Sample chapter from Julian Simon




The Great Toy Shortage

Forget it, Virginia. Santa won’t be leaving a “Star Wars” R2-D2 doll under the tree this year – just an I.O.U. promising you one at some vague time between February and June. Don’t count on a Mego Micronaut kit for building your own robot either, or a Milky the Marvelous Milking Cow, which drinks water when its tail is pumped, moos plaintively and squirts a tiny pailful of cloudy white “milk” from a detachable pink udder….

Not since the Grinch stole Christmas has there been such an unseasonable shortage.

Newsweek, December 19, 1977, p. 58


  • The Great Toy Shortage
  • Between Pig Copper and Dentistry
  • What Do We Mean by “Scarcity”?
  • What is the Best Way to Forecast Scarcity and Costs?
  • Will the Future Break With the Past?
  • A Challenge to the Doomsdayers to Put Their Money Where Their Mouths Are
  • Summary
  • Afternote 1: The “True” Cost (Price) of Natural Resources
  • Afternote 2: The Ultimate Shortage

The “Great Toy Shortage” of 1977 clearly was a freak event. We don’t worry that a scarcity of Hula-Hoops, pencils, dental care, radios, or new musical compositions will last. And we don’t fear that a larger population will reduce the supply of these goods; manufacturers will make more. Yet people do worry about an impending scarcity of copper, iron, aluminum, oil, food, and other natural resources. According to a typical pronouncement by Paul Ehrlich, the best-known contemporary doomster, “In the early 1970s, the leading edge of the age of scarcity arrived. With it came a clearer look at the future, revealing more of the nature of the dark age to come”. That we are entering an age of scarcity in which our finite natural resources are running out, that our environment is becoming more polluted, and that population growth threatens our civilization and our very lives – such propositions are continually repeated with no more evidence than that “everyone knows” they are true. Is there a fundamental economic difference between extractive natural resources and Hula-Hoops or dental care? Why do people expect that the supply of wheat will decline but the supplies of toys and drugs will increase? These are the questions that this chapter explores. The chapter draws examples from the metallic raw materials, which are relatively unencumbered by government regulations or international cartels and which are neither “burned up” like oil nor grown anew like agricultural products. Energy, food, and land will be given special treatment in later chapters.


There is an intuitive difference between how we get Hula-Hoops and copper. Copper comes from the earth, whereas a Hula-Hoop does not seem to be a “natural” resource. Copper miners go after the richest, most accessible lodes first. Therefore, they dig into lodes bearing successively lower grades of ore. If all else were equal, this trend would imply that the cost of extracting copper from the ground must continually rise as poorer and less accessible lodes are mined. Hula-Hoops and dental care and radios seem different from copper because most of the cost of a radio, a Hula-Hoop, or dental care arises from human labor and skill, and only a small part arises from the raw material – the petroleum in the plastic hoop or the silver in the tooth filling. For good reason we do not worry that human labor and skill comes from progressively less accessible reservoirs. But all this neat theorizing about the increasing scarcity of minerals contradicts a most peculiar fact: Over the course of history, up to this very moment, copper and other minerals have been getting less scarce, rather than more scarce as the depletion theory implies they should. In this respect copper follows the same historical trend as radios, undershirts, and other consumer goods (figures 1-1a and 1-1b). It is this fact that forces us to go beyond the simple theory and to think more deeply about the matter.

At the end of this confrontation between theory and fact, we shall be compelled to reject the simple Malthusian depletion theory, and to offer a new theory. The revised theory will suggest that natural resources are not finite in any meaningful economic sense, mind-boggling though this assertion may be. That is, there is no solid reason to believe that there will ever be a greater scarcity of these extractive resources in the long-run future than now. Rather, we can confidently expect copper and other minerals to get progressively less scarce.


Here we must pause for an unexciting but crucial issue, the definition of “scarcity.” Ask yourself: If copper – or oil or any other good – were much scarcer today than it actually is, what would be the evidence of this scarcity? That is, what are the signs – the criteria – of a raw material being in short supply?

Upon reflection perhaps you will not expect a complete absence of the material as a sign of scarcity. We will not reach up to the shelf and suddenly find that it is completely bare. The scarcity of any raw material would only gradually increase. Long before the shelf would be bare, individuals and firms – the latter operating purely out of the self-interested drive to make profits – would be stockpiling supplies for future resale so that the shelf would never be completely bare. Of course the price of the hoarded material would be high, but there still would be some quantities to be found at some price, just as there always has been some small amount of food for sale even in the midst of the very worst famines.

