A review of Julian Simon's


by Herman Daly

From Steady State Economics (1991)

This book is an all-out attack on neomalthusian or limits-to-growth thinking and a plea for more population and economic growth, both now and into the indefinite future. It is not a shotgun attack. Rather it is an attack with a single-shot rifle aimed at a single (but critical) premise of the neomalthusian position.

If Simon hits the target, then neomalthusian arguments collapse. If Simon misses the target, then all neomalthusian first principles remain unscathed, and Simon's progrowth arguments collapse. The critical premise that Simon attacks is that of the finitude of resources, including waste absorption capacities. Other premises from which neomalthusians argue include the entropy law and the vulnerability of ecological life-support services.

Simon's theoretical argument against the finitude of resources is that:

"The word "finite" originates in mathematics, in which context we all learn it as schoolchildren. But even in mathematics the word's meaning is far from unambiguous. It can have two principal meanings, sometimes with an apparent contradiction between them. For example, the length of a one-inch line is finite in the sense that it bounded at both ends. But the line within the endpoints contains an infinite number of points; these points cannot be counted, because they have no defined size. Therefore the number of points in that one-inch segment is not finite. Similarly, the quantity of copper that will ever be available to us is not finite, because there is no method (even in principle) of making an appropriate count of it, given the problem of the economic definition of "copper," the possibility of creating copper or its economic equivalent from other materials, and thus the lack of boundaries to the sources from which copper might be drawn."

Two pages later he drives home the main point in connection with oil:

"Our energy supply is non-finite, and oil is an important example...the number of oil wells that will eventually produce oil, and in what quantities, is not known or measurable at present and probably never will be, and hence is not meaningfully finite."

The fallacy in the last sentence quoted is evident. If I have seven gallons of oil in seven one gallon cans, then it is countable and finite. If I dump one gallon of oil into each of the seven seas and let it mix for a year, those seven gallons would no longer be countable, and hence not "meaningfully finite," therefore infinite. This is straightforward nonsense.

The fallacy concerning the copper is obscured by the strange fact that Simon begins with a correct distinction regarding infinity of distance and infinity of divisibility of a finite distance, and then as soon as he moves from one-inch lines to copper with nothing but the word "similarly" to bridge the gap, he forgets the distinction. It would be a wonderful exercise for a class in freshman logic to find the parallel between Simon's argument and Zeno's paradox of Achilles and the tortoise. Recall that Zeno "proved" that Achilles could never catch up with a tortoise that had a finite head start on him. While Achilles traverses the distance from his starting point to that of the tortoise, the tortoise advances a certain distance, and while Achilles advances this distance, the tortoise makes a further advance, and so on, ad infinitum. Thus Achilles will never catch up.

Zeno's paradox confounds an infinity of subdivisions of a distance, which is finite, with an infinity of distance. This is exactly parallel to what Simon has done. He has confused an infinity of possible boundary lines between copper and noncopper with an infinity of amount of copper. We cannot, he says, make an "appropriate count" of copper because the set of all resources can be subdivided in many ways with many possible boundaries for the subset copper because resources are "infinitely" substitutable. Since copper cannot be simply counted like beans in a jar, and since what cannot be counted is not finite, it "follows" that copper is not finite, or copper is infinite.

Simon has argued from the premise of an "infinite" substitutability among different elements within a (finite) set to the conclusion of the infinity of the set itself. But no amount of rearrangement of divisions within a finite set can make the set infinite. His demonstration that mankind will never exhaust its resource base rests on the same logical fallacy as Zeno's demonstration that Achilles will never exhaust the distance between himself and the tortoise. Simon's argument therefore fails even if we grant his premise of infinite substitutability, which gets us rather close to alchemy. Copper is after all an element, and the transmutation of elements is more difficult than the phrase "infinite substitutability" implies! Indeed, Simon never tells us whether "infinite substitutability" means infinite substitutability at declining costs, constant costs, increasing costs, or at infinite costs! Of course Simon could simply assert that the total set of all resources is infinite, but this would be a bald assertion, not a conclusion from an argument based on substitutability, which is what he has attempted.

