The Fiction of An Economic “Rationality”
Cornelius Castoriadis (1991) ‘Philosophy, Politics, Autonomy,’ Odeon, pp 187-191
Castoriadis’ short essay written some 20 years ago is still relevant to our situation today in that it uncovers certain illusions about economics that are perpetually foisted onto us in a wide range of types of discourse—from the Financial Times to the pronouncements of think tanks every time statements are made about GNP or how we might assess ‘the economy’. It would be interesting to read someone refuting what was said below.
Perhaps it is not difficult to understand why it is that the economy has for two centuries been considered as the realm and paradigm of “rationality” in human affairs. Its subject matter is what has become the central activity of society, its discourse to prove (and for opponents like Marx, to disprove) the idea that this activity is achieved in the best possible manner in the framework of, and by means of the existing social system. But also—by a happy “accident”—the economy provided the apparent possibility of mathematization, since it is the only field of human activity where phenomena appear to be measurable in a manner which is not trivial, and even where this “measurability” seems to be—and to a certain extent is—the essential aspect in the eyes of the human agents concerned. The economy deals in “quantities”; on this point the economists have always fallen into line (though from time to time they have been forced to discuss the question: Quantities of what?). So, economic phenomena seem to lend themselves to an “exact” treatment and one which is amenable to the application of mathematical tools, the tremendous effectiveness of which has been demonstrated day after day in physics.
Within this domain, identifying the maximum (or extremum) and the optimum seems the obvious thing to do—and it has quickly been done. There was a product to maximize, and costs to minimize. Thus there was a difference to maximize: the net saleable product for the firm, the net “surplus” for the overall economy (“surplus” appearing under the guise of “goods” or of the growth in “leisure” measured in “free time,” without consideration of the use or content of this “free time”).
But what is the “product,” and what are the “costs”? Nuclear bombs are included in the net product—because the economist “is not concerned with use values.” Equally included are the costs of publicity, by means of which the people are induced to buy the junk which otherwise they probably would not buy; and of course, this junk itself. There are also the expenses accrued from having Paris cleaned of industrial soot; and for every road accident, the net national product is increased on several scores. It is equally augmented every time that a firm decides to nominate an extra vice-president drawing a substantial salary (because, ex hypothesi, the firm would not have nominated him if his net marginal product was not at least equal to his salary). More generally, the “measure” of a product reflects the valuations of various objects and of various types of work performed in the existing social system-valuations which themselves, of course, in their turn reflect the existing social structure. GNP is what it is also because a business manager earns twenty times what a street-sweeper earns. But even if these valuations are accepted, the measurability of economic phenomena, trivialities apart, is only a misleading appearance. The “product,” on any definition, is measurable “instantaneously,” in the sense that one can always add up, for the whole of the economy and at a given moment, the quantity of produced goods multiplied by the corresponding prices. But if the relative prices and/or the composition of the goods changes (which, in fact, is always the case) the successive “measurements” taken at different moments in time cannot be compared (any more than they can be compared between different countries, for the same reason). Strictly speaking, the expression “growth in GNP” is nonsense, except, and only, in the fictional case where there is a similar expansion in all types of products. Particularly, in an economy undergoing technical change, “capital” cannot be measured in any way which makes sense, except by means of ad hoc hypotheses which are highly artificial and contrary to the facts.
All this immediately leads to the conclusion that it is equally impossible to really measure “costs” (since the “costs” for one are for the most part the “products” of another). There are other reasons why “costs” cannot be measured: because we cannot apply the classical idea of imputing one part of the net product to this or that “factor of production,” and/or this product to this arrangement of the means of production. Imputing parts to “factors of production” (labor and capital) involves postulates and decisions which largely go beyond the domain of the economy. Imputing costs to a given product cannot be done because of various types of indivisibility (which the classical and neoclassical economists treat as exceptions, though they are present everywhere), and because of the existence of all sorts of “externalities.” “Externalities” signify that the “cost for the firm” and the “cost for the economy” do not coincide, and that a nonattributable (positive or negative) surplus appears. What is even more important, these “externalities” are not confined within the economy as such.
We are accustomed to think of most of the environment (its totality, with the exception of land under private ownership) as a “free gift of nature.” Similarly, the social framework, general learning, the behavior and motivations of individuals are implicitly treated as “free gifts of history.” The environmental crisis has only made obvious something which was always true (as Liebig knew over a century ago): an “appropriate state” of the environment is not a “free gift of nature” in all circumstances and without regard to the type and to the expansion of the economy considered. Nor is it a “good” to which one can assign a “price” (real or “dual”)—since, for example, no one knows the cost of refreezing the polar icecaps, should they melt. And the case of non “developing” countries shows that we cannot treat Judaism, Christianity, and Shintoism as “free gifts of history”—since history made a “gift” to other peoples of Hinduism and fetishism, which up till now appear rather as “obstacles to development,” given freely by history.
