of it is inarticulate anyway. (We discuss inarticulate knowledge below.)
This is the knowledge problem of central planning. It explains why comprehensive socialism must necessarily always fail: the central planners cannot get the knowledge they need in order to plan effectively.
Now, let’s change the thought experiment slightly. Instead of Commissar of Railroads in the old Soviet Union, you are now the chief operations officer of a for-profit railroad company somewhere in the capitalist West. You face the same problem. You want to run a railroad line from City C to City D, and there is a mountain range between them, so you must go either through or around. How would you decide on your route? Again, assume that all the other costs and benefits come out the same on both routes, so that the only variables to consider are the different amounts of steel and engineering you would use. Answer for yourself before you go on.
* * * * *
Unless you are quite unusual, you would decide according to what’s cheapest. You would calculate the total cost of each route, in each case multiplying the amount of steel required by the price of steel, and adding that to the amount of engineering required times the price of engineering. Whichever route gives the lower total is the one you’d choose.
Typical, greedy capitalist! All you care about is the company’s profits, the bottom line. Like capitalists the world over, you would give no consideration to the overall good of the nation. You would ignore the question of whether steel or engineering is more valuable in other uses, even though that question is crucial to the overall productivity and living standards of your society. You would just do whatever is cheapest, focusing on your company’s profits and ignoring the well-being of other people.
But—and here’s the marvel—in doing whatever is cheapest in a free economy, you unwittingly do take into consideration every single piece of information and every bit of human knowledge about the values of thousands of alternative uses of steel and engineering. You learn from examining a few numbers all that the central planner would be powerless to find out in months of investigation. And choosing the cheaper route saves what’s more valuable in other uses for those uses. How? Because all the relevant knowledge available is embodied in the prices of steel and engineering. The lower total cost of one route, as compared to the other, tells us that the combined total of steel and engineering needed for the cheaper route is less valuable in other uses than the combination needed for the more expensive route.
Suppose many existing hospitals are small and out of date (in a free market for hospital care). The stronger the desire for new space, the more health care companies and philanthropists will be willing to pay contractors for the erection of new hospitals; and therefore, the more the contractors will be willing and able to pay, if necessary, for the steel girders with which to build them. The price of steel,
likewise, reflects the number of householders desiring to buy new steel pots and pans and the urgency of their desires. It reflects the amount any trucker will pay for a new delivery truck, which reflects the value of the produce deliveries he makes, which reflects the value of the groceries on the shelves, which reflects the value of the final consumer’s dinner. The number and urgency of all these direct and indirect desires for steel is reflected in the price offered for steel, as the various would-be buyers compete for the limited quantity of steel.
The price of engineering at any time is determined in the same manner. If new irrigation systems would improve farm output only a little, then farmers would not be willing to pay much for the engineers’ time, and the price of engineering would be correspondingly lower.
This railroad example illustrates the essential point that market prices give us a reliable gauge of the value of productive resources. To see why this is so, consider how market prices are determined: A particular price at a particular time results from competition among buyers and among sellers. Competing bids from would-be buyers push the price upward; competing offers from would-be sellers push the price downward. Each would-be buyer makes his bids based on his particular knowledge of the value of the resource to him. In a kind of implicit auction, he bids as little as he can to still remain in the running, and he never bids more than the good’s value to him. At the same time, each would-be seller makes her offers to sell based on her own particular knowledge of her costs—what she must give up to provide the good for sale. She asks as high a price as she can and still remain in the running, and she never asks a price lower than what it costs her to offer the good for sale. Out of that ongoing competitive process of bids and offers emerges the ever-fluctuating market price of the resource. In economic terminology, this is the process of price determination by the interaction of supply and demand. Notice that no individual really sets the market price; it emerges out of the interaction of many individuals. As economist Antony Davies says, “Prices are metrics that reflect value, not levers that set value.”
The market price reflects the value to the last buyer just barely willing to pay that price, and the cost to the last seller just barely willing to accept that price. Thus the market price (at that moment—remember prices constantly fluctuate) is the actual value in society of one more unit of that resource—what (someone in) society gives up in providing one more, and the benefit (someone in) society realizes from gaining one more.
Stated differently, the market price tells us what one more unit of a good is worth to others. Only with this knowledge can we calculate the least costly ways of producing the things we want, and thereby satisfy as many different human wants as possible.
With market prices to guide us, we can make use of the knowledge of everyone in society about the supplies of and uses for every different resource. In George Selgin’s useful term, prices serve as knowledge surrogates. They distill the knowledge of millions of people into a number. With market prices to guide us, each of those millions of people can stay well coordinated with all the others, never using for a less-important purpose what someone else needs for a more important purpose. Without market prices, however, we would be helpless, unable to coordinate our innumerable activities.
The Nature of Knowledge
The necessity of market prices arises out of the nature of human knowledge. The great Austrian economist F. A. Hayek laid out the connection in his famous article, “The Use of Knowledge in Society.” Hayek distinguishes two broad categories of knowledge: scientific knowledge and local knowledge. Scientific knowledge is much more systematic. It can be written down. It is relatively enduring, though scientific ideas sometimes are overturned by a new paradigm, as in the shift from earth-centered to sun-centered astronomy. Scientific knowledge can be effectively shared, passed wholesale from one person to another.
While scientific knowledge is of course very important, it is not the only kind of knowledge. The other sort, the local knowledge on which we focus here, is equally or more important to social well-being. This is what Hayek calls “knowledge of the particular circumstances of time and place.” Real estate agents, for example, specialize in knowing what various houses are for sale in their areas at what prices. Shipping agents know what space for freight is available on different trucks or trains headed for different destinations on different days. Farmers must know the moisture content in their recently-cut hay in order to bale it at the proper time. Each person in the job market knows his or her own aptitudes and qualifications better than anyone else.
In each of these cases, some particular individual (or perhaps just a few individuals) has specialized knowledge about some particular productive resources, and how they might be used to greatest advantage. The knowledge is valuable, but it’s local. It can’t be looked up. It is available only to that individual or a few individuals.
Local knowledge has several characteristics that lead to the knowledge problem for central planning, and make prices necessary for communicating knowledge.
First, knowledge is dispersed. Think, for example, of the knowledge needed to make a pencil. It is dispersed from Northern California to Sri Lanka to Wilkes-Barre to Indonesia. More important, it is dispersed among thousands of individuals who specialize in everything from logging to mining to chemistry to metallurgy to operating a kiln. Hayek emphasizes that each of us has special knowledge, known only to the individual you or me; good use might be made of that special knowledge if only we can somehow get it into the system.
Second, knowledge is incomplete. We are always learning, filling in the gaps of our understanding, discovering something we did not know the day before. (We pursue this point at length in the next chapter.)
Third, much valuable human knowledge is inarticulate, or tacit. That is, we know it, but we can’t say precisely what we know. For a trivial example, we all know how to tie our shoes. But suppose we were asked to write down precisely what we do, so that someone who did not know how to tie her shoes could read and follow our instructions. Could we do it? Probably few of us could make better than a bad job of it. Of course, we know how to tie our shoes, but we don’t know it in such a way that we can put it into words. Our knowledge is tacit, inarticulate—it is in our fingers, not in our heads.
This point deserves emphasis because so much important knowledge is tacit. Consider the tacit knowledge of a skilled personnel officer, the sort of person who, year after year, manages to size people up in their interviews, assess how well they’ll fit into the enterprise, and consistently hire the right people for the various jobs. Picture that person coming up to retirement. The boss says to her, “Stephanie, we are sure going to miss you. We don’t want to lose your skills, however, so in your last few weeks here, would you please write down what you do to size people up so well, so that we can fully instruct your replacement?” What an absurd request! No one can write down how she sizes someone up. The knowledge is a matter of hunches, judgment, interpreting hundreds of non-verbal clues. It is almost entirely inarticulate.
Here is a final example, taken from Hayek, of how tacit knowledge is: What do we mean by the knowledge of the skilled engineer? It is the knowledge of how to solve problems. It is not the knowledge of how to solve some particular problem, because once a particular problem has been solved, the solution can be written down and passed on. No, the knowledge we value in a skilled engineer is knowledge of how to figure out how to solve new problems. Picture the focused face under the hard hat, looking at a new construction or reinforcement problem, or the software systems engineer faced with integrating incompatible programs and operating systems. What sort of knowledge does such a person draw on to solve such problems? Surely it is not the kind of knowledge that can be written down. It’s experience. It’s what Hayek calls a technique of thought. It cannot be articulated.
A fourth characteristic of local knowledge that leads to the knowledge problem for central planning is that much human knowledge is latent at any particular tim
e. That is, we know, but we are not aware that we know. In such cases, we do not become consciously aware of our knowledge until it is somehow brought to our attention. I remember my high school physics teacher telling our class that we all “knew” the Doppler Effect—that the sound made by a moving object seems higher pitched to us when the object is approaching than when it is moving away. We all drew a blank. He smiled and made the sound every child makes when imitating a fast car or airplane going past. Sure enough, the pitch goes from higher to lower. Of course I knew that, but I had not been aware that I knew it.
Because local knowledge is dispersed, incomplete, tacit, and latent, it is not possible for the people who have that knowledge to communicate it clearly to some central planner. If that knowledge is to be used effectively, the people who have it must make the decisions that depend on it.
In order for overall coordination to be maintained, each local decision-maker must somehow take into account the relevant local knowledge of everyone else. In our example of building the new railroad line, the chief operating officer of the railroad company must somehow take into account the local knowledge of all the users of steel and engineering, and that of the users of all the goods and services produced with steel and engineering. Prices give us that vast amount of local knowledge. Or, rather, they do not give us the knowledge itself—that is so vast that it would overwhelm us—but instead, prices serve as surrogates for that knowledge. They embody in that simple number, the price, all we need to know of the vast amount of local knowledge held by others.
Hayek uses the example of a change in the supply of, or demand for, tin to illustrate how prices communicate to all, in a manner usable by all, the special knowledge of all. The long passage deserves quoting in its entirety:
Free Our Markets Page 4