The preceding observation points to a key sign of what we generally mean by increasing scarcity: a price that hasĀ  persistently risen. More generally, cost and price – whatever we mean by “price,” and shortly we shall see that that term is often subject to question – will be our basic measures of scarcity. In some situations, though, prices can mislead us. Governments may prevent the price of a scarce material from rising high enough to “clear the market” – that is, to
discourage enough buyers so that supply and demand come to be equal, as they ultimately will be in a free market. If so, there may be waiting lines or rationing, and these may also be taken as signs of scarcity. But though lines and rationing may be fair ways of allocating materials in the short run, in the longer run they are so wasteful that every sort of society tends to avoid them by letting the price rise enough to clear the market.

So increased scarcity of a raw material implies a higher price. But the converse need not be true; the price may rise even without a “true” increase in scarcity. For example, a strong cartel may successfully raise prices for awhile, as OPEC did with oil in 1973 even though the cost of producing oil remained unchanged. This suggests that, in addition to the price in the market, we should sometimes consider production costs as an index of scarcity. There are yet other reasons why price does not tell the full story about scarcity and our welfare. A product may be readily available, as measured by its price being low, and yet there may still be a problem. For example, a daily ration of vitamin X may be very cheap, but if people are not getting enough of it there is a problem. On the other hand, caviar may be unusually high-priced and scarce this year, but few would consider that to be a problem for society. Similarly, the price of a staple food such as grain may be higher than in a previous year, but this may not indicate a problem – if, for example, the price rises because of an increase in income and a resulting increase in the amount of grain fed to animals to produce meat. So, though
the price of food and social welfare are often connected, they are not identical.

A more personal – but often relevant – test of scarcity is whether you and I and others feel that we can afford to buy the material. That is, the relationship between price and income may matter. If the price of food stays the same but income falls sharply, then we feel that food is more scarce. By a similar test, if our wages rise while the price of oil remains constant, our fuller pockets lead us to feel that oil is getting less scarce.

A related test of scarcity is the importance of the material in your budget. You are not likely to say that salt has gotten appreciably more scarce even if its price doubles, because it accounts for an insignificant share of your expenditures.

So price, together with related measures such as cost of production and share of income, is the appropriate operational test of scarcity at any given moment. What matters to us as consumers is how much we have to pay to obtain goods that give us particular services; from our standpoint, it couldn’t matter less how much iron or oil there “really” is in the natural “stockpile.” Therefore, to understand the economics of natural resources, it is crucial to understand that the most appropriate economic measure of scarcity is the price of a natural resource compared to some relevant benchmark. Future scarcity is our interest. Our task, then, is to forecast future prices of raw materials.


There are two quite different general methods for forecasting costs of any kind: the economist’s method and the technologist’s (or engineer’s) method. The engineering method is commonly used in discussions of raw materials, but I shall argue that the conclusions about costs reached with it are usually quite wrong because it is not the appropriate method. With the technical engineering method, you forecast the status of a natural resource as follows: (1) Estimate the presently known physical quantity of the resource, such as copper in the earth that’s accessible to mining; (2) extrapolate the future rate of use from the current use rate; and (3) subtract the successive estimates of use in (2) from the physical “inventory” in (1). (Chapter 2 discusses technical forecasts in greater detail.) In contrast, the economist’s approach extrapolates trends of past costs into the future. My version of the economist’s method is as follows: (1) Ask: Is there any convincing reason to think that the period for which you are forecasting will be different from the past, going back as far as the data will allow? (2) If there is no good reason to reject the past trend as representative of the future as well, ask: Is there a reasonable explanation for the observed trend? (3) If there is no reason to believe that the future will be different than the past, and if you have a solid explanation for the trend – or even if you lack a solid theory, but the data are overwhelming – project the trend into the future. Given the wide disparity between the engineering and economic approaches, it behooves us to consider the conditions under which each is likely to be valid.

The forecasting situation is analogous to that of a businessperson who wishes to estimate costs of some piece of construction or production for the firm. Making sound cost estimates is the businessperson’s bread and butter, the difference between solvency and bankruptcy. The choice between the economic or the engineering method depends largely on whether the business has much or little experience with the type of project whose costs it wants to assess. Examples of jobs with which the organization already has a great deal of experience are (a) a construction firm preparing a bid on a small parking lot like many others that the firm has done, and (b) a chain franchising operation estimating the cost of adding another hamburger shop. In such cases the firm will, as a matter of course, estimate costs directly from its own records. In the case of the new hamburger shop, the estimate may simply be the average total cost of shops recently built, as already computed by the firm. In the case of the parking lot, the construction firm may be able to estimate with good accuracy the amounts required of the main components – labor and machine time – and their current prices. It is only when the firm does not have direct experience with the type of project being costed that it must and will make an engineering analysis of the project, together with estimates of the requirements for each element of the job. But the businessperson, and the academic analyst of natural resources, too, should turn to engineering cost-estimates only in the unfortunate absence of reliable data from the past, because engineering estimates are much harder to make accurately, for a variety of reasons. An analogy may help. In forecasting the speed of a three-year-old race horse, would you rely more on a veterinary and anatomical inspection of the horse’s bones and organs, or on the results of its past races?