Behind all this can be found the hidden hypothesis of total separability, as much within the domain of economics as between this domain and historical, social, and even natural processes. Political economy always supposes that it is possible, without absurdity, to separate the consequences of action X of firm A from the total flux of economic processes internal and external to the firm; as it also supposes that the effects of the presence or absence of a given “total” of “capital” and of “labor” can be separated from the rest of human and natural life in a meaningful fashion. But the moment we abandon this hypothesis the notion of an economic calculus, in other than a trite sense, collapses—and with it, the notion of economic “rationality” in the accepted sense of the term (as the achievement of an extremum or a family of extrema) as much at the theoretical level (the comprehension of facts) as at the practical level (the definition of an “optimal” political economy).
What is at stake here is not simply the “market economy” or “private capitalism,” but the “rationality,” in the sense just indicated, of the economy as such (of any expanding economy). This is because the ideas underpinning what I have just said apply as much to “nationalized” and “planned” economies, literally or mutatis mutandis.
To illustrate this last point, I will use another example, which touches upon the fundamental question of time. Time is taken into account in political economy only inasmuch as it can be treated as non-time, as a neutral and homogeneous medium. An expanding economy implies the existence of (“net”) investment, and investment is intimately related to time, since in investment the past, the present, and the future are brought into mutual relation. Now, decisions concerning investment can never be “rational,” except at the level of the firm and providing one takes a particularly narrow point of view. There are many reasons for this, of which I will mention only two. First, not only is the future “uncertain,” but the present is unknown (things are constantly happening everywhere, other firms are in the process of making decisions, information is partial and costly, and this to different degrees for different actors, etc.). Second, as already mentioned, the costs and the product cannot really be measured. The first factor may, in theory, be eliminated in a “planned” economy; the second could never be.
But, in any case, a much more important question arises: What is the correct overall rate of investment? Should society devote to (“net”) investment 10, 20, 40, or 80 percent of the (“net”) product? The classical response, for “private” economies, was that “the” interest rate constitutes the balancing factor between the supply and demand of savings, and is consequently the appropriate “regulator” of the rate of investment. As we know, this response is pure nonsense. (“The” interest rate does not exist; it is not possible to assume that the rate of interest is the main determinant of total savings, that price levels are stable, etc.) Von Neumann proved, in 1934, that, given certain hypotheses, the “rational” interest rate must be equal to the rate of growth of the economy. But what should the rate of growth be? Supposing that this rate of growth is a function of productive capacity, and knowing that this rate depends upon the rate of investment, we arrive back at the original question: What should the rate of investment be? We can make the additional hypothesis that the “planners” are set on maximizing “final consumption” for a given period. The question then becomes: What is the rate of investment which will maximize (under complementary hypotheses about the “physical productivity” of additional capital) in a “permanent” or “steady state” the integral of “final consumption” (be it individual or public, of “goods” or of “leisure”)? The value of this integral depends, of course, upon the interval of integration—which is to say, upon the temporal horizon which the “planners” have decided to take into account. If it is “instantaneous” consumption that is to be maximized (temporal horizon at zero distance) then the appropriate rate of investment is obviously zero. If consumption to be maximized is “forever” (temporal horizon at infinite distance) the appropriate rate of investment is nearly 100 percent of (“net”) product—assuming that the “marginal physical productivity” remains positive for all corresponding values of investment. Any answer which “makes sense” obviously lies between these two limits; but where exactly, and why? No “rational calculus” exists which can show that a temporal horizon of 5 years is (for society) less or more “rational” than one of 100 years. The decision would have to be made on the basis of considerations other than “economic” ones.
All this does not mean that everything that happens within the economy is “irrational” in the positive sense, still less that it is unintelligible; but it means that we cannot treat an economic process as a homogeneous flux of values, of which the only relevant aspect would be that they can be measured and ought to be maximized. This type of “rationality” is secondary and subordinate. We can make use of it in order to clear part of the terrain, to scotch some obvious absurdities. But the factors which today effectively fashion reality—among them, the decisions of governments, of firms, and of individuals—are not susceptible to this sort of treatment. And, in a new, alternative society, they would be of a completely different nature.