Simon appeals to the unlimited power of technology to increase the service yielded per unit of resource as further evidence of the essentially nonfinite nature of resources. If resource productivity (ratio of service to resources) were potentially infinite, then we could maintain an ever growing value of services with an ever smaller flow of resources. If Simon truly believes this, then he should join those neomalthusians who advocate limiting the resource flow precisely in order to force technological progress into the direction of improving total resource productivity and away from the recent direction of increasing intensity of resource use. Many neomalthusians advocate this even though they believe the scope for improvement is finite. If one believes the scope for improvement in resource productivity is infinite, then all the more reason to restrict the resource flow.

Those who are loud in their praise of Simon are the same people who would have bet on the tortoise, and are now betting on infinite resources. Simon's ultimate criterion for the validity of an argument seems to be willingness to "put your money where your mouth is." (See his grandstand offer on page 27 to bet anyone any amount, up to a $10,000 total, that the real price of any resource will not rise.) He suggests that the current heavy betting by speculators that the resource tortoise will stay ahead of the Achilles of demographic and economic growth is the best available evidence of the final outcome of the race. But it could in fact be the best available evidence that speculators are interested only in the short run, or that there is a sucker born every minute! In any case "put your money where your mouth is" is a challenge to intensity of belief, not correctness of belief. It is the adman's customary proof by bombastic proclamation.

But what about Simon's empirical evidence against resource finitude? It fares no better than his fallacious attempt at logical refutation. He leans heavily on two expert studies: "The Age of Substitutability" by Weinberg and Goeller (Science, February 20,1976), and Scarcity and Growth by Barnett and Morse.1 His use of these studies is amazingly selective.

From Weinberg and Goeller he quotes optimistic findings of "infinite" substitutability among resources, assuming a future low-cost, abundant energy source. This buttresses Simon's earlier premise of "infinite" subdivisibility or substitutability among resources. But it does not lend support to his fallacious conclusion that resources are infinite and therefore growth forever is possible. More to the point, however, is that Weinberg and Goeller explicitly rule out any such conclusion by stating in their very first paragraph that their "Age of Substitutability" is a steady state. It assumes zero growth in population and energy use at the highest level that Weinberg and Goeller are willing to say is technically feasible. And they express serious reservations about the social and institutional feasibility of maintaining such a high consumption steady state.

Furthermore, the levels envisioned by Weinberg and Goeller, though cornucopian by general consent, are quite modest by Simon's standards: world population in the Age of Substitutability would be only 2.5 times the present population, and world energy use would be only 12 times present use. This implies a world per-capita energy usage of only 70 percent of current U.S. per capita use. The very study that Simon appeals to for empirical support of his unlimited growth position explicitly rejects the notion of unlimited growth -- a fact that Simon fails to mention.

As further empirical evidence we are served a rehash of the Barnett and Morse study. Their finding was that the scarcity of most resources, as measure by per unit extractive costs and by relative prices, was decreasing rather than increasing from 1870 to 1957. Simon gives these arguments as evidence the resources are infinite.

There is no serious dispute about the Barnett and Morse numbers, but the conclusion that resources are becoming ever less scarce is hardly justified. The neomalthusians can reply that of course the prices of resources fall during a epoch of mineralogical bonanza. But the data cannot be decisive between these two views, since they cover only that epoch.

Barnett and Morse are careful to report an important exception to the general finding of falling resource prices: timber, whose price increased during the period. Simon's way of handling this exception is interesting. He first considers only mineral resources and applies the criterion of price as a measure of scarcity, explicitly rejecting all quantity-based indices. He thus shows, decline in scarcity of mineral resources. Later, in the context of food, he considers timber. This is a fair enough context, except that he switches his criterion of scarcity from price to quantity of timber growth. In this way he can show decreasing timber scarcity by applying quantity measures, while showing decreasing minerals scarcity by applying price measures.

But an equally shifty neomalthusian could use quantity remaining in the ground to prove increasing scarcity of minerals, and relative price to prove increasing scarcity of timber. There is a serious debate about the proper measure of scarcity, as the report by Resources for the Future, Scarcity and Growth Reconsidered,2 demonstrates, but Simon is not engaged in that serious discussion. He grabs whatever number may be moving in the direction that fits the needs of the argument at hand and baptizes it as an index of whatever he is talking about. Two examples will illustrate:

First, Simon claims, after warning us to "grab your hat," that pollution has really been decreasing rather than increasing. To test this hypothesis most investigators would probably look at parts per million of various substances emitted into the air and water by human activities to see if they have been rising or falling over time. Simon, however, takes life expectancy as his index of pollution: increasing life expectancy indicates decreasing pollution. If one suggests that the increase in life expectancy mainly reflects improved control of infectious diseases, Simon redefines "pollutant" to include the smallpox virus and other germs. In this way an increase in emissions of noxious substances from the economy (what everyone but Simon means by "pollution") would not register until after it more than offset the improvement in life expectancy brought about by modern medicine. Thus Simon "measures" pollution by burying it in an aggregate, the other component of which offsets and overwhelms it.

The second example is the claim (we are again told to grab our hats) that the combined increases of income and population do not increase "pressure" on the land. His proof: the absolute amount of land per farm worker has been increasing in the United States and other countries. One might have thought that this was a consequence of mechanization of agriculture and that the increasing investment per acre in machinery, fertilizer, and pesticides represented pressure on the land, not to mention pressure on mines, wells, rivers, lakes, and so on.

Simon's demonstration that resources are infinite is, in my view, a coarse mixture of simple fallacy, omission of contrary evidence from his own expert sources and gross statistical misinterpretation. Since everything else hinges on the now exploded infinite resources proposition, we could well stop here. But there are other considerations less central to the argument of the book that beg for attention.

If, Simon notwithstanding, resources are indeed finite, then the other premises of the neomalthusians remain in vigor. The entropy law tells us not only that coal is finite, but that you can't burn the same lump twice. When burned, available energy is irreversibly depleted and unavailable energy is increased along with the dissipation of materials. If nature's sources and sinks were truly infinite, the fact that the flow between them was entropic would hardly matter. But with finite sources and sinks, the entropy law greatly increases the force of scarcity.

Although the words "entropy" or "second law of thermodynamics" remarkably do not occur once in a 400-page book on The Ultimate Resource, the concept is occasionally touched upon. There is a comment made in passing that marble and copper can be recycled, whereas energy cannot. This raises hopes that Simon may not be ignorant of the entropy law. These hopes are soon dashed when he softens the statement to "energy cannot be easily recycled." Later he tells us that "man's activities tend to increase the order and decrease the homogeneity of nature. Man tends to bring like elements together, to concentrate them."

That is the only part of the picture that Simon knows about. But the entropy law tells us there is another part—that to increase order in one part of the system requires the increase of disorder elsewhere, and that in net terms for the system as a whole the movement is toward disorder. In other words, more order and more matter and energy devoted to human bodies and artifacts mean less matter and energy and less order for the rest of the system, which includes all the other species on whose life-support services we and our economy depend. Simon is quite prepared to ruin the habitats of all other species by letting them (and future generations) bear the entropic costs of disorders that our own continuing growth entails. For Simon, however, this problem cannot exist because he believes resources and absorption capacities are infinite. But after he has once mastered the paradox of Achilles and the tortoise concerning infinity, his next homework assignment should be to find out about entropy. Until he has done these two things he should stop trying to write books for grownups about resources and population.

Part II of the book is on population and is dedicated to the proposition that the ultimate resource is people. The more the better, indefinitely. We are told that: "Even the proposition that population growth must stop sometime may not be very meaningful (see Chapter 3 on 'finitude')." We have already seen Chapter 3 on finitude and have discovered that it is sheer nonsense. I will spare the reader a recitation of all the propositions about population that self-destruct with the demise of Chapter 3.

There is a puzzling methodological inconsistency between Parts I and II. In Part I Simon is the total empiricist, trusting only in the extrapolation of recent trends of falling resource prices. Any a priori argument from first principles about reversal of trends due to increasing cost, diminishing returns, the end of a bonanza, or even the S-shape of the logistic curve characteristic of all empirically observed growth processes simply does not warrant consideration by this hard-headed empiricist. Yet in Part II we find Simon refusing to project population trends and relying on the theory of demographic transition to reverse the recent trend of population growth. His own graphs, used to demonstrate the unreliability of past population predictions, also show that a simple linear trend would have yielded much more accurate predictions in the 1920s than did the then current "twilight of parenthood" theories. Once again, whatever epistemological posture serves the immediate needs of argument is adopted. One is certainly free to choose whatever balance of theory and empiricism one thinks is most effective in getting at the truth, but the balance should not fluctuate so wildly, so often, and so opportunistically.

Simon values human life. More people are better than fewer people because each additional person's life has value for that person, his loved ones, and for society as a whole should he turn out to be a genius: an increase of 4,000 people is more likely to yield another Einstein, Mozart, or Michelangelo than an increase of only 400 people.

While I personally give zero weight to the notion that more births among today's poor and downtrodden masses will increase the probability of another Einstein or Mozart (or Hitler or Caligula?), I do agree that, other things equal, more human lives, and more lives of other species, are better than fewer.

And I think that most of my fellow neomalthusians would agree than 10 billion people are better than 2 billion -- as long as the 10 billion are not all alive at the same time!

This is the crucial point: neomalthusian policies seek to maximize the cumulative total of lives ever to be lived over time, at a sufficient per-capita standard for a good life. Simon wants to maximize the number of people simultaneously alive -- and, impossibly, to maximize per-capita consumption at the same time. These two contradictory strategies are possible only if resources are infinite. If they are finite then maximizing the number of simultaneous lives means a reduction in carrying capacity, fewer people in future time periods, and a lower cumulative total of lives ever lived at a sufficient standard.

The difference is not, as Simon imagines, that he is "pro-life" and the neomalthusians are "anti-life." Rather it is that neomalthusians have a basic understanding of the biophysical world, whereas Simon still has not done his homework on Zeno's paradoxes of infinity, on the entropy law, on the importance of ecological life-support services provided by other species, and on the impossibility of the double maximization implied in his advocacy of "the greatest good for the greatest number."

Simon seems to believe that an avoided birth today implies the eternal nonexistence of a particular self-conscious person who would have enjoyed life. But as far as I know, the pairing of a particular self-consciousness with a particular birth is the greatest of mysteries. Perhaps birth control means that a particular existence is postponed rather than canceled. In other contexts, however, Simon proclaims that "birth control is simply a human right." When Kingsly Davis, Paul Ehrlich, or Garret Hardin advocate birth control they are sacrificing the unborn; but when Simon finds it convenient to his argument to endorse birth control, he is proclaiming a human right.

In this reviewer's opinion, Simon's book cannot stand up to even average critical scrutiny. Lots of bad books are written, and the best thing usually is to ignore them. I would have preferred to ignore this one, too, but judging from the publicity accorded Simon's recent articles, this book is likely to be hailed as a triumph by people who are starved for "optimism." Simon himself tells us that the optimistic conclusions he reached in his population studies helped to bring him out of a "depression of medically unusual duration," and he clearly wants to share the cure. But his cure is at best a sugar pill.

We must abandon the shallow, contrived optimism of growthmania once and for all. The end of growthmania is no cause for despair; it is a hopeful new beginning. To me the optimistic alternative is that of a steady state at a sufficient, sustainable level in which many future generations can rejoice in the loving study and care of God's creation.

Further prolongation of the current compulsive quest for infinite growth, power, and control is what I find depressing. We should learn to be good stewards of what is already under our dominion rather than seek always to enlarge that dominion. We who have done a poor job of managing a small domain should not trust ourselves to take over control of an ever larger "infinite" domain.

This review appeared originally in Bulletin of the Atomic Scientists, January 1982.


   1.  Harold Barnett, and Chandler Morse, Scarcity and Growth
       (Baltimore: Johns Hopkins Press, 1963)

   2.  V. Kerry Smith, ed., Scarcity and Growth Reconsidered
       (Baltimore: Johns Hopkins Press, 1979)

[back to previous page]