Mercury is an example of how one can go wrong with a nonhistorical, engineering forecast. Figure 1-2 shows a forecast made in 1976 by natural-scientist Earl Cook. He combined a then-recent upturn in prices with the notion that there is a finite amount of mercury on the earth’s surface, plus the mathematical charm of plotting a second-degree polynomial with the computer. Figures 1-3a and 1-3b show how the forecast was almost immediately falsified, and price continued its long-run decline. (Figure 1-3 also shows that reserves of mercury have increased rather than decreased over the
years, just the opposite of naive “finitist” ideas).

FIGURE 1-2[1-x from Cook], Figures 1-3a and 1-3b [1-y from Gunther] Fortunately, we don’t have to rely on such arbitrary guesswork. Considerable data showing trends in raw-material prices for a century or two, or longer, are available, as seen in various chapters and the Appendix to this book. Costs of extractive materials clearly have fallen over the course of recorded price history. The economist’s first-approximation forecast is that these trends toward less scarcity should continue into the foreseeable future unless there is some reason to believe that conditions have changed – that is, unless there is something wrong with the data as a basis for extrapolation. In brief, the economist’s and businessperson’s approach to price trends is to learn by experience – relevant experience. As P. T. Bauer put it, “it is only the past that gives us any insight into the laws of motion of human society and hence enables us to predict the future”.


Some people believe that we are now at a long-run turning point, and the past is not a guide to the future. Therefore we ask: How should one judge whether a historical trend is a sound basis for a forecast? Specifically, how can we judge whether the data from the many decades in the past showing declines in raw-material costs are a good basis for prediction? This question is a problem in scientific generalization. A sound principle is that you should generalize from your data if you can reasonably regard them as a fair sample of the universe about which you wish to draw conclusions. But prediction is a special type of generalization, a generalization from past to future. Prediction is always a leap of faith; there is no scientific guarantee that the sun will come up tomorrow. The correctness of an assumption that what happened in the past will similarly happen in the future rests on your wise judgment and knowledge of your subject matter.

A prediction based on past data is sound if it is sensible to assume that the past and the future belong to the same statistical universe, that is, if you can expect conditions that held in the past to remain the same in the future. Therefore, we must ask: Have conditions changed in recent years in such manner that the data on natural resources generated over the many past decades are no longer relevant? The most important elements in raw-material price trends have been (1) the rate of movement from richer to poorer ores and mining locations – that is, the phenomenon of “exhaustion”, and (2) the continued development of technology, which has more than made up for exhaustion. Is the rate of development of such new technology slowing up? To the contrary: the pace of development of new technology seems to be increasing. Hence if the past differs from the future, the bias is likely to be in the direction of understating the rate at which technology will develop, and therefore underestimating the rate at which costs will fall. The fall in the costs of natural resources, decade after decade and century after century, should shake us free from the idea that scarcity must increase sometime. And please notice that current prices do not mislead us about future scarcities. If there is reason to judge that the cost of obtaining a certain resource in the future will be much greater than it is now, speculators will hoard that material to obtain the higher future price, thereby raising the present price. So current price is our best measure of both current and future scarcity (more about this later). Figure 1-1(a and b) and others in the Appendix show the fundamental economic facts about natural resources. The costs and prices of most natural resources have been going down rather than up since at least 1800. But as noted earlier, cost and price can be slippery ideas and therefore require technical discussion in an afternote to this chapter. Here I’ll say but a few words on the matter.

