Georgescu-Roegen - The Entropy Law & the Economic Process (1973)

[M]y point is not that arithmetization of science is undesirable. 'Whenever arithmetization can be worked out, its merits are above all words of praise. My polnt is that wholesale arithmetization is impossible, that there is valid knowledge even without arithmetization, and that mock arithmetization is dangerous if peddled as genuine. Let us also note that arithmetization alone does not warrant that a theoretical edifice is apt and suitable. As evidenced by chemistry -- a science in which most attributes are quantifiable, hence, arithmomorphic -- novelty by combination constitutes an even greater blow to the creed "no science without theory." (p.15)

The verdict is indisputable: no social science can subserve the art of government as efficaciously as physics does the art of space travel, for example. Nevertheless, some social scientists simply refuse to reconcile themselves to this verdict and, apparently in despair, have come out with a curious proposal: to devise means which will compel people to behave the way "we" want, so that "our" predictions will always come true. The project, in which we recognize the continual striving for a ''rational" society beginning with Plato's, cannot succeed (not even under physical coercion, for a long time) simply because of its blatant petitio principii: the first prerequisite of any plan is that the behavior of the material involved should be completely predictable, at least for some appreciable period. But aggressive scholarship will never run out of new plans for the "betterment of mankind." Since the difficulties of making an old society behave as we want it can no longer be concealed, why not produce a new society according to our own "rational" plans? (p.16)

It is fashionable nowadays to indulge in estimating how large a population our earth can support. Some estimates are as low as five billions, others as high as forty-five billions. However, given the entropic nature of the economic process by which the human species maintains itself, this is not the proper way to look at the problem of population. Perhaps the earth can support even forty-five billion people, but certainly not ad infinitum. We should therefore ask "how long can the earth maintain a population of forty-five billion people?'' And if the answer is, say, one thousand years, we still have to ask "what will happen thereafter?" All this shows that even the concept of optimum population conceived as an ecologically determined coordinate has only an artificial value. [...] Man's natural dowry, as we all know, consists of two essentially distinct elements: (1) the stock of low entropy on or within the globe, and (2) the flow of solar energy, which slowly but steadily diminishes in intensity with the entropic degradation of the sun. But the crucial point for the population problem as well as for any reasonable speculations about the future exosomatic evolution of mankind is the relative importance of these two elements, For, as surprising as it may seem, the entire stock of natural resources is not worth more than a few days of sunlight! [...] In a different way than in the past, man will have to return to the idea that his existence is a free gift of the sun. (p.20ff)

Anatomically, theoretical science is logically ordered knowledge. A mere catalog of facts, as we say nowadays, is no more science than the materials in a lumber yard are a house. Physiologically, it is a continuous secretion of experimental suggestions which are tested and organically integrated into the science's anatomy. In other words, theoretical science continuously creates new facts from old facts, but its growth is organic, not accretionary. Its anabolism is an extremely complex process which at times may even alter the anatomic structure. We call this process "explanation" even when we cry out "science does not explain anything." Teleologically, theoretical science is an organism in search of new knowledge. (p.37)

There can be no doubt that the decumulation of a machine is not a mechanical spreading in time of the machine as is the case with the stock of provisions of an explorer, for instance. When we "decumulate" a machine we do not separate it into pieces and use the pieces one after another as inputs until all parts are consumed. Instead, the machine is used over and over again in a temporal sequence of tasks until it becomes waste and has to be thrown away. A machine is a material stock, to be sure, but not in the sense the word has in "a stock of coal." If we insist on retaining the word, we may say that a machine is a stock of services (uses). But a more discriminating (and hence safer) way of describing a machine is to say that it is a fund of services. 

The difference between the concept of stock and that of fund should be carefully marked, lest the hard facts of economic life be distorted at everyone's expense. If the count shows that a box contains twenty candies, we can make twenty youngsters happy now or tomorrow, or some today and others tomorrow, and so on. But if an engineer tells us that one hotel room will probably last one thousand days more, we cannot make one thousand roomless tourists happy now. We can only make one happy today, a second tomorrow, and so on, until the room collapses. [...] The use of a fund (i.e., its "decumulation") requires a duration. Moreover, this duration is determined within very narrow limits by the physical structure of the fund. We can vary it only little, if at all. If one wishes to "decumulate" a pair of shoes, there is only one way open to him: to walk until they become waste (of course, one may sell the shoes. But this would mean decumulation of the shoes as a stock not decumulation of the shoes as a fund of services.) In contrast with this, the decumulation of a stock may, conceivably take place in one single instant, if we wish so. And to put the dots on all significant i's, let us also observe that the "accumulation" of a fund too, differs from the accumulation of a stock. A machine does not come into existence by the accumulation of the services it provides as a fund: it is not obtained by storing these services one after another as one stores winter provisions in the cellar. Services cannot be accumulated as the dollars in a saving account or the stamps in a collection can. They can only be used or wasted.

Nothing more need be said to prove that also the use of the term "flow" in connection with the services of a fund is improper if "flow" is defined as a stock spread over time. In fact, the generally used expression "the flow of services" tends to blur -- at times, it has blurred -- the important differences between two mechanisms, that by which the prices of services and that by which the prices of material objects are determined. The inevitable trap of this ambiguous use of "flow" is that, because a flow can be stored up, we find it perfectly normal to reason that services are "embodied" in the product. Only the materials that flow into a production process can be embodied in the product. The services of the tailor's needle, for example, cannot possibly be embodied in the coat -- and if one finds the needle itself embodied there it is certainly a regrettable accident. The fact that in certain circumstances the value of services passes into the value of the product is to be explained otherwise than by simply regarding a machine as a stock of services that are shifted one after another into the product. 

The difference between flow and service is so fundamental that it separates even the dimensionalities of the two concepts. For this reason alone, physicists would not have tolerated the confusion for long. The amount of a flow is expressed in units appropriate to substances (in the broad sense) -- say pounds, quarts, feet, etc. The rate of flow, on the other hand, has a mixed dimensionality, (substance)/(time). The situation is entirely reversed in the case of services. The amount of services has a mixed dimensionality in which time enters as a factor, (substance) x (time). If a plant uses one hundred workers during a working day (eight hours), the total of the services employed is eight hundred man x hour. If by analogy with the rate of flow we would like to determine the ratio of service for the same situation, by simple algebra the answer is that this rate is one hundred men, period. The rate of service is simply the size of the fund that provides the service and consequently is expressed in elemental units in which the time factor does not intervene. (p.226ff)

A leading symptom is that purists maintain that thermodynamics is not a legitimate chapter of physics. Pure science, they say, must abide to the dogma that natural laws are independent of man's own nature, whereas thermodynamics smacks of anthropomorphism. And that it does so smack is beyond question. But the idea that man can think of nature in wholly nonanthropomorphic terms is a patent contradiction in terms. Actually, force, attraction, waves, particles, and, especially, interpreted equations, all are man-made notions. Nevertheless, in the case of thermodynamics the purist viewpoint is not entirely baseless: of all physical concepts only those of thermodynamics have their roots in economic value and, hence, could make absolutely no sense to a nonanthropomorphic intellect.

A nonanthropomorphic mind could not possibly understand the concept of order-entropy which, as we have seen, cannot be divorced from the intuitive grasping of human purposes. For the same reason such a mind could not conceive why we distinguish between free and latent energy, should it see the difference at all. All it could perceive is that energy shifts around without increasing or decreasing. It may object that even we, the humans, cannot distinguish between free and latent energy at the level of a single particle where normally all concepts ought to be initially elucidated.

No doubt, the only reason why thermodynamics initially differentiated between the heat contained in the ocean waters and that inside a ship's furnace is that we can use the latter but not the former. But the kinship between economics and thermodynamics is more intimate than that. Apt though we are to lose sight of the fact, the primary objective of economic activity is the self-preservation of the human species. Self-preservation in turn requires the satisfaction of some basic needs-which are nevertheless subject to evolution. The almost fabulous comfort, let alone the extravagant luxury, attained by many past and present societies has caused us to forget the most elementary fact of economic life, namely, that of all necessaries for life only the purely biological ones are absolutely indispensable for survival. The poor have had no reason to forget it. And since biological life feeds on low entropy, we come across the first important indication of the connection between low entropy and economic value. For I see no reason why one root of economic value existing at the time when mankind was able to satisfy hardly any non biological need should have dried out later on.

Casual observation suffices now to prove that our whole economic life feeds on low entropy, to wit, cloth, lumber, china, copper, etc., all of which are highly ordered structures. But this discovery should not surprise us. It is the natural consequence of the fact that thermodynamics developed from an economic problem and consequently could not avoid defining order so as to distinguish between, say, a piece of electrolytic copper -- which is useful to us -- and the same copper molecules when diffused so as to be of no use to us. We may then take it as a brute fact that low entropy is a necessary condition for a thing to be useful. (p.277ff)

The corresponding symptoms in analytical studies are even more definite. First, there is the general practice of representing the material side of the economic process by a closed system, that is, by a mathematical model in which the continuous inflow of low entropy from the environment is completely ignored. But even this symptom of modern econometrics was preceded by a more common one: the notion that the economic process is wholly circular. Special terms such as roundabout process or circular flow have been coined in order to adapt the economic jargon to this view. One need only thumb through an ordinary textbook to come across the typical diagram by which its author seeks to impress upon the mind of the student the circularity of the economic process.

The mechanistic epistemology, to which analytical economics has clung ever since its birth, is solely responsible for the conception of the economic process as a closed system or circular flow. As I hope to have shown by the argument developed in this essay, no other conception could be further from a correct interpretation of facts. Even if only the physical facet of the economic process is taken into consideration, this process is not circular, but unidirectional. As far as this facet alone is concerned, the economic process consists of a continuous transformation of low entropy into high entropy, that is, into irrevocable waste or, with a topical term, into pollution. The identity of this formula with that proposed by Schrödinger for the biological process of a living cell or organism vindicates those economists who, like Marshall, have been fond of biological analogies and have even contended that economics "is a branch of biology broadly interpreted.". The conclusion is that, from the purely physical viewpoint, the economic process is entropic: it neither creates nor consumes matter or energy, but only transforms low into high entropy. (p.281)

Low entropy is a necessary condition for a thing to have value. This condition, however, is not also sufficient. The relation between economic value and low entropy is of the same type as that between price and economic value. Although nothing could have a price without having an economic value, things may have an economic value and yet no price. For the parallelism, it suffices to mention the case of poisonous mushrooms which, although they contain low entropy, have no economic value. (p.282)

we cannot mine the stock of solar energy at a rate to suit our desires of the moment. We can use only that part of the sun's energy that reaches the globe at the rate determined by its position in the solar system. With the stocks of low entropy in the earth's crust we may be impatient and, as a result, we may be impatient-as indeed we are with their transformation into commodities that satisfy some of the most extravagant human wants. But not so with the stock of sun's energy. Agriculture teaches, nay, obliges man to be patient-a reason why peasants have a philosophical attitude in life pronouncedly different from that of industrial communities. (p.297)

In a broad perspective we may say that mankind disposes of two sources of wealth: first, the finite stock of mineral resources in the earth's crust which within certain limits we can decumulate into a flow almost at will, and second, a flow of solar radiation the rate of which is not subject to our control. In terms of low entropy, the stock of mineral resources is only a very small fraction of the solar energy received by the globe within a single year. More precisely, the highest estimate of terrestrial energy resources does not exceed the amount of free energy received from the sun during four days! [...] because the low entropy received from the sun cannot be converted into matter in bulk, it is not the sun's finite stock of energy that sets a limit to how long the human species may survive. Instead, it is the meager stock of the earth's resources that constitutes the crucial scarcity. Let S be this stock and r the average rate at which it may be decumulated. Clearly, S = r x t, where t stands for the corresponding duration of the human species. This elementary formula shows that the quicker we decide to decumulate S, the shorter is t. Now, r may increase for two reasons. First, the population may increase. Second, for the same size of population we may speed up the decumulation of the natural resources for satisfying man-made wants, usually extravagant wants.

The conclusion is straightforward. If we stampede over details, we can say that every baby born now means one human life less in the future. But also every Cadillac produced at any time means fewer lives in the future. Up to this day, the price of technological progress has meant a shift from the more abundant source of low entropy-the solar radiation to the less abundant one--the earth's mineral resources. True, without this progress some of these resources would not have come to have any economic value. But this point does not make the balance outlined here less pertinent. Population pressure and technological progress bring ceteris paribus the career of the human species nearer to its end only because both factors cause a speedier decumulation of its dowry. The sun will continue to shine on the earth, perhaps, almost as bright as today even after the extinction of mankind and will feed with low entropy other species, those with no ambition whatsoever. For we must not doubt that, man's nature being what it is, the destiny of the human species is to choose a truly great but brief, not a long and dull, career.

"Civilization is the economy of power [low entropy]," as Justus von Liebig said long ago, but the word economy must be understood as applying rather to the problems of the moment, not to the entire life span of mankind. Confronted, in the distant future, with the impending exhaustion of mineral resources (which caused Jevons to become alarmed about the coal reserves), mankind -- one might try to reassure us -- will retrace its steps. The thought ignores that, evolution being irrevocable, steps cannot be retraced in history. (p.303ff)

[T]he usual denunciation of standard economics on the sole ground that it treats of "imaginary individuals coming to imaginary markets with ready-made scales of bid and offer prices" is patently inept. Abstraction, even if it ignores Change, is "no exclusive privilegium odiosum" of the economic science, for abstraction is the most valuable ladder of any science. In social sciences, as Marx forcefully argued, it is all the more indispensable since there "the force of abstraction" must compensate for the impossibility of using microscopes or chemical reactions. However, the task of science is not to climb up the easiest ladder and remain there forever distilling and redistilling the same pure stuff. Standard economics, by opposing any suggestion that the economic process may consist of something more than a jigsaw puzzle with all its elements given, has identified itself with dogmatism. And this is a privilegium odiosum that has dwarfed the understanding of the economic process wherever it has been exercised. (p.319)

The question is why a science interested in economic means, ends, and distribution should dogmatically refuse to study also the process by which new economic means, new economic ends, and new economic relations are created. (p.320)

[T]he immense satisfaction which Understanding derives from arithmomorphic models should not mislead us into believing that their other roles too are the same in both social and natural sciences. In physics a model is also "a calculating device, from which we may compute the answer to any question regarding the physical behavior of the corresponding physical system." [Bridgman, The Nature of Physical Theory] The same is true for the models of engineering economics. The specific role of a physical model is better described by remarking that such a model represents an accurate blueprint of a particular sector of physical reality. But [...] an economic model is not an accurate blueprint but an analytical simile. Economists are fond of arguing that since no model, whether in physics or economics, is accurate in an absolute sense we can only choose between a more and a less accurate model. Some point out also that after all how accurate we need to be depends on our immediate purpose: at times the less accurate model may be the more rational one to use. All this is perfectly true, but it does not support the further contention -- explicitly stated by Pareto -- that it is irrelevant to point out the inaccuracy of economic models. Such a position ignores an important detail, namely, that in physics a model must be accurate in relation to the sharpest measuring instrument existing at the time. If it is not, the model is discarded. Hence, there is an objective sense in which we can say that a physical model is accurate, and this is the sense in which the word is used in" accurate blueprint." In social sciences, however, there is no such objective standard of accuracy. Consequently, there is no acid test for the validity of an economic model. And it is of no avail to echo Aristotle, who taught that a model is "adequate if it achieves that degree of accuracy which belongs to its subject matter." One may always proclaim that his model has the proper degree of accuracy. Besides, the factors responsible for the absence of an objective standard of accuracy also render the comparison of accuracy a thorny problem. (pg.332ff)

From the deterministic viewpoint, the notion of "rational behavior" is completely idle. Given his tastes, his inclinations, and his temperament, the person who smokes in spite of the warning that "smoking may be hazardous to your health" acts from a definite ground and, hence, cannot be taxed as irrational. And if we accept the conclusions biologists have derived from the study of identical twins, that every man's behavior is largely determined by his genotype, then criminals and warmongers are just as "rational" as the loving and peaceful people. But for a determinist even nurture (whether ecological, biotic, or cultural) cannot be otherwise than what it is: together with nature, nurture holds the individual in a predetermined and unrelenting grip. This is probably why, when a social scientist speaks of irrational behavior, he generally refers to a normative criterion. Take the villagers in some parts of the world who for the annual festival kill practically all the pigs in the village. They are irrational-we say-not only because they kill more pigs than they could eat at one feast but also because they have to starve for twelve months thereafter. My contention is that it is well-nigh impossible to name a behavior (of man or any other living creature) that would not be irrational according to some normative criterion. This is precisely why to an American farmer the behavior of a Filipino peasant seems irrational. But so does the behavior of the former appear to the latter. The two live in different ecological niches and each has a different Weltanschauung. The student of man should know better than to side with one behavior or another. The most he can do is to admit that the two behaviors are different, search for the reasons that may account for the differences, and assess the consequences. (p.345ff)

Like the social insects, man lives in society, produces socially and distributes the social product among his fellows. But, unlike the social insects, man is not born with an endosomatic code capable of regulating both his biological life and his social activity. And since he needs a code for guiding his complex social activity in a tolerable manner, man has had to produce it himself. This product is what we call tradition. By tradition man compensates for his "birth defect," for his deficiency of innate social instincts. So, man is born with an endosomatic (biological) code but within an exosomatic (social) one. It is because of the endosomatic code that a Chinese, for example, has slanted eyes and straight hair. It is because of the exosomatic code that a Filipino peasant cultivates his fields in the manner all Filipino peasants do, participates in the extravagant festivals held by his village at definite calendar dates, and so on. A biological process sees to it that the pool of genes is transmitted from one generation to another. Tradition does the same for what we call "values" or, more appropriately, "institutions," i.e., the modes by which every man acts inside his own community. (p.359)