The basic measure of the cost of, say, copper, is the ratio between the price of copper and the price of another product. One such measure – the most important measure – of the terms of trade is the price of copper relative to wages, as shown in figure 1-1a. This price has declined very sharply. This means that an hour’s work in the U.S. has purchased increasingly more copper from 1800 to the present. The same trend has almost surely held throughout history for copper, and other raw materials. Consider a very different measure of cost: the decreasing price of copper relative to nonextractive products. This means that it now takes smaller quantities of haircuts or underwear to buy a ton of copper than it did a century ago. The relative price between copper and other products is like a price that has been adjusted for the cost of living. We must remember that these other, nonextractive items also have been produced progressively more cheaply over the years. Yet minerals have declined in price even faster than have the other products. Still another way to think about the cost of natural resources is the proportion of our total income we must pay to get them. This measure also reveals a steady decline in cost. The absolute physical quantities of natural resources extracted have been rising, decade after decade, century after century, and the kinds of resources used have been increasing in number. But our expenditure on them has been falling as a proportion of total expenditures. “The gross value of extractive output [including agriculture, oil, and coal] relative to value of national product has declined substantially and steadily from 1870 to the present. In 1890, the extractive share was nearly 50 percent. By the turn of the century, it had fallen to 32 percent, and, by 1919, to 23 percent. In 1957, the figure was 13 percent and still trending downward”; by 1988 the figure had fallen all the way to 3.7 percent. In 1988, minerals plus energy (but excluding food) accounted for only 1.6 percent of the U.S. GNP, and minerals alone (excluding energy sources) accounted for less than half a percent of total extractive value, or only about an amazing .0002 (a fiftieth of a percent of GNP. This trend makes it clear that the cost of minerals – even if it becomes considerably higher, which we have no reason to expect – is almost irrelevant to our standard of living, and hence an increased scarcity of minerals is not a large danger to our peacetime standard of living.

A vivid illustration of the changing role of natural resources is a floppy computer disk: with a standard word processing program on it, it sells for $450. It embodies only a cent’s worth of petroleum and sand. This is the closest-to-home indicator of changing raw-materials costs. Every measure leads us to the same conclusion, but these calculations of expenditures for raw materials as a proportion of total family budgets make the point most strongly. Raw materials have been getting increasingly available – less scarce – relative to the most important element of life, human work time. Another piece of evidence confirms the price data. Figure 1-4 shows the total world “known reserves” of copper, and the ratio of reserves to total production. The figures show that reserves not only do not go down, they even go up – just as the “reserves” in your pantry go up when you are expecting company. The explanation for this anti-commonsensical trend may be found in chapter 2. Figure 1-4. Total Copper Reserves, and as a Ratio to Consumption Taken together, the various data suggest the anti-intuitive conclusion that, even as we use coal and oil and iron and other natural resources, they are becoming less scarce. Yet this is indeed the appropriate economic way of viewing the situation. (To explore further these measures of price and scarcity, see Afternote 1 to this chapter.)

In lectures on the foregoing section I commonly ask the audience to raise their hands to indicate whether they think natural resources have been getting more or less scarce. Then I show the price data. One woman responded that I had “tricked” the audience by drawing their attention away from the physical aspects of depletion. But it is the inappropriate concepts of physical science that have “tricked” this woman and many of the prominent doomsters. Like Cook predicting a scarcity of mercury, they are led into error by the seductive but unsound theory of “natural” physical limits that draws their attention away from the experience of falling raw-material prices throughout history.


Some respond that these price data are “old” and do not reflect what happened last year or last week, and that the long-term trends no longer hold. There is no way of conclusively proving them wrong. But the long, sad history of alarming analyses that run against previous long-term trends shows the alarms to be wrong far more often than simple extrapolations of the long-term trends. Forgive me for sounding strident, but talk is cheap, especially scare talk that
gets newspaper attention and foundation grants. Where I come from, when we feel that someone is talking without taking responsibility for the results, we say, “Put your money where your mouth is.” I am prepared to back my judgment with my own cash. If I am wrong about the future of natural resources, you can make money at my expense. If mineral resources such as copper will be more scarce in the future – that is, if the real price will rise – you can make money by buying the minerals now and selling them later at the higher prices. This is exactly what is done by speculators who
believe that the prices of commodities will rise (although for convenience’ sake, they really buy contracts for the commodities, or “futures,” rather than physical stocks). Please notice that you do not have to wait ten or twenty years to realize a profit, even if the expected changes in supply and demand will not occur for ten or twenty years. As soon as information about an impending scarcity becomes known and accepted, people begin buying the commodity; they bid up the present market price until it reflects the expected future scarcity. Current market prices thus reflect the best
guesses of professionals who spend their lives studying commodities, and who stake their wealth and incomes on being right about the future.

A cautionary example is how Japan let itself be panicked by fears of future scarcities in the wake of the OPEC oil embargo and the resulting general price rise in 1973.

The Japanese, and above all Japanese officialdom, were seized by hysteria in 1974 when raw materials shortages were cropping up everywhere. They bought and bought and bought [copper, iron ore, pulp, sulfur, and coking coal]. Now they
are frantically trying to get out of commitments to take delivery, and have slashed raw materials imports nearly in half. Even so, industrial inventories are bulging with high-priced raw materials.

Japan began paying heavily for this blunder.
page # \ultres\ tchaq01a December 23, 1993

Read the whole thing here:

Related posts: